Verbal Prompting as a Method for Improving Everyday Cognition in MCI and Unimpaired Older Adults

MISSING IMAGE

Material Information

Title:
Verbal Prompting as a Method for Improving Everyday Cognition in MCI and Unimpaired Older Adults
Physical Description:
1 online resource (103 p.)
Language:
english
Creator:
Thomas, Kelsey R
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Psychology, Clinical and Health Psychology
Committee Chair:
Marsiske, Michael
Committee Members:
Bowers, Dawn
Pereira, Deidre B
Janicke, David M

Subjects

Subjects / Keywords:
aging -- cognition -- impairment -- prompting
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre:
Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
This study investigated the effect of verbal prompting on elders' 5-year longitudinal change in everyday cognition. Differential effects of prompting associated with impaired cognitive status (amnestic/non-amnestic) were also examined. At baseline, 2,802 participants (mean age=73.6 years, mean education=13.5 years, 26% African American, Mini Mental Status Exam > 23) from the ACTIVE clinical trial were classified as unimpaired, amnestic mild cognitive impairment (MCI) or non-amnestic MCI based on psychometric algorithm. Participants were given the Observed Tasks of Daily Living (OTDL; a behavioral measure with tasks like change making, phone dialing) at baseline and at 1-, 2-, 3- and 5-year follow-ups. When participants said "I don't know" or did not respond to an item, they received a standardized verbal prompt. At each occasion, Unprompted (sum of items correct without prompting) and Prompted (sum of items correct including both prompted and unprompted) scores were derived for each participant. Multi-level modeling, adjusting for demographics, health and training group, was used to determine the trajectories of OTDL performance. Results showed a linear and quadratic time pattern of initial gain followed by decline in later years across participants. The Prompted score was higher at all occasions and showed a flatter five-year trajectory than the Unprompted. Cognitive impairment was associated with poorer OTDL performance (amnestic < non-amnestic < unimpaired). Prompting interacted with impairment status, such that the impaired participants' Prompted scores improved more than those of the unimpaired participants. Random time effects were observed, indicating individual differences in rates of change. Brief verbal prompts boosted everyday cognitive performance, particularly in MCI participants, and seemed to produce scores that were less affected by practice or decline. Discussion focuses on how prompting might be incorporated into elders' daily lives to improve or maintain independent functioning.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Kelsey R Thomas.
Thesis:
Thesis (M.S.)--University of Florida, 2012.
Local:
Adviser: Marsiske, Michael.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-05-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044158:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 VERBAL PROMPTING AS A METHOD FOR IMPROVING EVERYDAY COGNITION IN MCI AND UNIMPAIRED OLDER ADULTS By KELSEY THOMAS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIR EMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Kelsey Thomas

PAGE 3

3 To my devoted, caregiving m other and m y truly courageous grandparents

PAGE 4

4 ACKNOWLEDGMENTS First, I thank my mentor, Michael Marsiske, whose constant gu idance, unwavering support, and continued patience has greatly impacted not only my direction, but the quality of my journey. His persistent prompting is greatly appreciated and has slowly but surely started to internalize I also thank my lab mates, Sh annon Sisco and Anna Yam for being warm and welcoming and making my transition to Florida so smooth. I extend a special thank you to Jared for his continued support and love. His genuine kindness and old man humor makes my life full of joy Lastly, I th ank my family especially my mom and brother, for always being there to talk to me on my drive home, keeping me grounded, and putting things into perspective.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPT ER 1 LITERATURE REVIEW ................................ ................................ .......................... 12 Overview ................................ ................................ ................................ ................. 12 Measures of Everyday Cognition ................................ ................................ ............ 12 Relationship between Well Structured Measures of Everyday Cognition and Everyday Functioning ................................ ................................ .......................... 16 Negative Age Association with Everyday Cognition ................................ ................ 16 Mild Cognitive Impairment ................................ ................................ ...................... 18 Relationship between Everyday Cognition and Cognitive Performance ................. 19 Preserving Everyday Functioning ................................ ................................ ........... 22 Summary ................................ ................................ ................................ ................ 25 2 STATEMENT OF PROBLEM ................................ ................................ .................. 27 Overview ................................ ................................ ................................ ................. 27 Number of Verbal Prompts Given on the OTDL by Cognitive Status ...................... 28 Aim 1 ................................ ................................ ................................ ................ 28 Hypothesis 1 ................................ ................................ ................................ ..... 28 Prompted Vs. Unprompted Performance on the OTDL from Baseline to Year 5 .... 28 Aim 2 ................................ ................................ ................................ ................ 28 Hypothesis 2 ................................ ................................ ................................ ..... 29 Longitudinal Relationship of Cognitive Status Group and Prompted vs. Unprompte d Performance on the OTDL ................................ .............................. 29 Aim 3 ................................ ................................ ................................ ................ 29 Hypothesis 3 ................................ ................................ ................................ ..... 29 3 RESEARCH DESIGN AND METHODS ................................ ................................ .. 31 Overview ................................ ................................ ................................ ................. 31 ACTIVE Study ................................ ................................ ................................ ......... 31 Design ................................ ................................ ................................ .............. 31 Recruitment ................................ ................................ ................................ ...... 32 Exclusion Criteria ................................ ................................ ............................. 32 Measures ................................ ................................ ................................ ................ 33

PAGE 6

6 Cognitive Domains ................................ ................................ ........................... 33 Memory domain ................................ ................................ ......................... 34 Attention domain ................................ ................................ ........................ 35 Language domain ................................ ................................ ...................... 35 Visuospatial domain ................................ ................................ ................... 36 Complex cognition domain ................................ ................................ ......... 37 Everyday Cognition ................................ ................................ .......................... 38 Additional Covariates ................................ ................................ ....................... 39 Procedures ................................ ................................ ................................ ............. 40 Group Classification ................................ ................................ ......................... 40 Participants ................................ ................................ ................................ ....... 41 Planned Analyses ................................ ................................ ................................ ... 42 Aim 1 ................................ ................................ ................................ ................ 42 Aim 2 ................................ ................................ ................................ ................ 42 Aim 3 ................................ ................................ ................................ ................ 44 4 RESULTS ................................ ................................ ................................ ............... 50 Overview ................................ ................................ ................................ ................. 50 Aim 1: Number of Verbal Prompts Given on the OTDL by Cognitive Status Group ................................ ................................ ................................ ................... 50 Post hoc Analysis of the Relationship between OTDL Item Difficulty and Number of Prompts ................................ ................................ ............................. 50 Aim 2: Prompted vs. Unprompted Performance on the OTDL at Baseline and over 5 Years ................................ ................................ ................................ ........ 51 Fixed Effects ................................ ................................ ................................ ..... 52 Random Effects ................................ ................................ ................................ 53 Aim 3: Longitudinal Relationship of Cognitive Group and Prompted vs. Unprompted Performance on the OTDL ................................ .............................. 54 Fixed Effects ................................ ................................ ................................ ..... 54 Rand om Effects ................................ ................................ ................................ 56 5 DISCUSSION ................................ ................................ ................................ ......... 64 Overview ................................ ................................ ................................ ................. 64 Review of Findings ................................ ................................ ................................ 64 Aim 1: Number of Verbal Prompts Given on the OTDL by Cognitive Status Group ................................ ................................ ................................ ............ 64 Aim 2: Prompted Vs. Unprompted Performance on the OTDL from Baseline to Year 5 ................................ ................................ ................................ ........ 65 Aim 3: Longitudinal Relationship of Cognitive Status Group and Prompted vs. Unprompted Performance on the OTDL ................................ .................. 66 Major Theoretical Issues ................................ ................................ ......................... 69 Limitations ................................ ................................ ................................ ............... 72 Future Directions ................................ ................................ ................................ .... 74 Final Conclusions ................................ ................................ ................................ ... 77

PAGE 7

7 APPENDIX: OBSERVED TASKS OF DAILY LIVING ................................ .................... 79 LIST OF REFERENCES ................................ ................................ ............................... 93 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 103

PAGE 8

8 LIST OF TABLES Table page 3 1 Cognitive measures by domain ................................ ................................ .......... 46 3 2 Unprompted and prompted OTDL score calculation sample .............................. 47 3 3 Baseline demographic mean (standard deviations) for full analytic sample and by cog nitive status group ................................ ................................ ............. 48 4 1 Fixed and random estimates for the full longitudinal model examining prompted vs. unprompted OTDL performance ................................ ................... 58 4 2 Fixed and random estimates for the full longitudinal model examining prompted vs. unprompted OTDL performance by cognitive group ..................... 59 4 3 Nested models ................................ ................................ ................................ .... 60

PAGE 9

9 LIST OF FIGURES Figure page 3 1 ACTIVE study design ................................ ................................ ......................... 49 4 1 Relationship between item difficulty and prompt ing on the OTDL. ..................... 61 4 2 Trajectory of prompted vs. unprompted performance on the OTDL ................... 62 4 3 Trajectory of prompted vs. unprompt ed performance on the OTDL by cognitive status group. ................................ ................................ ........................ 63 5 1 Possible reasons that participants needed prompting ................................ ........ 78

PAGE 10

10 Abstract of Thesis Presented to the Gradu ate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science VERBAL PROMPTING AS A METHOD FOR IMPROVING EVERYDAY COGNITION IN MCI AND UNIMPAIRED OLDER ADULTS By Kelsey Thomas May 2012 Chair: Michael Marsiske Major: Psychology This study investigated year longitudinal change in everyday cognition. Differential effects of prompting associated with impaired cognitive status (amnestic/non amnestic) were also examined. At baseline, 2,802 participants (mean age=73.6 years, mean education=13.5 years, 26% African American, Mini Mental Status Exam > 23) from the ACTIVE clinical trial were classified as unimpaired, amnestic mild cognitive impair ment (MCI) or non amnestic MCI based on psychometric algorithm. Participants were given the Observed Tasks of Daily Living (OTDL; a behavioral measure with tasks like change making, phone dialing) at baseline and at 1 2 3 and 5 year follow ups. When respond to an item, they received a standardized ve rbal prompt. At each occasion, U nprompted (sum of items correct without prompting) and P rompted (sum of items correct including both prompted and unprompted) sco res were derived for each participant. Multi level modeling, adjusting for demographics, health and training group, was used to determine the trajectories of OTDL performance. Results showed a linear and quadratic time pattern of initial gain followed by decline in later years across

PAGE 11

11 participants. The P rompted score was higher at all occasions and showed a flatter five year trajectory than the U nprompted. Cognitive impairment was associated with poorer OTDL performance (amnestic < non amnestic < unimpaire d). Prompting interacted with impairment status, such t rompted scores improved more than those of the unimpaired participants. Random time effects were observed, indicating individual differences in rates of change. Brief verbal prompts boosted everyday cognitive performance, particularly in MCI participants, and seemed to produce scores that were less affected by practice or decline. Discussion focuses on rove or maintain independent functioning.

PAGE 12

12 CHAPTER 1 LITERATURE REVIEW Overview This study examined whether a minimal level of support (i.e., verbal prompting) helped to improve performance on everyday tasks in older adults. We then investigated whether scores on a measure of everyday cognition after verbal prompting were more resistant to age related change in everyday cognitive function, relative to the un prompted score. Additionally, participants were classified using a psychometric algorithm as unim paired, amnestic mild cognitive impairment (MCI) or non amnestic MCI. Comparisons between groups were made on (1) the number of supportive prompts required on a performance based measure of everyday cognition, (2) whether MCI groups are initially more f unctionally compromised, (3) whether the impaired groups proceed to become disproportionately compromised, and (4) whether there are differential effects of prompting associated with impaired cognitive status. This investigation will contribute to the cu rrent everyday cognition literature by examining the longitudinal performance of unimpaired and cognitive ly impaired (amnestic and non amnestic) older adults on measures similar to activities performed in ts of prompting on these tasks may increase support for minimal prompting strategies to be clinically implemented in order to attenuate or delay the decline of independent everyday functioning that is frequently seen in older adults due to age related or p athological cognitive decline. Measures of Everyday Cognition The Advanced Cognitive Training of Independent and Vital Elders ( ACTIVE ) trial included measures of everyday cognition because these measures more closely

PAGE 13

13 resemble activities that older adults perform in their everyday life than more basic cognitive measures. This is important and innovative because it allowed the ACTIVE study to investigate real world functions (e.g., comprehension of medication labels) that are related to the cognitive domain s of the intervention (i.e., memory, reasoning, speed) in a laboratory setting, as well as examine whether the effects of cognitive training interventions generalized to more realistic functions. The Observed Tasks of Daily Living (OTDL; Diehl, Willis, & Schaie, 1995; Diehl, Marsiske, Horgas, & Saczynski 1998; Diehl et al., 2005 ) is a measure of everyday cognition used in ACTIVE and is the primary outcome of this current investigation. Everyday cognition measures attempt to capture the application of cogn itive skills and resources to important everyday (survival) tasks Everyday cognition has been conceptualized by Allaire & Marsiske (1999) as the performance of individuals on problems using natural (ecological) stimuli (e.g., read food package labels or official documents). Thus, most measures of everyday cognition are constructed to be similar in content and employ similar cognitive processes as tasks that are necessary for older adults to perform in their daily lives (e.g., identifying nutrition inform ation or comparing the value of different financial products). The concept of everyday cognition has been investigated as a reaction to the concern that laboratory based psychometric measures of cognition may not be aptly assessing the cognitive functionin g of older adults in real world situations (Demming & Pressey, 1957; Denney, 1989; Salthouse, 1990; Schaie, 1978; Sinnott, 1989; Sternberg & Wagner, 1986; Wagner, 1986; Willis & Schaie, 1986). In the context of a real world environment, cognitive performan ce of older adults can be more appropriately assessed since the individual can draw on their experience and

PAGE 14

14 practice to enhance their problem solving outcomes ( Berg & Klaczynski, 1996; Blanchard Fields & Chen, 1996; Cornelius & Caspi, 1987; Wagner, 1986; W illis, 1996; performance on prospective memory tasks in laboratory settings vs. non laboratory settings (e.g., in their home or a familiar environment). These findings rev ealed that older adults perform better when in the context of a more familiar environment (Rendell & Craik, 2000; Rendell & Thomson, 1999). There is, however, a clear consequence of relying so heavily on familiar experience. This may, for example, hinder adapt and function in novel situations, which would significantly restrict the context in which someone would feel comfortable and capable of function ing independently. While most laboratory based psychometric tests measure specific domai ns of basic cognition, real world problems do not typically involve only one specific area of cognition. In fact, it is suggested that a single real world problem may require several basic cognitive domains, including attention, working memory, processing speed, reasoning and executive functioning (Marsiske & Margrett, 2006). Because of its dependence on several domains of basic cognitive functioning in addition to amassed experience, everyday cognition is considered a higher order, multidimensional const ruct (Allaire & Marsiske, 1999; Marsiske & Willis, 1995; Willis & Marsiske, 19 91; Willis & Schaie, 1986, 1993). There has been great variability in the measurement of everyday cognition across studies, which has made it difficult to make generalizations about performance. One specific area of heterogeneity is in the structure of the measurement strategy. One approach to everyday problems, which is referred to as the ill structured approach,

PAGE 15

15 suggests that there is rarely a single, correct way to approach and solve a problem (Neisser, 1978; Wagner & Sternberg, 1985; Wagner, 1986). Thus, these ill structured assessments tend to be open ended vignettes that measure fluency, which is defined as the number of reasonable and effective solutions produced by the participant. Alternatively, the other approach has been labeled well structured, which is more similar to the traditional psychometric scoring system and assumes that the problems that occur in the real world can usually be solved with a single, effectiv e solution (Allaire & Marsiske, 1999, 2002; Hartley, 1989; Hershey & Farrell, 1999; Kirasic, Allen, Dobson, & Binder 1996; Morrell, Park, & Poon 1989; West, Crook, & Barron 1992; Willis, 1991, 1996 a ; Willis & Marsiske, 1991; Willis & Schaie, 1986, 1993). Those in support of the well defined approach have typically viewed everyday cognition as the result of underlying basic cognitive abilities; therefore, hypothesizing that everyday cognitive performance is a result of a fusion of multiple cognitive capac ities (Allaire & Marsiske, 1999, 2002; Berry & Irvine, 1986; Marsiske & Willis, 1998; Willis & Marsiske, 1991; Willis & Schaie, 1986, 1993). This current study used the Observed Tasks of Daily Living (OTDL) as a measure of everyday cognition It is consid ered a well structured measure because it assumes that there is a single best response for each item (e.g., when a participant is shown a medication bottle and asked how many days a refill will last, this measure assumes there is only one correct response) As the OTDL was the focus of this current investigation, well structured measures will be the focus for the remainder of this study.

PAGE 16

16 Relationship between Well Structured Measures of Everyday Cognition and Everyday Functioning In an effort of validate we ll structured measures of everyday cognition, Willis (1996b) suggested the use of individual ability differences in instrumental activities of daily living (IADL; e.g., medication use, financial management, food preparation; Lawton & Brody, 1969). The per formance on a measure of everyday cognition, the OTDL and IADL scores were positively correlated, r = 0.50 (Diehl et al. 1995). However, there are some limitations with this method of validation. First, IADL ratings are dependent on self or caregiver reports, which open up the possibility of reporting bias or inaccurate in community dwelling older adults who are functioning independently, there is an implie d ceiling effect in IADLs. Therefore, this limited range of IADLs in community dwelling elders cannot be expected to show extremely strong associations with measures of everyday cognition (Marsiske & Margrett, 2006). Thus, it is likely that measures of e veryday cognition are more sensitive to decline and pre clinical vulnerabilities in functioning than self or proxy reported IADL ratings. Negative Age Association with Everyday Cognition Like many aspects of cognition, everyday cognition usually shows ev idence of cross sectional and longitudinal decline in advanced old age. Trajectory findings vary with type of task, but well structured problems requiring one correct answer seem to be particularly vulnerable to decline. This decline may pre date self re ported decline in functional measures of cognition. A question is whether everyday cognition may signal pre clinical vulnerabilities in the ability to handle real life challenges even before participants and their proxies become aware of them. As describ ed by Marsiske &

PAGE 17

17 Margrett (2006), there have been several studies that have investigated the relationship between normal aging and performance on measures of everyday cognition. While some of these studies have used behavioral observation and others have used paper and pencil tests, most have investigated the domain of IADLs. There have been several cross sectional design studies of older adults (Allaire & Marsiske, 1999; Diehl et al., 1995, 2005 ; Lindenberger, Mayr, & Kliegl 1993; Marsiske & Willis, 199 5; Owsley Sloane, McGwin & Ball 2001 ) and middle aged to older adults (Whitfield, Baker Th omas, Heyward, Gatto & Williams 1999) that have shown a negative age association ranging from r = 0.23 to r = 0.45. Longitudinally, studies have shown modest declines, with a greater magnitude of decline in older participants (Willis, Jay, Diehl & Marsiske ., 1992; Willis & Marsiske, 1991; Willis, 1996 a ). Overall, a meta analysis conducted by Thornton and Dumke (2005) suggested that there is a negative age re lationship with older adults performing more poorly than middle aged or younger d = 0.48; Marsiske & Margrett, 2006). This suggests that, similar to the age related decline seen in many domains of basic cognition there is an age related decline observed on measures of everyday cognition, Importantly, compared to age related decline in basic cognition, the decline in every day cognition may be somewhat, but not fully, buffered by the relative preservation of knowledge based functioning since these tasks are often similar to activities performed in daily life. While measures of everyday cognition include tasks that seem fami liar, it is important to note that the tasks are still performed in a novel environment without external compensatory strategies (e.g., reminder notes, visual cues, reminder from

PAGE 18

18 re sensitive measure of true everyday cognitive performance. Mild Cognitive Impairment Just as everyday cognition measures may signal pre clinical vulnerabilities in adaptive cognition, there has been a great deal of diagnostic attention paid on identifyin g preclinical states of cognitive loss that may be pre morbid for imminent cognitive impairment and dementia. One of the most widely used classifications in recent research is Mild Cognitive Impairment (MCI). Some of the other terms that have been used t o describe the accelerated cognitive decline that occurs for a subset of older adults include: age related cognitive decline (American Psychiatric Association, 1994), age associated memory impairment (American Psychiatric Association, 1994) and cognitive impairment no dementia (CIND; Ebly, Hogan & Parhad, 1995; Graham, Rockwood & Beattie, 1997). MCI can be broken down into sub types based on the specific cognitive domain(s) that is impaired (e.g., single domain amnestic MCI, non amnestic MCI, multi domain MCI). The reason for classifying the vari ous domains of impairment is because there has been some evidence to suggest that different areas of impairment may be prodromal states of differ ent types of dementia (e.g., am n e stic MCI I n fact, a National disease (Albert et al., 2011). The definition of MCI is not wid ely agreed upon. One of the most commonly used is the Petersen (2004) criteria, which include: (1) subjective mean), (3) essentially preserved general cognitive functio n, (4) mostly intact functional

PAGE 19

19 abilities, (5) no dementia. The International Working Group on MCI (Winblad et al., their definition does not require a subjective cognitive complaint. The method of measuring these cognitive criteria is somewhat flexible, as there are no specific most widely used, there have been many studies that have sh own the poor predictive value of subjective complaints in older adults (Cook & Marsiske, 2006; Crowe et al., 2006; DeJager, Blackwell, Budge & Sahakian, 2005; Jorm et al., 1997). Jak and defined as having at least 2 impaired scores (defined as at least 1 SD below normative data) on neuropsychological measures within the same domain (i.e., memory, attention, language, visuospatial, or executive). There are no specific criteria for the meth od of deciding who is considered cognitively impaired. In fact, depending on the study, MCI has been classified numerous ways. This study will use an algorithmic approach that has been used in the ACTIVE sample ( Cook, et al., under review ; Wadley et al., 2007) as well as in other studies with large sample sizes ( Ganguli, Dodge, Shen, & DeKosky 2004; Ganguli, et al., 2010; Ritchie, Artero & Touchon, 2001). Similar to the other studies that have used this algorithmic approach in the ACTIVE sample, this i nvestigation will not include daily functioning performance as a criterion for MCI classification since, at baseline, all participants had intact ADLs. Relationship between Everyday Cognition and Cognitive Performance What is relatively unknown is whether, as one might expect, low functioning and/or declines in everyday cognition are associated with cognitive impairment. It is possible

PAGE 20

20 to hold competing hypotheses. then it ought to be sensitive to pr e clinical impairment classifications. On the other hand, if everyday cognition decline, then one ought to find that is shows stability even when more basic cognitive measures do not. If ever yday cognition is preserved, it is likely due to the high familiarity (i.e., regularly practiced) and relevance (i.e., high motivational value) of such tasks. The latter hypothesis is supported by the work of Cornelius & Caspi (1987) who suggested that th ere is a linear increase in everyday problem solving from young to late adulthood. However, as discussed above, in cross sectional studies, everyday cognition has been shown to be negatively associated with age, which is similar to the relationship betwee n basic cognitive measures and age. Thus, as age increases, it is assumed that someone has had more practice in these daily tasks; however, everyday cognition still appears to decline. Interestingly, longitudinal studies seem to demonstrate greater stabil ity in performance on measures of everyday cognition than is observed in the mean difference analysis of cross sectional designs. In a study by Willis et al. (1992), there were large individual differences in time and rate of change on the ETS basic skill s test; however, over 7 years, only 38% of the sample showed analysis confirms that while young and middle aged adults perform similarly in everyday problem solving/decision making effectiveness (EPSE), possibly due to the balance of accumulated knowledge and the beginning of basic cognitive decline seen in middle aged adults, EPSE performance is significantly lower in older adults. This is most likely due to memory and reasoning declines that eventually surpass the gains of accumulated knowledge.

PAGE 21

21 In order to support the proposition that decline in everyday functioning is sensitive to decline in basic cognition, it is important to investigate the relationship between measures of everyday cognition and tr aditional measures of cognition. While the majority of these studies are cross sectional and are very diverse in both the measures of everyday and basic cognition that were used, there is a small literature showing a shared variance of 50 to 80% between b asic cognitive abilities and well structured measures of everyday cognition (Marsiske & Margrett, 2006). Although reasoning/executive and memory abilities, followed by processing speed, are the most frequently studied basic cognition constructs, they also appear to be the greatest predictors of everyday cognition (A llaire & Marsiske, 1999; Burton, Strauss, Hultsch, & Hunter, 2006; Diehl et al., 1995; Owsley et al., 2001; Thornton, Deria, Gelb, Shapiro & Hill 2007 ; Wood et al., 2005). In a longitudinal st udy done by Willis et al. (1992), the relationship between fluid and crystallized intelligence, memory, and speed and a measure of everyday cognition called the Test of Basic Skills (TBS; Educational Testing Service, 1977) was investigated over 7 years. F indings revealed that baseline fluid reasoning was the best predictor of TBS performance (partial R 2 =0.52), with a total of 70% of the variance explained when all basic cognition measures and demographics were added to the model. Also, while mostly intact 4) MCI criteria, there is literature to suggest that performance based measures of everyday cognition, particularly those that focus on IADLs, were able to discriminate people with various levels of cognitive impairment and correla ted with neuropsychological measures that were sensitive to cognitive impairment (Marsiske &

PAGE 22

22 Margrett, 2006). Furthermore, research from the A CTIVE study has shown that being classified in a MCI subgroup (amnestic, non amne stic, or multi domain) predicted greater three year decline in both self reported activities of daily living (ADLs; e.g., feeding, hygiene, dressing) and self reported IADLs perceived difficulty and performance (Wadley et al., 2007). There have been sever al studies in older adults that have investigated the relationship between basic cognition, particularly executive functioning, and performance on IADLs. These studies have suggested that cognitive decline is associated with decline in IADL performance (C ahn Weiner, Malloy, Bole, Marran, & Salloway 2000; Cahn Weiner, Boyle & Malloy, 2002; Grigsby, Kaye, Baxter, Shetterly & Hamman 1998; Gross, Rebok, Unverzagt, Willis & Brandt 2011 ; Insel Morrow, Brewer & Figueredo 2006). Also, there is more recen t evidence indicating that individuals with MCI demonstrate impaired complex activities of daily living (IADLs) compared to non impaired older adults (Allaire & Willis, 2006; Goldberg et al., 2010; Perneczky et al., 2006; Tuokko, Morris, & Ebert, 2005). F or example, Allaire & Willis (2006) showed that older adults who were not impaired scored 0.66 standard deviations higher on a performance based measure of everyday cognition (Everyday Problems Test for Cognitively Challenged Elderly; EPCCE) than older adu lts classified as impaired. Then, over two years, impaired participants and participants who transitioned from non impaired to impaired status had significantly greater decline on the EPCCE compared to the participants who remained cognitively intact. Pre serving Everyday Functioning The benefits associated with preserving everyday functioning, or at least delaying there are cu rrently 5.2 million Americans

PAGE 23

23 jump to an estimated 7.7 million. Naturally, this incredibly large prevalence leads to significant financial concern amounting to about 183 bi llion dollars annually for costs such as healthcare and caregivers. Thus, there are significant financial benefits to remaining independent in the home longer before transitioning into assisted living facilities. In 2010, there were 17 billion hours of u npaid caregiver services provided to those with dementia. Caregiver burden can be seen in terms of emotional and physical stress. Therefore, it seems logical that by decreasing the rate of decline in everyday functioning, it would help alleviate some of t he burden experienced by caregivers. If everyday cognition is sensitive to cognitive impairment, and everyday cognition has functional/adaptive value in predicting independence, then it would be ideal to find compensatory strategies that can help preserv e and extend effective functioning. One compensatory strategy that has received attention, particularly in the dementia s. External prompting can be used to assist when areas of cognition such as prospective memory, self initiation processes, or working memory break down. Alarm clocks, calendars, automated reminders, pictures, or even a caregiver giving a reminder cue are various forms of external prompting, but are in no way the complete list. There have been a number of intervention studies that have investigated the use of external prompting devices in assisting with ADLs or IADLs. As discussed earlier, ADLs and IADLs are thought to involve several steps that require various domains of cognition; thus, the goal of these studies has been to break down the basic steps of a task so an older adult can successfully complete the task.

PAGE 24

24 The use of assisted technology for cog nition (ATC) has been the focus of much of the literature on external prompting. In a review by LoPresti, Mihailidis & Kirsch (2004), it is suggested that ATC can increase the efficiency of support practices (e.g., caregiving) by enabling the person to in dependently engage in the task as well as potentially increase the array of environmental contexts in which a task can be completed. One area that has been investigated is the use of an artificial intelligence computerized device to assist with hand washi ng in older adults with moderate to severe dementia. The studies have shown that the majority of participants did show some improvement with the device, and that the number of tasks completed without caregiver support increased (Mihailidis, Barbenel, & Fe rnie, 2004; Mihailidis, Fernie, & Cleghorn, 2000). Further work by this same group investigated the effectiveness of verbal cueing alone vs. audiovisual cueing. They found that, overall, these two methods were not significantly different from one another (Labelle & Mihailidis, 2006). In a different study by Bewernitz, Mann, Dasler, and Belchior (2009), completion of three types of self care tasks (e.g., drinking water, brushing teeth, and upper body dressing) were investigated in participants with dement ia. The study used machine delivered cognitive assistance. Over the three tasks, using the prompting device, participants were able to complete the tasks approximately 86% of the time; however, most participants required more assistance than the initial start directive to complete the task, although this varied individually. There are limitations in relying on these external prompts for support in daily functioning in cognitively impaired populations. One concern is that many of these prompting devices require the older adult to interact. For example, once the task is

PAGE 25

25 completed, the person may have to press a button to indicate to the prompting device that the task has been completed. While this may be feasible for someone who is fairly cognitive inta ct, this could be a concern and an area of distress for someone with a more severe dementia who is not comfortable with having to interact with the device, let alone remember what steps need to be completed when dealing with the device. In these cases whe n the individual has already experienced significant cognitive decline, awareness would be more appropriate (LoPresti et al. 2004). While external verbal prompts have been shown to reduce caregiver responsibility (Bew ernitz et al., 2009; Flannery & Rice, 1997), another concern is that people may become too dependent on the external prompts, particularly if the prompts give directives at each step of a task. This dependence is often seen in a patient caregiver relation r than allowing the person to perform th e task independently (Vogelpohl, Beck, Heacock, & Mercer 1996). Excess disability is described as the level of self care or cognitive function that is below what is expected from a physiological perspective (Kahn, 1971 1975). This concept is particularly prevalent in nursing homes where caregivers tend to reinforce dependent behaviors and provide ADL support even if the person could probably do the task independe ntly (Baltes, Burgess & Stewart, 1980; Barton, Baltes & Orzech 1980 ). Thus, it seems that it would be ideal to have a verbal prompting system that did not automatically give cues, but waited until a prompt was needed to assist the person in completing the task. Summary Cross sectional studies have shown that there are moderate relationships between everyday cognition and basic cognition, and eve ryday cognition and everyday

PAGE 26

26 functioning (i.e., self reported). In harmony with these findings, there is some research showing that older adults classified as cognitively impaired perform worse on performance based measures of instrumental activities of d aily living than non impaired older adults. However, there are lingering questions regarding the longitudinal change of IADL performance in older adults such as whether functioning remains relatively intact or whether it follows a trajectory more similar to that of more basic cognition, which tends to decline with age. Also of interest is whether any change in everyday With the increasing number of people over 65 yea rs old, it will be importa nt to develop ways to keep older adults functioning independently longer. Verbal prompting is a tool that has been investigated and found somewhat effective in the dementia population, but has had almost no attention in the normal to MCI older adult popul ation. Thus, this investigation also examined whether prompting serves as an intervention for improving performance on tasks of everyday cognition in a heterogeneous group of healthy and mildly impaired older adults. Additionally, the question of whether MCI (amnestic and non amnestic) participants respond particularly poorly or well to the prompts compared to unimpaired participants was investigated. The next chapter will discuss the specific aims of this study

PAGE 27

27 CHAP T ER 2 STATEMENT OF PROBLEM Overview While many current definitions of MCI include the criterion of intact functional abilities (Petersen, 2004; Winblad et al., 2004), more recent work has suggested that, in fact, those classified as MCI perform significantly lower on both measures of everyda y cognition and self reported IADL functioning (Allaire & Willis, 2006; Goldberg et al., 2010; Perneczky et al., 2006; Tuokko et al. 2005; Wadley et al., 2007). Also, to our knowledge no one has examined the benefits of verbal prompts for everyday cognit ion in adults in the normal to mild cognitive impairment range. While this tool has been investigated in the dementia literature, it seems logical to investigate low intensity prompting in older adult populations prior to the decline in basic activities o f daily living. It seems essential that, if an external prompting device is to be used, developing familiarity and understanding the use of a device while still mostly cognitively intact, may decrease the failure rate of the external prompt later on. Also of importance is that in the stages prior to significant decline is ADLs, a minimal prompting intervention would be ideal so the one does not become too dependent on these external cues. Thus, it is tool to refocus the person on what they should be doing or provide a minimal hint for how to initiate the activity. Also, the prompt should only be given when needed, rather than automatically. Given the research on ev eryday cognition in older adults with cognitive impairment and the lack of research investigating the use of minimal verbal prompts in this normal to MCI population, the current study will investigate whether there are differences in

PAGE 28

28 late life cognitive st atus groups (unimpaired, amnestic MCI and non amnestic MCI) on a measure of everyday cognition. Furthermore, the present study will explore the possible benefits of minimal verbal prompting on everyday cognition and whether these benefits differ over time and across cognitive status group. This investigation will use longitudinal data collected over 5 occasions (baseline, 1 2 3 5 year follow ups). There are three primary aims of this study. Number of Verbal Prompts G iven on the OTDL by Cognitive St atus Aim 1 Determine if the number of verbal prompts that were given on the OTDL at baseline differed by baseline MCI classification unimpaired, amnestic, or non am nestic. Hypothesis 1 Overall, participants in the impaired classifications (amnestic MCI and non amnestic MCI) will require more verbal prompts on the OTDL than the non impaired participants. Specifically, while the amnestic participants with additional impairments (multidomain amnestic) will likely be the lowest functioning overall, as a group, we hypothesize that the non amnestic group will require the most prompts. This hypothesis is based on the literature suggesting that reasoning and executive abilities are more related to everyday functioning than memory (e.g., Gross et al., 2011; Royall et al., 2007). Prompted V s. Un prompted P erformance on the OTDL from Baseline to Year 5 Aim 2 This analysis investigated if there is a difference in performance on the Observed Tasks of Daily Living (OTDL) when the U n prompt ed score (raw number of items co rrect,

PAGE 29

29 not including correct items after prompts) is compared to the Prompt ed score ( Unprompt ed score plus items correct after prompting). This analysis will use the ACTIVE BL, 1 2 3 and 5 year follow up data to examine the main effects of P rompting Status (i.e., getting help improves performance), O ccasion, and the interactions. Next, because there was evidence that there is a difference in Prompted and Unprompted scores at BL, we investigated whether five year scores on the OTDL were more resista nt to age related change as a function of prompting Hypothesis 2 We predicted that the Prompted scores would be higher than the Unprompted scores and that any 5 year OTDL change observed would be smaller under Prompted conditions than Unprompted condition s ( Prompting Status x Occasion ). Concretely, this will show the potential compensatory power of prompting, and show that age related change in everyday problem solving could be reduced with minimal ly directive standardized verbal support. Longitudinal Re lationship of Cognitive Status G roup and Prompted vs. Un prompted P erformance on the OTDL Aim 3 Determine if there are level differences in OTDL performance based on cognitive impairment status (i.e., unimpaired, amnestic, non amnestic), and if these level differences changed over time. Also, the differential effects of prompting associated with cognitive impairment status were examined. Hypothesis 3 It was hypothesized that the Cognitive Group main effect will show that the non amnestic group performs th e worst, followed by the amnestic group, with the

PAGE 30

30 unimpaired group performing the best at all occasions (unimpaired > amnestic > non amnestic) due to the stronger relationship between executive/reasoning abilities and functioning than memory and functioni ng (described above). However, because Gross and colleagues (2011) found that memory is more important for determining everyday functional change over time, we thought there would be a Cognitive Group x O ccasion interaction such that the level difference between unimpaired and impaired groups will get larger over 5 years, particularly with the amnestic group performing increasingly worse than the other groups. There will again be a main effect of P rompting Status but there will also be a significant P rom pting Status x Cognitive G roup interaction such that the impaired groups will not benefit as much from prompting as the unimpaired group.

PAGE 31

31 CHAPTER 3 RESEARCH DESIGN AND METHODS Overview The current study represents 5 year longitudinal data for the availab le analytic sample of participants enrolled in the Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) study and described by Cook et al. ( under review ). The ACTIVE study design, recruitment, and eligibility sections are first described followed by the measures, procedures, study sample and statistical analyses. ACTIVE Study Design All participants (N = 2802) randomized in the ACTIVE clinical trial were considered for inclusion in this investigation. Recruitment procedures, sample c haracteristics, and study design have been described elsewhere (Ball et al., 2002; Jobe et al., 2001; Willis e t al., 2006) Briefly, ACTIVE is a randomized, controlled, single blind trial evaluating the effectiveness of three cognitive training interventions in improving cognitive abilities and daily functioning in independently living adults over age 65. Pa rticipants were randomly assigned to one of four training conditions either memory training, reasoning training, speed training, or no contact control. Approximately 11 months following the completion of the cognitive training, participants in from each t raining intervention (memory, reasoning, and speed) who completed at least 80% of the initial training sessions were randomly selected to receive four additional booster sessions. Participants underwent psychometric testing to measure memory, reasoning, a nd perceptual speed at baseline, post 10 session cognitive intervention (or no contact control), and one, two, three and five years post intervention. Figure 3 1 illustrates the

PAGE 32

32 ACTIVE study design through the 10 year follow up (in progress); however, only data through the 5 year follow up was available for this investigation. Recruitment Participants were recruited from March 1998 through October 1999 at six field sites and included over sampling of African Americans. The goal was to obtain a representati ve sample of older adults who have not yet experienced significant functional decline. Jobe et al. (2001) described the following recruitment sites. The University of Alabama recruited participants through the Department of Public Safety and through UAB eye clinics. The Hebrew Rehabilitation Center for Aged, located in Boston, recruited from congregate and senior housing sites, senior centers, and a volunteer research registry. Indiana University recruited through a group of facilities providing activit ies and social services to seniors, as well as local churches and senior organizations, senior centers, churches, senior housing, and through programs offering or coordinatin g wellness or service programs for seniors. Pennsylvania State University recruited through a state funded pharmaceutical assistance program for low income elders. Wayne State University recruited from a large range of community organizations, churches, ho spital based senior assessment centers, senior housing sites, and driver registration lists. All procedures were approved by the Institutional Review Boards at each collaborating site and informed consent was obtained prior to participation. Exclusion Cri teria Exclusion criteria were used to identify physically and cognitively intact older adults that would be able to participant in the study longitudinally. Participants were not

PAGE 33

33 included in the ACTIVE sample if they a) were <65 years old at the initial screening; b) experienced a substantial cognitive decline defined as a Mini Mental State Exam (MMSE; Folstein, Folstein & McHugh, 1975) score of < 23 or a prior diagnosis of dementia; c) had self reported ADL limitations (Morris et al., 1997; e.g., bathing dressing, and personal hygiene); d) had a medical condition that was likely to result in imminent functional decline or mortality before the second annual follow up assessment (e.g., stroke, certain cancers); e) experienced severe sensory losses (e.g., s elf report of extreme difficulty reading newspaper print or performance based vision test scores exceeding 20/50; Mangione et al., 1992); f) had communication difficulties extensive enough to prevent participation in study protocol (based on interviewer ra ting); g) had recent cognitive training; or h) were unavailable during the testing and training phases of the study. Measures Cognitive Domains Cognitive domains were created for memory, attention, language, visuospatial, and complex cognition domains. T hese domains were created to be closely comparable to the domains used by Jak et al. (2009) to define MCI subgroups. However, there was relatively little cognitive domains with those used by Jak and colleagues This was due to the defined set of measures that were available in the ACTIVE study, and this secondary analysis study re purposes those measures for reasons that were not originally intended. For example, Jak et al. had a well n consisting of measures of mental switching, inhibition, problem solving, and cognitive flexibility. However, ACTIVE did not have this range of measures, so this investigation used

PAGE 34

34 domain consisting of measures of reasoning and ever yday problem solving. The cognitive domains of the current study are consistent wit h those used by Cook et al. (under review) and are outlined further below. Table 3 1 shows the breakdown of what individual measures were compiled into each respective dom ain composite score. Test retest reliability statistics a re also displayed in Table 3 1. Memory domain The memory measures focused primarily on episodic verbal memory that were included in the memory domain score include the Hopkins Verbal Learning Test, Related Word Lists (HVLT; Brandt, 1991); Rey Auditory Verbal Learning Test, Unrelated Word Lists and Delayed Recognition (AVLT; Rey, 1941); and the Rivermead Behavioral Memory Test, Paragraph Recall task (RBMT PR; Wilson et al., 1985). The HVLT is a list learning measure that required participants to recall a list of 12 words that could be semantically grouped into three categories. There were three trials across trial s, with potential scores ranging from 0 to 36. For the AVLT, the participants heard a list of 15 words that were read aloud with an eight second pause between each word. Participants were then given two minutes to write down as many words as they could re member from the list. This process of word then totaled across the five trials, with potential scores ranging from 0 to 75. The participants were then read a list of words, some of these were new words and others were those from the original list, and the participants were asked to determine which words were on the original list of words that was read to them.

PAGE 35

35 The RBMT PR is a story memory measure where participants listened to a pre recorded story and then ask ed to write down as much of the story as they could remember as soon as the audiotape ended. The passage was divided into 21 key idea ot recalled, 0.5=approximately accurate, 1= completely accurate). Alternate stories were administered at each occasion. Attention d omain The attention measures that were included in the attention domain score were the Digit Symbol Copy (Wechsler, 1981) a Field of View test (UFOV, task 1; Ball & Owsley, 1993) The Digit Symbol Copy task required participants to simply copy symbols in a box immediately beneath a target symbol. This measure was prorated to reflect the number of correct items in 90 seconds simply required the participant to identify a target of either a truck or a car that is presented at a central fixation point on the screen. The participant responded by touchin g the screen to indicate their response. Scores were recorded based on the minimum stimulus duration in which the participant responded correctly 75% of the time. Language domain The Language domain was comprised of the Vocabulary subtest of the Kit of Fa ctor Referenced tests (Ekstrom et al. 1976) and the sum score for 5 language items from the Mini Mental Status Examination (Folstein et al. 1975). The Vocabulary subtest required participants to recognize words and their definitions Scores ranged for 0 to 18, with higher scores indicating more items correct.

PAGE 36

36 The MMSE items that were included in the language domain were object naming following items (i.e., take pa per in right hand, fold in half, place on floor), and writing a sentence. Scores potentially ranged from 0 to 7. Visuospatial d omain The Visuospatial domain was comprised of the mean of Tasks 2 and 3 on the Useful Field of View test (Ball & Owsley, 1993) and the Digit Symbol Substitution test from the WAIS (Wechsler, 1981). This conceptually differs from the Jak et al. (2009) visuospatial domain in that construction tasks are not included and each of these tasks involve a processing speed component. In UF two dimensional picture of either a car or a truck at the center of the computer screen. At the same time, a picture of a car appeared at one of eight locations around the peri phery of the computer screen. The task was to first choose if it was a car or truck that was presented in the center of the screen, and then selected where the car appeared around the periphery of the screen. The presentation of the center car or truck a additional distraction. In this task there were small triangles that served as visual clutter that were in the space on the screen between the center car/truck and the peripheral car. For both tasks, the score is the minimum stimulus duration in which the participant responded correctly 75% of the time.

PAGE 37

37 The Digit Symbol Substitution task required participants to use a legend on the top of the page to fill in a symbol in the lower box that corresponds with the number in the upper box. The goal was to fill in as many correct symbols possible in 90 seconds Complex cognition d omain Complex cognition was assessed using the Word Series Measure (Gonda & Schaie 1985 ) and the Everyday Problems Test (EPT; Willis & Marsiske, 1993). While this domain measures reasoning and everyday problem solving ability, as described above, this domain deviated f rom Jak (2009) conceptualization of the Executive domain in that measures of switching, inhibition, cognitive flexibility, and perseveration were not included. The Word Series measure has a series of days of the week or months of the year, and the participant was asked to determine which day or month came next in the sequence from five choices. Participants had six minutes to complete 30 problems, and the total raw score was based on the correct items (possible range 0 30). The EPT is a pencil an d paper measure of everyday problem solving where the participant was presented 14 everyday stimuli (e.g., medication labels, transportation schedules, telephone rate charts, Medicare benefits information) and asked to answer questions about the stimuli. There are 7 domains of daily living that are represented on the EPT: food preparation, medication use and health behaviors, telephone use, shopping and consumerism, financial management, housekeeping and laundry, and transportation. EPT scores range from 0 28 and are based on the number of corrected responses.

PAGE 38

38 Everyday Cognition Everyday cognition and response to verbal prompts was measured in this study using the Observed Tasks of Daily Living (OTDL; Diehl et al., 1995 1998). In the OTDL, participants are presented with nine tasks and a total of 13 questions (some questions have multiple parts) in the domains of medication use, telephone use, and financial management. This task is expected to be most closely related to reasoning and memory. In this m given. The prompts were not designed to give the participant the answer, but rather, to either just serve as a reminder, motivation, or to help initiate the process of finding the answer. For example, on one of the items, the participant was given a package insert according to this package insert, how many tablets of aspirin Is there any Performance on th e OTDL measure was scored by a certified scorer and the range of scores possible is 0 28. Diehl et al. (2005) reported that the Kuder total measure was 0.71 and the internal consistency of the total measure was 0.81. Ba ll et al. (2002) report an internal consistency of 0.75. See Appendix A for the OTDL measure OTDL unprompted vs. prompted score calculation. There were two different OTDL scores calculated and used in this investigation an U n prompt ed score and a Prompt e d score. The Unprompted score was calculated by summing the number of items correct without a prompt The Prompted score was the Unprompted score plus

PAGE 39

39 any items where the participant responded correctly after receiving a prompt. Table 3 2 displays a sam Unprompted and Prompted scores were calculated. Participants who did not receive a prompt were dropped from the prompted score analyses; however, at BL, only 6.2% (n=173) of the participants did not receive a prompt. In the BL sa mple (N=2,799), the Unprompted scores ranged from 1 28 with the mean (SD) score was 17.00 (4.62) and the Prompted score ranged from 3 28 with a mean (SD) score of 20.51, which is a difference in score of 3.51 points. The number of prompts given at BL rang ed from 0 22, with the mean (SD) number of 4.81 (3.51). Additional Covariates Baseline demographic and health covariates were included in the model to adjust for selective attrition (e.g., age, education, race, vision, general health) and significant gro up differences not specifically related to cognition (e.g., depression, gender). General health that is non specific to cognition was measured using the General Health subscale of the MOS Short Form Health Survey (SF 36; Ware & Sherbourne, 1992). Vision in ACTIVE was measured using a scale that ranged from 0 90, with higher scores indicating better vision. Furthermore, a 7 level categorical intervention group variable was created to control for any potential and differential cognitive training and boost er session gains (coded as 1=Memory training, no booster; 2=Reasoning training, no booster; 3=Speed training, no booster; 4=Memory training, with booster; 5=Reasoning training, with booster; 6=Speed training, with booster; 7=control group, no training). T he Center for Epidemiological Studies Depression 12 scale (CES D) was used to assess depressive symptom severity (Radloff, 1977). In addition to these measures used as covariates in the analyses, other measures were used simply to compare cognitive group differ ences at baseline. These include the Mini Mental State

PAGE 40

40 Exam (MMSE; both with language included and without in order to avoid circular data since language items were used to classify cognitive status group), which was used to measure global functioni ng (Folstein et al., 1975), the Physical Functioning subscale of the SF 36 ( Ware & Sherbourne, 1992), and a composite of everyday functioning that is composed of the Instrumental Activities of Daily Living (IADL) performance and difficulty scores and Activ ities of Daily Living (ADL) score (ADL/IADL, Morris et al., 1997) was used to examine functional status. This was created by summing the points associated higher score s correspond to worse functioning. Procedures Group Classification This study reports data from baseline (BL), first annual (A1), second annual (A2), third annual (A3), and fifth annual (A5) follow up evaluations of the ACTIVE study. However, to classify participants into cognitive status groups, only BL data was used and these methods are described fully in Cook et al. ( under review). Briefly, the psychometrically defined MCI groups were determine by first computing the baseline score distribution for each measure separately for each combination of age (65 69.9, 70 74.9, 75 79.9, and 80+ years), education (0 11.9 years, 12 15.9 years, and 16+ years), and race (African American, White). Each participant received a percentile rank for each measure based on their age/education/race group. computed in each of the five cognitive domains (Memory, Attention, Language, Visuospatial Processing, Complex Cognition). Each measure w ithin the composite score is weighted equally. Individuals whose average percentile was 15.87 (~1.0 SD

PAGE 41

41 below the mean ) or lower were classified as impaired in that domain. A percentile cut off was used to account for the non normal distribution of some o f the measures. This cut off of 15.87 (~1 SD) below the mean was implemented because this closely appro ach differed, however, in that a composite was used rather than requiring t wo or more performances within a domain to fall at least 1 SD below the normative mean. Participants who performed above impairment threshold in all domains were considered unimpaired. Participants with impairment in the memory domain were considered amne stic MCI. These participants could also be impaired in other domains. Lastly, participants with at least one impairment in a non memory domain and who were not impaired in memory were considered non amnestic MCI. Participants The full ACTIVE analytic sample included 2,802 healthy older adults ranging from 65 to 94 years old and were all considered for use in this investigation. Participants were randomly assigned to the memory training, reasoning training, speed training, or no contact control groups following baseline assessment, which resulted in 703 participants randomized to memory training, 699 to reasoning training, 702 to speed training, and 698 to the no contact control group. In this study, three participants were excluded at baseline for mi ssing OTDL data. Demographic characteristics for the full 2,799 participants included at BL as well as by cognitive group are depicted in Table 3 3 Because the ACTIVE sample is a community based, primarily healthy sample of older adults, there were more unimpaired than cognitive impaired participants. At BL, there were 1,872 unimpaired participants, 225 amnestic participants, and 703 non amnestic participants. Impaired participants generally showed poorer functioning in

PAGE 42

42 regard to vision, MMSE scores, d epression, physical functioning, and activities of daily living than the unimpaired participants and were more likely to drop out of the study early. Between the impaired groups the amnestic group was the oldest, had more males, performed more poorly on the MMSE, and reported more difficulty performing activities of daily living. Of t he amnestic participants, 58.7 % were also impaired in at least one non memory domain (i.e., multidomain amnestic) and 32.5 % of the non amnestic participants were impaired in more than one non memory domain (i.e., multidomain non amnestic). Planned Analyses Aim 1 To determine if the number of OTDL verbal prompts that were given differed by cognitive impairment status. E was summed at BL Then a one way Analysis of Variance was run to determine the significance of the mean differences in number of prompts between the unimpaired, amnestic, and non amnestic participants at BL Bonferroni post hocs were used to adjust for multiple comparis ons. The number of OTDL prompts given was calculated for each participant at each occasion. The maximum possible number of prompts was 28, but the sample ranged from 0 22 prompts. It is important to note that at BL only 6.3% of the sample did not require any prompts. Aim 2 To investigate if prompting significantly improved performance on the OTDL over all five occ asions, we conducted the analysi s using a multilevel model (MLM) approach, which is also known as a mixed effects model or hierarchical linear model ( Bryk &

PAGE 43

43 Raudenbush, 1992 ). MLM is an extension of the general linear model and does not require the observations to be independent. It is very flexible and well suited for a study design with multiple occasions since it allows all available data t o be utilized without having case wise exclusion for a missing data point ( Singer & Willett, 2003 ). Thus, bias of the results due to selective attrition can be minimized using a full information maximum likelihood approach. Furthermore, we attempted to further minimize attrition bias by including baseline covariates that have been shown to predict attrition (e.g., age, education) in this sample. Visual inspection of the longitudinal data suggested both linear and quadratic effects of time, such that th e Unprompt ed score appears to be more quadratic and Prompted appears more linear. Thus, two occasion variables were used L inear (coded as 0, 1, 2, 3, 5) reflects the time spacing of the testing occasion and Q uadratic, which was computed by squaring the li near occasion variable and taking the residual. Because attrition was not completely random in this sample (older, more impaired persons were more likely to drop out), baseline covariates were also included since the baseline sample is the occasion with the least amount of bias as a result of selective attrition not yet occurring. These covariates were similar to those used by Wolinsky et al. (2009) minus the cognitive predictors and include: BL A ge, Y ears of E ducation, R ace (African American), BL Vi sio n score, and BL SF 36 G eneral H ealth index score. Additionally, due to significant group differences in levels of depression and the evidence that there may be gender differences in the way men and women respond to external support (Margrett & Marsiske, 2002), these covariates were also included as covariates in the model. Multilevel modeling is able to use these less biased covariates

PAGE 44

44 to give the best estimate for missing data at future occasions. MLM allows both the fixed and random effects of the mod el to be examined. Fixed effects are the group or model were: Intercept, Prompting S tatus ( Prompted vs. Unprompted ), Replicate, Site, Intervention Group, Occasion, O ccasion 2 Prompting S tatus x Occasion, Prompting S tatus x Occasion 2 in addition to the BL covariates described above. On the other hand, random effects examine whether a particular individual significantly differs from the obtained fixed effect. The r andom effects that were included in this model were: Intercept, Occasion, and Occasion 2 ( Prompting S tatus x Occasion and Prompting S tatus x Occasion 2 was also included, but could not be reliably estimated). Aim 3 To determine if there were level difference s in OTDL performance based on cognitive impairment status (i.e., unimpaired, amnestic, non amnestic), if these level differences changed over time, and if there were differential effects of prompting associated with cognitive impairment status, a MLM very similar to Aim 2 was used. In fact, the model was identical, except the cognitive group of the participants was now added to the model, as well as the additional 2 and 3 way interactions. The BL covariates described above were also included in this mod el. The fixed effects that were included in this model were: Intercept, Prompting S tatus ( P rompted or U nprompted), Replicate, Site, Intervention Group, Occasion, Occasion 2 Prompting S tatus x Occasion, Prompting S tatus x Occasion 2 Prompting S tatus x Gro up classification, Cognitive Group x Occasion, Cognitive Group x Occasion 2 Prompting S tatus x Cognitive Group x Occasion, and Prompting S tatus x Cognitive Group x Occasion 2 in addition to the B L covariates described above. The

PAGE 45

45 random effects that were i ncluded in this model were: Intercept, Occasion, Occasion 2 Prompting Status Cognitive Group x Occasion, Cognitive Group x Occasion 2 Prompting Status x Cognitive Group, and Prompting S tatus x Cognitive Group x Occasion ( Prompting Status x Occasion, Promp ting Status x Occasion 2 and Prompting Status x Cognitive Group x Occasion 2 were also included, but could not be reliably estimated).

PAGE 46

46 Table 3 1: Cognitive measures by domain Domain Test Reliability* Published Source Memory HVLT, Recall AVLT, Recall AVLT, Recognition RBMC, Paragraph Recall 0.73 a 0.78 a 0.36 b 0.60 a Brandt, 1991 Rey, 1941 Rey, 1941 Wilson et al., 1985 Attention Digit Symbol Copy UFOV, task 1 0.62 c 0.69 d Wechsler, 1981 Ball & Owsley 1993 Language Vocabulary MMSE language items § 0.87 e Ekstrom French, Harman, & Derman, 1976 Folstein, Folstein & McHugh, 1975 Visuospatial UFOV, tasks 2/3 composite Digit Symbol Substitution 0.78 e 0.82 f Ball & Owsley, 1993 Wechsler, 1981 Complex Word Series Everyday Problems Test 0.84 a 0.87 a Gonda & Schaie,1985 Willis & Marsiske, 1993 Note. HV LT=Hopkins Verbal Learning Test; AVLT=Auditory Verbal Learning Test; RBMT= Rivermead Behavioural Memory Test ; UFOV=Useful Field of View; MMSE=Mini Mental Status Exam. Some memory tests were modified for group ad ministration (AVLT, RBMT), and all memory tests employed audiotape administration, written responses by participants, and no delayed trials. *Reliabilities are all test retest correlations; § I tems extracted from MMSE; published reliabilities not available ; a Ball et al., 2002; b Schmitt, 2004, one year interval ; c Calc ulated using ACTIVE control group, one year interval ; d Edwards et al., 2005; e Calculated using ACTIVE control group 12 week interval ; f Wechsler, 1981.

PAGE 47

47 Table 3 2 Unprompted and p rompted OTD L score calculation sample Item Number Unprompted (raw score without prompting) Prompted (Unprompted score + items correct with prompting) 1 1 1 2 1 1 3 0 1 4 1 1 5 0 1 6 0 0 Total 3 5 Note A rrow indicates item where participan t required a prompt and answered the item correctly following the prompt.

PAGE 48

48 Table 3 3. Baseline demographic mean (standard deviations) for full analytic sample and by cognitive status group Full Sample (N=2,799) Amnestic (N=225) Non Amnestic (N=703) Unim paired (N=1872) F statistic / 2 § df p value p 2 BL Age 74.11(5.90) 75.09 (6.91) a 74.25 (6.14) ab 73.95 (5.67) b 3.93 2, 2796 .020 .003 Education 13.53 (2.70) 13.43 ( 2.79 ) a 13.46 ( 2.79 ) a 13.56 ( 2.66 ) a 0.54 2, 2796 .586 <.001 Female, n (%) 2124 (75.88%) 130 (57.78%) 523 (74.50%) 1471 (78.58%) 43.89 § 2 <.001 African American, n (%) 727 (25.97%) 55 (24.44%) 198 (28.21%) 474 (25.32%) 2.48 § 2 .289 BL Vision Score 73.12 (11.49) 70.38 ( 12.11 ) a 71.78 ( 11.87 ) a 73.96 ( 11.16 ) b 16.36 2, 2796 <.001 .012 BL MMSE 27.31 (2.00) 26.10 (1.87) a 2 6.53 ( 2.09 ) b 27.75 ( 1.83 ) c 155.45 2, 2796 <.001 .100 BL MMSE No Language 19.70 (1.87) 18.62 (1.83) a 19.20 (2.00) b 20.02 (1.73) c 95.49 2, 2796 <.001 .064 BL CES D 5.22 (2.13) 6.30 ( 5.11 ) a 6.16 ( 5.18 ) a 4.76 ( 5.06 ) b 24.59 2, 2796 <.001 .017 BL SF 36 Gener al Health 69.00 (19.24) 67.74 ( 19.27 ) ab 65.54 ( 19.51 ) a 70.45 ( 18.97 ) b 17.07 2, 2746 <.001 .012 BL SF 36 Physical Functioning 68.83 (24.06) 64.25 ( 25.71 ) a 65.91 ( 24.27 ) a 70.46 ( 23.63) b 13.40 2, 2754 <.001 .010 BL IADL/ADL 6.01 (6.41) 8.53 ( 7.69 ) a 6.78 ( 7. 13 ) b 5.42 ( 5.82 ) c 31.20 2, 2796 <.001 .022 Dropout by A5, n (%) 922 (32.94%) 107 (47.56%) 283 (40.31%) 532 (28.42%) 55.06 § 2 <.001 Note Matching superscript letters indicate values are not significantly different, p>.05, using Bonferroni corrected o ne way ANOVAs. Chi squared tests were used to measure dichotomous variables and are denoted with superscript § Significant overall chi square statistics were followed up with logistic contrasts, using the Unimpaired group as the reference value; signific ant differences after adjusting for the 2 comparisons to the reference group are denoted with superscript*. MMSE=Mini Mental Status Exam; CES D=Center for Epidemiological Studies Depression 12; IADL/ADL= Instrumental Activities of Daily Living/Activities o f Daily Living

PAGE 49

49 Figure 3 1. ACTIVE study d esign

PAGE 50

50 CHAPTER 4 RESULTS Overview This study investigated whether minimally directive standardized verbal prompts were successful in increasing performance on everyday tasks in community dwelling older adults. Furthermore, the study addressed whether people who had extremely low cognitive functioning experienced benefits of prompting that were similar to those with higher cognitive functioning. The ACTIVE clinical trial longitudinal data from BL, A1, A2, A3, a nd A5 was used to answer these questions. Aim 1: Number of Verbal Prompts Given on the OTDL by Cognitive Status Group The first goal of this study was to understand if the amount standardized verbal prompts that were provided differed by cognitive status g roup. A one way Analysis of Vari ance (ANOVA) that compared the N umber o f P rompts received (DV) across Cognitive G roup (3: amnestic, non amnestic, unimpaired) at the first occasion (1: BL) was conducted This analysis showed a significant main effect of B L Cognitive G roup classification on the number of prompts participants received at BL, F (2, 2796)=81.49, p <.001, p 2 =.055 Bonferroni adjusted post hoc analyses revealed that the un impaired group received fewer verbal prompts (mean=4.23, SD=3.05) than eit her the amnestic (mean=6.02, SD=3.94) and non amnestic (mean=5.97, SD=4.07) groups. The number of prompts received by the amnestic and non amnestic groups did not significantly differ at BL. Post hoc Analysis of the Relationship between OTDL Item Difficu lty and Number of Prompts The group differences tell us about person factors that predict higher rates of prompting. Are there also characteristics of the testing stimuli themselves that predict a

PAGE 51

51 higher rate of prompts? To look at this, a post hoc analy sis was run to investigate the association between individual item difficulty on the OTDL and Prompting. A simple linear regression was done to analyze the relationship between the percentage of participants who got an item correct without a prompt and th e percentage of people who needed a prompt. A regression wa s used in order to correct for the potential r squared bias due to the small sample Only 19 of the 28 items were included because for several of the items, the number of prompts was summed acros s a group of items. As such, for that group of items, the prompt s could not be specifically linked to a single item and, therefore, the percent of participants who responded correctly and the percent of participants who were prompted could not be accurate ly separated in the data for those items. There was a strong inverse relationship ( = .71, p =.001) such that the percent of participants who needed a prompt on an item explained a significant portion of the variance in the percent of participants who responded correctly to an item, R 2 =0 51 adjusted R 2 =0. 48 F (1, 18)=17.59, p =.001. Ther efore, items where a larger percentage of participants responded correctly without a prompt, required fewer prompts. Figure 4 1 illustrates this relationship. Thus, item difficulty accounts for a large portion of the variance in explaining the reasons be hind why participants required prompts. Aim 2: Prompted vs. Un prompted Performance on the OTDL at Baseline and over 5 Years Multilevel modeling (MLM) was done to examine the difference in score on the OTDL when items that were prompted and subsequently ans wered correctly were con sidered correct as part of the Prompted score. The Prompted score is compared to the U nprompted score which is the raw score where prompted items were considered incorrect. Full details of this analysis are described in the Metho ds section Briefly, th is

PAGE 52

52 analysis was a 2 (Prompting S tatus: Prompted vs. Unprompted, within subjects) x 5 (Occasion: BL, A1, A2, A3, A5, within subjects) MLM with Replicate, Site Intervention Group BL Age Education Race, Gender, BL Vision, BL Depress ion, and BL Gener al Health as covariates. OTDL P erformance was the dependent variable (DV). T he intraclass correlation (ICC) is an index for the amount of between and within person variance to be explained (Bryk & Raudenbush, 1992) For the Unconditio nal Means Model (UM M) in which only the fixed and random intercept were included in the in the model where OTDL P erformance was the DV, the intraclass correlation (ICC) was 0.49. This indicates that 49% of the variability in OTDL P erformance was between pe rson and 51% was within person. Table 4 1 displays the level one (within subjects) and level two (between subject) parameters for the full model. Fixed Effects A fixed effect provides the overall effect across all participants, and is similar to effect s one might find in OLS regression or ANOVA. The majority of the BL covariates that were included in order to help control for selective attrition and between subject differences (as described in the Methods section) had significant main effects ( F statist ics displayed in Table 4 1 p <.05), with the exception of the I ntervention G roup variable, suggesting that there is not a unique effect of I ntervention G roup on OTDL P erformance. There was a significant effect of Prompting status F (1,2660.80 )= 531 8.08, p <.001, such that the Prompted scores, across occasions and participants, were approximately 3.0 7 points higher than the Unprompted OTDL scores using the covariate adjusted estimates ( Unprompted mean=17.37, s.e.=0.08 ; Prompted mean=20.4 4 s.e.=0.09 ). The L inear and Q uadratic effects of time were both significant F (1, 1900.77 )= 41.39 p <.001 and F (1, 2214.82 )= 108.86 p <.001, respectively. T he

PAGE 53

53 L inear effect had a fairly flat, but positive slope while the Q uadratic O ccasion estimation was negative, which sugge sts an inverse U shape, indicating initial gain followed by decline. Figure 4 2 depicts the trajectory of the Prompted versus Unprompted OTDL scores. As shown in the figure, the Prompted scores have a flatter trajectory while the Unprompted scores appear to show a more quadratic effect of time such that there is initial gain followed by a decline in OTDL performance at the later occasions. The Prompting S tatus x L inear O ccasion interaction was non significant, but the Prompting S tatus x Q uadratic O ccasio n interaction was significant [ F (1, 11202.47 )= 154.11 p <.001 ] suggesting that the Prompted performance across occasions, while the Unprompted scores had a greater initial practice effect followed by greater decline from Year 3 to Year 5. Random Effects A random effect assesses whether there are individual differences in the strength of a relationship between a within subjects IV and an outcome. Both the L inear and Q uadratic random effects of time were signif icant ( Wald Z= 16.59 p <.001 and Wald Z= 17.36 p <.001, respectively), which suggest that there are individual differences in both linear and quadratic rates of change. Prompting S tatus also showed a significant random effect ( Wald Z= 9.05 p <.001 ) suggestin g that there are inter individual differences in the effect of prompting (i.e., prompting was differentially effective by person). Random Prompting S tatus x Occasion interactions (both Linear and Quadratic) were included in the model; however, they could n ot be reliably estimated. Additionally, there was significant between and within subject variance that remained to be accounted for in future models. This model explained 48.5% of the within subject variance and 36.9% of the between subject variance. R elativ e to the Unconditional

PAGE 54

54 Means M odel this model is a significant improvement in fit [ 2 (33)=9095.06]. Incremental model fit is displayed in Table 4 3. Aim 3: Longitudinal Relationship of Cognitive Group and Prompted vs. Un prompted Performance o n the OTDL This multilevel model was done to examine the Prompted versus Unprompted OTDL Performance by C ognitive G roup. Full details of this analysis are described in the Methods section. Briefly, th is analysis was a 2 (Prompting S tatus: Prompted vs. Unpr ompted, within subjects) x 5 (Occasion: BL, A1, A2, A3, A5, within subjects) x 3 (Cognitive Group: amnestic, non amnestic, unimpaired, between subjects) MLM with Replicate, Site Intervention Group BL Age Education Race, Gender, BL Vision, BL Depression and BL Gener al Health as covariates. OTDL P erformance was the dependent variable (DV). This model is nested within the Aim 2 model in that the models are the same with the exception that the Aim 3 model also adds the between subjects Cognitive Group v ariable and its interactions. Table 4 2 displays the fixed and random effects of the level one (within subjects) and level two (between subject) variables for the full model. Fixed Effects The majority of the BL covariates that were included in order to help control for selective attrition, between subject differences, and between group differences (e.g., depression) had significa nt main effects except for the R eplicate variable (F statistics are displayed in T able 4 2) There was again a significant m ain effect of Prompting Status F (1, 2619.41 )=540 9.51 p <.001, with Prompted scores being 3.08 points higher than Unprompted scores, across occasions and participants using the covariate adjusted estimates ( Unprompted mean= 16.49 s.e.=0.09 ; Prompted mean= 1 9.57

PAGE 55

55 s.e.=0.09 ). Similar to Aim 2, both L inear and Q uadratic main effects of time were significant F (1, 1584.79 )= 35.03 p <.001 and F (1, 2203.11 )= 102.87 p <.001, respectively. Again, Linear Occasion had a relatively flat, but positive slope and Quadratic Occasion showed an inverted U shape. There was also a significant main effect of C ognitive G roup classification, F (2,2698.27 )= 205.80 p <.001, such that the covariate adjusted estimate of erformance was the best (mean=19.7 9 s.e.=0.09 ), followed by the non amnestic (mean=17.61, s.e.=0.12 ), with amnestic persons having the lowest performance (mean=16.72, s.e.=0.19 ). These differences between groups were all significantly different after Bonferroni alpha corrections. The Promp ting Status x Cognitive Group in teraction was also significant [ F (1, 3764.52 )= 33.28 p <.001] such that prompting improved the non amnestic OTDL P (3.35points). The unimpaired p articipants had the smallest change between the Unprompted and Prompted OTDL scores, improving by 2.83 points with prompting. However, while there was a smaller gap between the unimpaired participants and Prompted scores than Unprom pted the unimpaired Prompted performance was still better than that of the amnestic and non amnestic groups The Prompting Status x L inear O ccasion was again non significant, but the Prompting Status x Q uadratic O ccasion interaction was significant, F (1, 111 11.95 )=93.8 7 p <.001 This suggest s that while the Prompted OTDL P erformance remains fairly consistent over time, the Unprompted performance shows a quadratic trajectory of ini tial gain followed by decline. Figure 4 3 displays the trajectories of Prom pted versus Unprompted OTDL performance by Cognitive Group The Cognitive Group x Occasion

PAGE 56

56 interactions (both Linear and Quadratic) were non significant suggesting that there is not a di fference in trajectory of OTDL Performance between Cognitive G roup s Additionally, the three way Prompting Status x Cognitive Group x Occasion (both L inear and Q uadratic ) interactions were non significant. Random Effects Both the L inear and Q uadratic random effects of time were significant ( Wald Z= 9.84 p <.001 and Wald Z= 1 2.26 p <.001, respectively), which suggests that there are both linear and quadratic individual rates of change that are not accounted for by the fixed effects. The random effect of Prompting Status was significant ( Wald Z= 4.17 p <.001 ) suggesting that the re are inter individual differences in the effect of prompting. The random Prompting Status x Occasion (Linear and Quadratic) interactions were again included in the model, but were unable to be reliably estimated. The Prompting Status x Cognitive Group i nteraction was non significant, indicating that there were not significant inter individual differences in response to prompting by cognitive group that is not already accounted for by the mean (fixed) effect. The Cognitive Group x Linear O ccasion random effect was significant ( Wald Z =6.21, p <.001), indicating that regardless of prompting, there are individual differences in five year linear trajectories On the other hand, Cognitive Group x Quadratic Occasion effect was non significant. The three way Cog nitive Group x Prompting Status x Linear Occasion interaction was significant ( Wald Z =3.19, p =.001), suggesting there are inter individual differences in linear trajectory of Prompted versus Unprompted scores by Cognitive Group There was additional and s ignificant between and within subject variance that was left unexplained by this model. This model explained 49.2% of the within subject variance and 50.2% of the between subject variance. Relative to the model in Aim 2 this model,

PAGE 57

57 which included the Co gnitive Group and its interactions, is a significant improvement in model fit [ 2 (12)=572.09]. Incremental model fit is displayed in Table 4 3.

PAGE 58

58 Table 4 1. Fixed and random estimates for the full longitudinal model examining prompted vs. un prompted OTDL performance Fixed Effect Estimate S.E. df F statistic p Random Effect Estimate S.E. Wald Z p Level 2 (Between Subjects Effects ) Replicate 2592.97 2.69 0.020 Site 2599.10 19.32 <.001 Intervention Group 2599.59 1.54 0.160 BL Age 0.16 0.23 0.01 2623.68 232.79 <.001 BL Ed ucation 0.43 0.26 0.02 2637.90 345.23 <.001 African American 1.40 0.30 0.16 2629.78 79.82 <.001 Female 0.78 0.18 0.14 2620.71 33.25 <.001 BL Vision 0.04 0.09 0.01 2624.79 44.62 <.001 BL CES D 0.07 0.08 0.01 2636.16 31.89 <.001 BL SF 36 General Health 0.01 0.04 0.00 2628.79 9.12 0.003 Intercept 23.45 0.30 1.06 2628.01 428.43 <.001 Between subjects Residual 6.51 0.23 27.80 <.001 Level 1 (Within Subjects Effects) Prompting Status 3.06 0.68 0.04 2660.01 5318.08 <.001 Prompting Status 0.53 0.06 9.05 <.001 Occasion 0.11 0.05 0.02 1900.77 41.39 <.001 Occasion 0.30 0.02 16.60 <.001 Occasion 2 0.11 0.06 0.01 2214.82 108.86 <.001 Occasion 2 0.14 0.01 17.36 <.001 Prompting x Occasion 0.03 0.01 0.02 11933.67 1.33 0.248 Prompting x Occasion Prompting x Occasion 2 0.16 0.05 0.01 11202.47 154.11 <.001 Prompting x Occasion 2 Within subjects Residual 5.51 0.08 69.89 <.001 BL=Baseline occasion; CES D=Center for Epidemiological Studi es Depression 12 item scale; SF 36 General Health=MOS Short Form Health Survey, General Health subscale

PAGE 59

59 Table 4 2. Fixed and random estimates for the full longitudinal model examining prompted vs. unprompted OTDL performance by cognitive group Fixed Ef fect Estimate S.E. df F statistic p Random Effect Estimate S.E. Wald Z p Level 2 (Between Subjects Effects) Replicate 2541.99 1.90 0.092 Site 2549.18 19.99 <.001 Intervention Group 2548.31 2.30 0.033 BL Age 0.17 0.23 0.01 2591.20 306.65 <.001 BL Education 0.42 0.25 0.02 2615.30 402.15 <.001 African American 1.29 0.28 0.14 2584.80 81.29 <.001 Female 0.51 0.12 0.12 2573.96 16.69 <.001 BL Vision 0.03 0.07 0.01 2583.29 27.95 <.001 BL CES D 0.04 0.05 0.01 2588.43 15.48 <.001 BL SF 36 General Health 0.01 0.03 0.00 2583.82 5.78 0.016 Cognitive Group 2698.27 205.80 <.001 Intercept 23.49 0.03 0.97 2597.16 537.92 <.001 Between subjects residual 5.14 0.20 25.95 <.001 Level 1 (Within Subjects Effects) Prompting Status 3.08 0.68 0.04 2619.41 5409.51 <.001 Prompting Status 0.31 0.07 4.18 <.001 Occasion 0.11 0.05 0.02 1584.79 35.03 <.001 Occasion 0.19 0.02 9.84 <.001 Occasion 2 0.11 0.06 0.01 2203.11 102.87 <.001 Occasion 2 0.13 0.01 12.26 <.001 Prompting x Occasion 0.03 0.01 0.03 11814.82 1.34 0.247 Prompting x Occasion Prompting x Occasion 2 0.15 0.05 0.02 11111.95 93.87 <.001 Prompting x Occasion 2 Prompting x Gr oup 0.38 0.04 0.07 3764.52 33.28 <.001 Prompting x Group 0.31 0.17 1.81 0.071 Group x Occasion 0.05 0.01 0.03 1223.01 3.24 0.072 Group x Occasion 0.22 0.03 6.21 <.001 Group x Occasion 2 0.02 0.01 0.01 1747.01 2.26 0.133 Group x Occasion 2 0.00 0.01 0.4 3 0.671 Group x Prompting x Occasion 0.01 0.00 0.03 1195.83 0.28 0.596 Group x Prompting x Occasion 0.04 0.01 3.20 0.001 Group x Prom p ting x Occasion 2 0.02 0.01 0.02 10582.48 2.3 0.129 Group x Prompting x Occasion 2 Within subjects residu al 5.44 0.08 69.60 <.001 BL=Baseline occasion; Group Classification=unimpaired, non amnestic, or amnestic; CES D=Center for Epidemiological Studies Depression 12 item scale; SF 36 General Health=MOS Short Form Health Survey, General Health subscale; Grou p=Cognitive Group Classification (amnestic, non amnestic, or unimpaired)

PAGE 60

60 Table 4 3: Nested m odels Model 2LL AIC BIC 2 compared to previous 2LL parameters p value R 2 between R 2 within UM M 104477.35 104487.35 104506.91 3 Aim 2 95382.29 95454.29 95736.46 9095.06 36 <.001 0.37 0.49 Aim 3 94810.20 94906.20 95282.43 572.09 48 <.001 0.50 0.49 UM M=unconditional mean s model; See text for full Aim 2 and Aim 3 model descriptions

PAGE 61

61 Figure 4 1. Relationship between item difficulty and prompting on the OTDL. r = 0.71

PAGE 62

62 Figure 4 2 Trajectory of prompted vs. u nprompted performance on the OTDL Note Error bars represent the 95% confidence interval

PAGE 63

63 Figure 4 3 Trajectory of prompted vs. u n prompted performance on the OTDL by cognitive status group. Note Error bars represent the 95% confidence interval.

PAGE 64

64 CHAPTER 5 DISCUSSION Overview The pre sent study investigated minimally directive standardized verbal prompts on a measure of everyday cognition (OTDL), and whether this support led to better performance on the OTDL in older adults. The effect of prompting was investigated longitudinally over five years, which allowed the age related change of everyday cognitive functioning to be examined for both the Unprompted (raw score) and Prompted (raw score plus correct items after prompting) scores. Additionally, this study allowed an examination of w hether Prompting modified the trajectory of change relative to the Unprompted performance. P a rticipants were classified on an impairment gradient a ccording to a psychometric algorithm developed by Cook et al. ( under review) and were defined as unimpaired, amnestic mild cognitive impairment (MCI) or non amnestic MCI. Using these cognitive status groups, c omparisons between groups were first made on the number of supportive prompts required on a performance based measure of everyday cognition Then, comp arisons between groups examined initial everyday cognitive functioning performance, followed by a longitudinal investigation of whether the impaired groups proceed to become disproportionately compromised in everyday cognition and whether there are differ ential effects of prompting associated with impaired cognitive status. Review of Findings Aim 1: Number of Verbal Prompts G iven on the OTDL by Cognitive Status Group The first aim of this study was a preliminary investigation to determine whether there w ere differences in the number of verbal prompts on the OTDL received by

PAGE 65

65 participants at BL based on cognitive group classification (i.e., amnestic, non amnestic, unimpaired). The results showed that at BL, the unimpaired participants received significantl y fewer prompts than either of the impaired groups. In addition, the number of prompts received by the amnestic and non amnestic groups did not significantly differ from one another. This result is slightly different than the hypothesized result that non amnestic persons would require more prompts at BL. However, the actual result is not unexpected given that more than half of the amnestic participants were also impaired in non memory domains (multidomain amnestic). In previous research, multidomain amn estic persons have been shown to be the most impaired of the MCI subgroups on instrumental activities of daily living, while single domain amnestic participants did not differ from non amnestic subtypes (IADLs; Kim et al., 2009) Aim 2: Prompted Vs. Unpromp ted Performance on the OTDL from Baseline to Year 5 This aim investigated if there was a difference in performance on the OTDL when the U n prompted score is compared to the Prompted score over all five occasions. The analysis revealed that Prompting was be neficial in boosting performance on the OTDL by approximately 3 points across participants and occasions. It is important to note that the Prompted score was still approximately 7.5 points below the maximum score of 28, indicating that the Prompted score was still relatively far from the ceiling for this measure. Both Linear and Quadratic Occasion effects were significant (show in Figure 4 2) such that the linear trend is relative flat, with a slightly positive improvement in OTDL performance over time. The quadratic trend can be observed as having as initial improvement in performance at the earlier occasions, followed by decline at later occasions. This is likely attributed to initial practice effects followed by decline as a

PAGE 66

66 result of age related chan ge as most participants transitioned from young old to old old age. This decline is consistent with the trajectory that is often seen in more basic cognitive function in that performance starts to decline as people transition into old old age (e.g., Singe r Verhaeghen, Ghisletta, Lindenberger, & Baltes, 2003). Indeed, the ACTIVE sample, on average, moved into old old age throughout the course of the study. The quadratic effec ts of time, however, were qualifie d by the significant Prompting S tatus (Prompted vs. Unprompted) by Quadratic Occasion interaction, in which the Unprompted score primarily showed the quadratic trajectory over the five years, while the Prompted score remai ned stable over time. The flatter trajectory of the Prompted score suggests that the use of a minimal amount of verbal support may in fact have allowed everyday cognition to be more resistant to the negative effects of aging. In this model that examined Prompted and Unprompted performance over time, there were significant random effects that indicate intra individual differences that were not accounted for by the fixed effects of the model. The slopes of both the linear and quadratic time differed across individuals. Also, there were individual differences in responses to Prompting with some participants likely improving more than others. The model explained approximately 36.2% of the between subject variance and 48.5% of the within subject variance. T his model was a significant impr ovement in fit relative to the Unconditional Means M odel. Aim 3: Longitudinal Re lationship of Cognitive Status G roup and Prompted vs. Unprompted P erformance on the OTDL This aim investigated if there is a difference in per formance on the OTDL based on cognitive status classification and whether impaired participants (amnestic or non amnestic) benefit more or less from prompting than the unimpaired group. T here were

PAGE 67

67 level differences between the unimpaired and impaired grou ps in performance on the OTDL with the unimpaired performing the best, followed by the non amnestic group, with the amnestic group performing the worst. Gross et al. (2011) found that there was greater variance explained by tests of reasoning than memory on measures of everyday functioning. Thus, we expected that the non amnestic group would be the lowest performing, particularly at BL. Instead, we found that the amnestic group performed the worst on the OTDL. However, the fact that the amnestic group was the lowest performing is not surprising given previous work demonstrating that amnestic MCI persons, especially those are who are impaired in memory in addition to other domains, are generally poorer functioning than non amnestic MCI participants (e.g., K im et al., 2009; Teng ., Becker, Woo, Cummings, & Lu, 2010; Wadley et al., 2007) and are more likely to progress to a dementia (e.g., Palmer, Backman, Winblad, & Fratiglioni, 2008; Yaffe, Petersen, Lindquist, Kramer, & Miller, 2006). This also makes sens e given that the majority of the amnestic participants ( 58.7 %) were in fact impaired in multiple domains, as opposed to only 32.5% of non amnestic participants. Additionally, as described in Cook et al. (under review), a larger proportion of amnestic pers ons progressed to dementia at later occasions compared to non amnestic and unimpaired participants. The Prompted performance represented an improvement over the Unprompted scores across all three cognitive groups. The trajectories of Prompted and Unpromp ted scores did not differ by Cognitive Group. The shape of the trajectories were all similar to those described above in Aim 2 with Prompted scores remaining stable over the five years (flatter) and Unprompted scores showing initial gain followed by decli ne at the

PAGE 68

68 later occasions (inverted u shape). Interestingly, there was a significant Prompting Status by Cognitive Group interaction such that the non amn estic and amnestic OTDL performance showed greater improvement (3.64 and 3.35 points, respectively) t han the surpass appears that with prompting, impaired participants were able to perform at or above the functional level of those participants without impairment. As with the previous model in Aim 2, this model as had significant random effects that indicate intra individual differences in both linear and quadratic rates of change in performance on the OTDL as well as individual differences in responses to prompting. Additionally, whil e the overall (fixed) Cognitive Group by L inear O ccasion interaction was not significant, there was person to person variation in rates of change within groups Furthermore, while there was not a significant overall Prompting Status by C ognitive G roup by L inear O ccasion interaction, the re were individual differences in linear trajectory of Prompted versus U nprompted scores by Cognitive Gr oup. This model that included the Cognitive Group variable and its interaction is a significant improvement from the pre vious (Aim 2) model. It explained approximately 49.5% of the between subject variance and 49.2% of the within subject variance. Thus, this model improves the understanding of between subject differences in OTDL performance However, the addition of the gr oup variable does not seem to moderate the time trend, which is consistent with the non significant Cognitive Group by O ccasion

PAGE 69

69 interactions. As such, the amount of within subject variance explained in this model is almost identical to that explained in t he Aim 2 model. Major Theoretical Issues This investigation offers support to the idea that performance based measures of instrumental or complex activities of daily living may, in fact, be rather sensitive to decline in everyday functioning that is not ye report of functioning. Particularly at BL, for inclusion in the study all participants w all living independently and screened o ut if they had an MMSE 23 or reported extreme difficulty with basic ADLs such as bathing, dressing, or personal hygiene Moreover, about 50% of participants endorsed no self reported IADL limitations at BL ( Michael Ma rsiske, personal communication). However, there is still a large range of OTDL performance scores with 99% of the sample incorrectly responding to at least two items at BL Thus, the performance based OTDL appears to be rather sensi tive to errors/difficulties that other assessment schemes miss Very little wor k has been done to investigate the longitudinal trajectory of everyday cognition (Willis et al., 1992; Willis, 1996). This investigation, however, allowed for the examination of the 5 year trajectory of performance on a measure of everyday cognition and a ttempted to address the question of whether complex functioning showed relative preservation or decline in older adults over time The trajectory of the OTDL performance showed relative absolute improvement in that the Year 5 performance was, on average, higher than the BL performance. Specifically, the steady improvement from BL through Year 3 was greater than the decline that was observed from Year 3 to Year 5 fo r the Unprompted performance. The initial improvement is likely a result of increasing fami liarity and practice with the tasks on the OTDL This

PAGE 70

70 early trajectory is consistent with a psychometric measure pattern of practice when ite ms have been repeatedly tested. The repeated exposure to the items on the OTDL allowed participants to become fam iliar with the items and potentially remember it ems from occasion to occasion. This trajectory of improvement at the early occasions is important to address because it has implications for repeated clinical usage of the OTDL to assess everyday cognitive f unctioning. For example, while there may be true subtle decline in everyday cognition over time, this may not be captured by the OTDL if there is repeated use, particularly within a short period of time since there are not alternative forms of the OTDL. On the other hand, the Prompted score trajectory did not show the practice effects that were observed in the Unprompted trajectory. This is likely a result of prompting boosting performance to a level that is consistent with personal best which sti ll appears to be several points below the measure ceiling of 28 points. This finding of a maximum individual performance is consistent with the concept They suggested that there is a max imum baseline performance depending on varying task demands and support. Prompting may act as a support similar to practice ; therefore, with the OTDL prompts, there is no more ro om for incremental gains due to practice since both prompting and practice co mmonly activate the baseline reserve. Nevertheless, the decline in Unprompted OTDL performance at later occasions (Year 3 through Year 5) of this study suggests that practice and familiarity is challenged by age related decline. While there was initial ga in at the early occasions, as the mean age of participants transitioned from young old to old old age, we began to see a more universal or normative decline from Year 3 to Year 5. This is consistent with previous

PAGE 71

71 cross sectional work that demonstrated age related decline in everyday cognition (Marsiske & Margrett, 2006; Thornton and Dumke, 2005; Willis & Marsiske, 1991). This trajectory of decline also maps on to the longitudinal find ings of Wills (1996) in which the resistance to decline begins to disappea r as older adults transition from young old t o old Thus, this investigation further supports the notion that normative decline is observed in everyday cognition and suggests that change in everyday cognitive performance is similar to the age related decline that is observed in more basic cognitive functioning (e.g., Singer et al., 2003). This study also supports the work that has reported that IADLS are sensitive to the early cognitive decline seen in MCI participants (Allaire & Wi llis, 2006; Goldberg et al., 2010; Perneczky et al., 2006; Tuokko, Morris, & Ebert, 2 005). We were surprised that it was the amnestic group that performed the worst on the OTDL. Our initial hypothesis that the non amnestic group would have the lowest per formance was based on work that suggested that reasoning explained the most variance in everyday cognitive performance (Burton et al., 2006; Gross et al., 2011; Willis et al., 1992). However, the non am n e stic participants are a very heterogeneous mix of n ot only participants with reasoning and problem solving impairments (Complex cognition domain), but also Attention, Language, and Visuospatial impairments. Similarly, the am n e stic group consists of multidomain amnestic participants. As such, because thes impairment groups ( i.e., amnestic group has 58.7 % multidomain amnestic participants ; non amnestic has 32.5 % multidomain non amnestic participants) and multidomain amnestic participants are generally the lowest functioning MCI subtype, the current finding is reasonable.

PAGE 72

72 Limitations The present study has several limitations that will be important to consider when developing future studies to replicate and elaborate on these findings. As frequently observed in longitudinal studies of older ad ults, s elective attrition can be a source of significant bias. It is typically seen that participants demonstrating greater impairment (e.g., health related, dementia) are more likely to drop out of the study prior to conclusion (e.g., Siegler & Botwinick 1979). This pattern was also observed in this sample with cognitively impaired persons (amnestic and non amnestic) being more likely to dropout prior to the 5 year follow up than the unimpaired participants (Cook et al., under review ). In this study, v ariables that are predictive of attrition in this sample (covariate dependent attrition) were controlled for in the analyses, which could be considered a strength of this investigation. However, future analyses of this or a similar sample may consider usi ng pattern mixture modeling (i.e., where separate estimates are derived for groups with different missing data patterns, and then the final estimates are pooled across these missing data subgroups) for another way of estimating data due to attrition ( Littl e, 1993; Hedeker & Gibbons, 1997 ). The sample used in this investigation was specifically selected for participation in longitudinal cognitive intervention study. As described in the methods section above, participants were generally healthy and able t o function independently, as to ensure that they could comply with the demands of the cognitive training intervention. However, this selection bias produced a group of older adults who are less likely to experience physical, cognitive or functional proble ms compared to the general population. Because this was a sample of generally healthy older adults, the sample size of the unimpaired group to the amnestic and non amnestic groups is considerably unbalanced. Then,

PAGE 73

73 again, due to the attrition bias, impair ed participants were even more underrepresented at later occasions. conditions ( e.g., in groups, written versus verbal responding in memory tests) in order to account for the substantial number of participants (i.e., 2,802) tested at multiple sites with multiple follow up evaluations. While this approach is more widely used in this literature, it is also acknowledged that, from a neuropsychological perspective, the deviation f rom standard administration may introduce additional error that, in essence, believe valuable information can be attained through the use of the cognitive impairment groups identified by Coo k and colleagues ( under review) that are used in this study the ACTIVE study and the cognitive measures were not designed for the purpose of classifying cognitive impairment. While the OTDL is generally a good representation of the construct of everyday 1995, 1998 ) this measure does have some limitations. First, the measure was delivered in a research context and was not an observation of functioning at home, which limits the generaliza tion to real world functioning since the environment was unfamiliar. Furthermore, there are components of the delivery of the measure that could introduce additional variance to the measure. For example, the participant was required to read each item off of an index card. Thus, while vision was corrected for in Aims 2 and 3, literacy may affect the While this could be considered a limitation, work by Kirsch & Mosenthal (1990) su ggest ed that

PAGE 74

74 Also, prompts were delivered if the participant said that they did not know the answer or took too long t o respond; however, they were not given if the participants simply gave the wrong answer. Therefore, participants who were impulsive may not have received prompts because they gave an answer quickly, even when they did not know if it was correct. The OTD L has no way of capturing who gave these types of impulsive responses. T his study does not specifically answer the question behind why people needed prompts or why verbal prompting was effective. The idea that verbal prompting is an effective way to impro ve everyday cognitive functioning across time and impairment type is interesting and a potential strength from a broader, public health perspective. However, from a neuropsychological perspective, this study does not sufficiently investigate specific mech anisms. For example, when developing cognitive intervention programs, it would be beneficial to know if simple verbal prompts are effective for people with poor self initiation or problems with attention or both. Future Directions This investigation provi des a basis for continued research in the use of prompting as a method for improving everyday cognition. First, it showed that simple verbal prompts can not only improve performance on a task of everyday cognition, but that they can delay the decline of p erformance over five years. This finding was true across participants, including those considered cognitive ly impaired; however, the mechanism by which verbal prompting was effective was not examined in this investigation. Future research should address possible mechanisms for the effectiveness of verbal prompting on tasks of everyday cognition, including the possibility that the prompting serves as an

PAGE 75

75 environmental support and improve s deficient self initiated processing (Craik, 1983, 1986; Craik and Jen nings, 1992 ), reorients participants to the task if they are having difficulty focusing their attention, or potentially provides motivation for a participant to keep trying, particularly if they experience low self efficacy. Figure 5 1 displays some possib ilities that should be further explored in future work. Just o ne of the many potential reason s that participants needed prompts is due to item difficulty. This was investigated in a post hoc analysis, which was described in the results section. This cor relation revealed that item difficult was strongly associated with the need for prompts, and is s hown in Figure 4 1 It is important to understand the mechanisms for prompting success for the following reasons. First, this would allow the content of the prompt to focus in on the reason the prompt is needed. For example, if low self efficacy Second, it is important to know why prompting works so that cognitive rehabilitation programs can teach prompting strategies that are proven effective and most likely to maximize the intervention success. It is likely that participants with differ ent impairments or difficulty with everyday cognitive functioning require prompting for different reasons. As such, future studies may include an analysis examining how the mechanism of verbal prompting may differ by cognitive impairment type. This would be useful when developing cognitive rehabilitation programs in order to tailor the program to the person. While this investigation offers a unique opportunity to examine the effect of verbal prompting on a performance based measure of everyday cognition, it is limited by the

PAGE 76

76 retrospective use of study measures to classify impaired participants. Future investigations of mild cognitive impairment and the effects of verbal prompting on their ability to perform tasks of everyday cognition should use a prospec tively clinically classified group of MCI participants. Additionally, based on this study, MCI participants respond well to the verbal prompts as administered in a structured environment; however, what remains to be answered is whether people with MCI can learn to use external prompting strategies in real world, less structured environments. In the age of technology such as computers and smart phones, future research is needed on the use of these tools in implementing a prompting system that can be used by the masses. For example, can an external prompting device such as daily reminder alarms medication as prescribed? If this is shown to be effective and a feasible tool for older adults to learn, this could potentially allow older adults to maintain their independence for longer, and, in turn, reduce the hours of caregiver assistance needed or delay the transition to assisted living. While the uses of assistive technology have be en been limited work exami ning the broad range of possible types of prompts used in terms in that prompts were only given when needed and did not provide direct instructions to the participants dementia research, prompts include detailed step by step instructions. These types of prompts make sense for severely impaired persons; however, in normal to mildly

PAGE 77

77 impaired people, would this level of detail actually be detrimental in that it promotes a dependence on the prompts? Along the lines of the risk of becoming depen dent on prompting, it might be worth investigating if prompting could be used as a meta cognitive external to internal training approach. For example, could you start with the use of prompting as a training tool, but then over time, fade the prompts away while teaching the person to use internal away from depending on external prompts to function and an increase in the use of internal prompts that the person has learne d through the initial exposure to external support. This analysis would examine the longitudinal use of external prompting with the question of whether we can train people to prompt themselves, effectively reducing the need for prompts. Final Conclusions This study showed that prompt provision increased performance across all participants. The effect of prompting differed by cognitive status (impaired vs. unimpaired) and was highest in the impaired participants. Prompt provision did boost level of perfo rmance ( by 19% across all participants ), but while the gap was narrowed, it did not eliminate impairment group differences in performance. Prompt use did attenuate rate of decline over five years

PAGE 78

78 Figure 5 1. Possible reasons that participants neede d prompting

PAGE 79

79 APPENDIX: OBSERVED TASKS OF DA ILY LIVING

PAGE 80

80

PAGE 81

81

PAGE 82

82

PAGE 83

83

PAGE 84

84

PAGE 85

85

PAGE 86

86

PAGE 87

87

PAGE 88

88

PAGE 89

89

PAGE 90

90

PAGE 91

91

PAGE 92

92

PAGE 93

93 LIST OF REFERENCES Allaire, J. C., & Marsiske, M (1999). Everyday cognition : Age and intellectual ability correlates Psychology and Aging, 14 627 644. Allaire, J.C., & Marsiske, M. (2002). Well and ill defined measures of everyday cognition: Relat Psychology and Aging, 17 101 115. Allaire, J. C., & Willis, S. L. (2006). Competence in everyday activities as a predictor of cognitive risk and mortality. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology and Cognition 13 (2), 207 224. Albert, M.S., DeKosky, S.T., Dickson, D., Dubois, B., Feldman, H.H., Fox, N.C., endations from the National Institute on Aging disease. The Journal of the Alzheimer's Association 7 270 279. Alzheimer's Association. (2011). 2011 Alzheim er's Disease Facts and Figures. Alzheimer's & Dementia: The Journal of the Alzheimer's Association 7 208 44. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (4 th ed.). Washington, DC: American Psychiatric As sociation Press. Willis, S.L. for ACTIVE Study Group (2002). Effects of cognitive training interventions with older adults: a randomized controlled trial. JAMA: The Journal of the American Medical Association 288 (18), 2271 2281. Ball, K. & Owsley, C. (1993). The Useful Field of View Test: A new technique for evaluating age related declines in visual functioning. Journal of the American Optometric Association, 64 71 79. Baltes M. M., Burgess, R.L, & Stewart, R. B. (1980). Dependence and independence in self care behaviors in nursing home residents: An operant observational study. International Journal of Behavioral Development, 3 ,489 500. Barton, E.M., Baltes, M.M., & Orzech, M.J. (1980). Etiology of dependence in older nursing home residents during morning care: The role of staff behavior. Journal of Personality and Social Psychology, 38( 3), 423 431. Berg C.A. & Klaczynski, P. (1996). Practical intelligence and problem solving In: F. Blanchard Fields & T.M. Hess (Eds .). Perspectives on cognitive change in adulthood and aging (pp. 323 357). New York: McGraw Hill.

PAGE 94

94 Berry J & Irvine S. (1986). Bricolage: Savages do it daily. In: R.J. Sternberg & R. Wagner (Eds.). Practical intelli gence ( pp. 236 270) New York: Cambridge University Press. Bewernitz, M.W., Mann, W.C., Dasler, P., & Belchior, P. (2009) Feasibility of Machine Based Prompting to Assist Persons With Dementia. Assistive Technology, 21(4), 196 207. Blanchard Fields, F. & Chen, Y. (1996). Adaptive cognition and aging. American Behavioral Scientist, 39 231 248. Brandt, J. (1991). The Hopkins Verbal Learning Test: Development of a new memory test with six equivalent forms. Clinical Neuropsychology 5, 125 142. Bryk, A. S. & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks: Sage. Burton, C. L., Strauss, E., Hultsch, D. F., & Hunter, M. A. (2006). Cognitive functioning and everyday problem solving in older adults. The Clinical Neuropsychologist 20 (3), 432 52. Cahn Winer, D.A., Malloy, P.F., Bole, P.A., Marran, M., & Salloway, S. (2000). Prediction of functional status from neuropsychological tests in community dwelling elderly individuals. The Clinical Neuropsychologi st 14 187 195. Cahn Weiner, D.A., Boyle, P.A., & Malloy, P.F. (2002). Tests of executive function predict instrumental activities of daily living in community dwelling older individuals. Applied Neuropsychology, 9, 187 191. Cook, S. E. & Marsiske, M. (20 06). Subjective memory beliefs and cognitive performance in normal and mildly impaired older adults. Aging and Mental Health, 10 413 423. Cook, S.E., Marsiske, M., Thomas, K.R., Unverzagt, F.W., Wadley, V.G., Langbaum, J.B.S., Crowe, M. ( under review ). Id entification of Mild Cognitive Impairment in ACTIVE: Algorithmic Classification and Stability. Journal of the International Neuropsychological Society. Cornelius, S.W. & Caspi, A. (1987). Everyday problem solving in adulthood and old age. Psychology and Ag ing, 2 144 153. Craik, F. I. M. (1983). On the transfer of information from temporary to permanent memory. Philosophical Transactions of the Royal Society, London, Series B, 302, 341 359. Craik, F. I. M. (1986). A functional account of age differences in memory. In F. Klix & H. Hagendorf (Eds.), Human memory and cognitive capabilities (pp. 409 442). New York: Elsevier Science.

PAGE 95

95 Craik, F. I. M., & Jennings, J. M. (1992). Human memory. In F. I. M. Craik & X A. Salthouse (Eds.), Handbook of cognition and agin g ( pp 51 110). Hillsdale, NJ: Erlbaum. Crowe, M., Andel, R., Wadley, V., Cook, S., Unverzagt, F., Marsiske, M. & Ball, K. (2006). Subjective cognitive function and decline among older adults with psychometrically defined amnestic MCI. International Journa l of Geriatric Psychiatry, 21 1187 1192. DeJager, C., Blackwell, A. D., Budge, M. M. & Sahakian, B. J. (2005). Predicting cognitive decline in healthy older adults. American Journal of Geriatric Psychiatry, 13, 735 740. Demming, J.A. & Pressey, S.L. (195 Journal of Counseling Psychology, 4 144 148. Denney N.W. (1989). Everyday problem solving: Methodological issues, research findings, and a model. In: L.W. Poon, D.C. Rubin, B.A. Wilson (Eds.), Everyday cognition in adulthood and late life (pp. 330 351). New York: Cambridge University Press. Diehl, M., Marsiske, M., Horgas, A.L., Rosenberg, A., Saczynski, J.S., & Willis, S.L. (2005). The revised observed tasks of daily living: A performance based assess ment of everyday problem solving in older adults. Journal of Applied Gerontology 24, 211 230. Diehl, M., Marsiske, M., Horgas, A.L., Saczynski, J.S. (1998). Psychometric Properties of the Revised Observed Tasks of Daily Living (OTDL R). Paper presented a t: Annual Meeting of the Gerontological Society of America; November 20 23, 1998. Philadelphia, PA. Diehl, M., Willis, S.L., & Schaie, K.W. (1995). Everyday problem solving in older adults: Observational assessment and cognitive correlates. Psychology a nd Aging 10, 478 491. Ebly, E. M., Hogan, D. B., & Parhad, I. M. (1995). Cognitive impairment in the nondemented elderly: Results from the Canadian Study of Health and Aging. Archives of Neurology 52 612 619. Educational Testing Service (1977). Test of basic skills Princeton, NJ: Educational Testing Service. Edwards, J. D., Vance, D. E., Wadley, V. G., Cissell, G. M., Roenker, D. L., & Ball, K. K. (2005). Reliability and validity of the Useful Field of View test scores as administered by personal comput er. Journal of Clinical and Experimental Neuropsychology, 27 529 543.

PAGE 96

96 Ekstrom, R. B., French, J. W., Harman, H., & Derman, D. (1976). Kit of Factor Referenced Cognitive Tests rev. ed. Princeton, NJ: Educational Testing Service. Flannery, M., & Rice, D. (1 997). Using available technology for reminding. Proceedings of the Rehabilitation Engineering Society of North America (RESNA) (pp. 5 17 5 19). Arlington, VA: RESNA Press. Folstein, M.F., Folstein, S.E., McHugh, P.R. (1975). Mini mental state: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12 189 198. Ganguli, M., Chang, C C. H., Snitz, B. E., Saxton, J. A., Vanderbilt, J. & Lee, C W. (2010). Prevalence of Mild Cognitive Impairment by mult iple classifications: The Monongahela Youghiogheny Healthy Aging Team (MYHAT) Project. American Journal of Geriatric Psychiatry, 18, 674 683. Ganguli, M., Dodge, H. H., Shen, C. & DeKosky, S. T. (2004). Mild cognitive impairment amnestic type: An epidemio logic study. Neurology, 63, 115 121. Technology for Cognition: A Systematic Review. Journal of the International Neuropsychological Society, 18 1 19. Goldberg, T.E., Koppel, J ., Keehlisen, L., Christen, E., Dreses Werringloer, U., Conejero Goldberg, C., Gordon, M.L., & Davies, P. (2010). Performance Based Measures of Everyday Function in Mild Cognitive Impairment. The American Journal of Psychiatry, 167, 845 853. Gonda, J., & S chaie K. W. (1985). Schaie Thurstone Mental Abilities Test: Word Series Test Palo Alto, CA: Consulting Psychologists Press. Graham, J. E., Rockwood, K., & Beattie, B. L. (1997). Prevalence and severity of cognitive impairment with and without dementia in an elderly population. Lancet, 349, 1793 1796. Grigsby, J., Kaye, K., Baxter, J., Shetterly, S.M., & Hamman, R. (1998). Executive cognitive abilities and functional status among community dwelling older persons in the San Luis Valley Health and Aging Study Journal of the American Geriatrics Society, 46 590 545. Gross, A.L., Rebok, G.W., Unverzagt, F.W., Willis, S.L., & Brandt, J. (2011). Cognitive predictors of everyday functioning in older adults: results from the active cognitive intervention trial. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 66(5) 557 566. Hartley, A.A. (1989). The cognitive ecology of problem solving. In: L.W. Poon, D.C. Rubin, & B.A. Wilson (Eds.). Everyday cognition in adulthood and late life ( pp. 300 329). New York: Cambridge University Press.

PAGE 97

97 Hedeker, D., & Gibbons, R.D. (1997). Application of random effects pattern mixture models for missing data in longitudinal studies. Psychological Methods, 2, 64 78. Hershey, D.A. & Farrell, A.H. (1996). A ge differences on a procedurally oriented test of practical problem solving. Journal of Adult Development, 6 87 104. Insel, K., Morrow, D., Brewer, B., & Figueredo, A. (2006). Executive function, working memory, and medication adherence among older adults Journals of Gerontology: Psychological Sciences and Social Sciences, 61B (2), 102 107. Jak, A.J., Bondi M.W., Delano Wood, L., Wierenga, C., Corey Bl oom, J., Salmon, D.P., et al. ( 2009). Quantification of five neuropsychological approaches to defining mild cognitive impairment. American Journal of Geriatric Psychiatry 17 368 375. Jobe, J.B., Smith, D.M., Ball, K., Tennstedt, S.L., Kleinman, K. (2001). ACTIVE: A cognitive intervention trial to promote independence in olde r adults. Controlled Clinical Trials 22 453 479. Jorm, A. F., Christensen, H., Korten, A. E., Henderson, A. S., Jacomb, P. A. & Mackinnon, A. (1997). Do cognitive complaints either predict future cognitive decline or reflect past cognitive decline? A lo ngitudinal study of an elderly community sample. Psychological Medicine, 27 91 98. Kahn, R. L. (1971). Psychological aspects of aging. In I. Rossman (Ed.), Clinical geriatrics. J. B. Lippincott, Philadelphia. Kahn, R.L. (1975). The Mental Health System a nd the Future Aged. The Gerontologist, 15, 24 31. Kim, K.R., Lee, K.S., Cheong, H.K., Eom, J.S., Oh, B.H., & Hong, C.H. (2009). Characteristic Profiles of Instrumental Activities of Daily Living in Different Subtypes of Mild Cognitive Impairment. Dem entia and Geriatric Cognitive Disorders, 27 278 285. Kirasic, K.C., Allen, G.L, Dobson, S.H., Binder, K.S. (1996). Aging, cognitive resources, and declarative learning. Psychology and Aging, 11 658 670. Kirsch, I.S. & Mosenthal, P.B. (1990) Exploring doc ument literacy: Variables underlying the performance of young adults. Reading Research Quarterly, 25 5 30. Kliegl, R. & Baltes, P.B. (1987). Theory guided analysis of mechanisms of development and aging through testing the limits and research on expertise In C. Schooler & K.W. Schaie (Eds.), Cognitive functioning and social structure over the life course (pp. 95 119). Norwood, NJ: Ablex.

PAGE 98

98 Labelle, K. & Mihailidis, A. (2006). The use of automated prompting to facilitate handwashing in persons with dementia. American Journal of Occupational Therapy, 60 (4), 442 450. Lawton, M. P., & Brody, E. M. (1969). Assessment of older people: Self maintaining and instrumental activities of daily living. The Gerontologist, 9 (3), 179 186. Lindenberger, U., Mayr, U., & Klie gl, R. (1993). Speed and intelligence in old age. Psychology and Aging 8, 207 220. Little, R. (1993), Pattern mixture models for multivariate incomplete data, Journal of the American Statistical Association 88 125 134. Lo Presti, E, F., Mihailidis, A., & Kirsch, N. (2004). Assistive technology for cognitive rehabilitation: State of the art. Neuropsychological Rehabilitation, 14 (1/2), 5 39. Mangione, C.M., Phillips, R.S., Seddon, J.M., Lawrence, M.G., Cook, E.F., Dailey, R. & Goldman, L. (1992). Developm measure of visual functional status. Medical Care, 30 111 126. Margrett, J. A., and Marsiske, M. (2002). Gender differences in adults' everyday cognitive collaborations International Journal of Behavioral Development, 26, 45 59. Marsiske, M. & Margrett, J. A. (2006). Everyday problem solving and de cision making In J. E. Birren & K. W. Schaie (Eds.). Handbook of the Psychology of Aging (6th Edition, pp. 315 342). New York: Academic Press. Marsiske, M. & Willis, S. L. (1995). Dimensionality of everyday problem solving in older adults Psychology and Aging, 10 269 283. Marsiske, M. & Willis, S. L. (1998). Practical creativity in older adult problem solving: Life span perspectives In C. E. Adams Price (Ed.), Creativity and successful aging: Theoretical and empirical approaches (pp. 73 113). New York: Springer. Mihailidis A., Barbenel, J, C., & Fernie, G. (2004). The efficacy of an intelligent cognitive orthosis to facilitate handwashing by persons with moderate to severe dementia, Neuropsychology Rehabilitation, 14 (1/2), 135 171. Mihailidis, A., Fernie, G., & Cleghborn, W. (2000). The development of a computerized cueing dev ice to help people with dementia to be more independent. Technology and Disability, 13 23 40. Morrell, R.W., Park, D.C., & Poon, L.W. (1989). Quality of instructions on prescription drug labels: Effects on memory and comprehension in young and old adults. The Gerontologist, 29 345 354.

PAGE 99

99 Morris, J.N., Fries, B.E., Steel, K., Ikegami, N., Bernabei, R., Carpenter, G.I., Topinkov, E. (1997). Comprehensive clinical assessment in community setting: Applicability of the MDS HC. Journal of the American Geriatr ic Society, 45 (8) 1017 1024. Neisser, U. (1978) Memory: What are the important questions? In: M.M. Gruneberg, P.E. Morris, R.N. Sykes (Eds). Practical aspects of memory (pp. 3 24). London: Academic Press. Owsley, C., Sloane, M.E., McGwin, G.J., & Ball, K. (2001). Timed instrumental activities of daily living tasks: Relationship to cognitive function and everyday performance assessments in older adults. Gerontology, 48 254 265. Palmer, K., Backman, L., Winblad, B., & Fratiglioni, L. (2008). Mild cognitive impairment in the general population: Occurrence and progression to Alzheimer disease. American Journal of Geriatric Psychiatry 16 603 611. Perneczky, R., Pohl, C., Sorg, C., Hartmann, J., Komossa, K., Alexopoulos, P., Kurz, A. (2006). Complex activit ies of daily living in mild cognitive impairment: Conceptual and diagnostic issues. Age and Ageing 35 240 245. Petersen, R. C. (2004). Mild cognitive impairment as a diagnostic entity. Journal of Internal Medicine, 256, 183 194. Radloff, L.S. (1977). The CES D Scale: A self report depression scale for research in the general population. Applied Psychological Measurement, 1 385 401 Rendell, P.G., & Craik, F.I. (2000). Virtual week and actual week: Age related differences in prospective memory. Applied Co gnitive Psychology, 14 S43 S62. Rendell, P.G. & Thomson, D.M. (1999). Aging and prospective memory: Differences between naturalistic and laboratory tasks. Journal of Gerontology : Psychological Sciences, 54B 256 269. Rey, A. (1941) psychologique Archives de Psychologie 28, 21. Ritchie, K., Artero, S. & Touchon, J. (2001). Classification criteria for mild cognitive impairment: A population based validation study. Neurology, 56, 37 42. Royall, D.R., Laut erbach, E.C., Kaufer, D., Malloy, P., Coburn, K.L., Black, K.J., & The Committee on Research of the American Neuropsychiatric Association. (2007). The Cognitive Correlates of Functional Status: A Review From the Committee on Research of the American Neur opsychiatric Association. The Journal of Neuropsychiatry and Clinical Neurosciences, 19, 249 265

PAGE 100

100 Salthouse T. A. (1990). Cognitive competence and expertise in aging. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the psychology of aging (3rd ed., pp. 310 319). San Diego, CA: Academic Press. Schaie, K. W. (1978). External validity in the assessment of i ntellectual development in adulthood. Journal of Gerontology, 33, 695 701. Schmidt, M. (2004). Rey Auditory and Verbal Learning Test: A handbook. Los Ang eles, CA: Western Psychological Services. Siegler, I.C. & Botwinick, J. (1979). A long term longitudinal study of intellectual ability of older adults: The matter of selective subject attrition. Journal of Gerontology, 34(2), 242 245. Singer, J.D., & Wille tt, J. B. (2003). Applied longitudinal data analysis : modeling change and event occurrence Oxford ; New York: Oxford University Press. Singer, T ., Verhaeghen, P., Ghisletta, P., Lindenberger, U., Baltes, P.B. (2003 ). The fate of cognition in very old age: Six year longitudinal findings in the Berlin Aging Study (BASE ). Psychology and Aging 18(2) 318 331. Sinnott J.D. (1989). Summary: Issues and directions for everyday problem solving research. In: Sinnot JD, editor. Everyday problem solving: Theory and application ( pp. 300 306) New York: Praeger. Sternberg R.J. & Wagner R.K. (1986). Practical intelligence: Nature and origins of competence in the everyday world. New York: Cambridge University Press. Teng, E., Becker, B.W., Woo, E., Cummings, J.L., & Lu, P.H. (2010). Subtle Deficits in Instrumental Activities of Da ily Living in Subtypes of Mild Cognitive Impairment Dementia and Geriatric Cognitive Disorders 30, 189 197. Thornton, W. L., Deria, S., Gelb, S., Shapiro, R. J., & Hill, A. (2007). Neuropsychological mediators of the links among age, chronic illness, and everyday problem solving. Psychology and Aging, 22 (3), 470 481. Thornton, W.L., & Dumke, H.A. (2005). Age differences in everyday problem solving and decision making effectiveness: A meta analytic review. Psychology and Aging, 20, 85 99. Tuokko, H., Morri s, C. & Ebert, P.(2005). Mild cognitive impairment and everyday functioning in older adults. Neurocase, 11( 1), 40 47. Promoting independence through individualized s trategies. Journal of Gerontological Nursing, 22 (3), 39 42.

PAGE 101

101 Wadley, V. G., Crowe, M., Marsiske, M., Cook, S. E., Unverzagt, F. W., Rosenberg, A. L. & Rexroth, D. (2007). Changes in everyday function among individuals with psychometrically defined Mild Cog nitive Impairment in the Advanced Cognitive Training for Independent and Vital Elderly Study. Journal of the American Geriatrics Society, 55, 1192 1198. Wagner R.K. (1986). The search for interterrestrial intelligence. In: Sternberg RJ, Wagner RK, editors. Practical intelligence: Nature and origins of competence in the everyday world ( pp. 51 83) New York: Cambridge University Press. Wagner, R.K. & Sternberg, R.J. (1985). Practical intelligence in real world pursuits: The role of tacit knowledge. Journal of Personality and Social Psychology, 49 436 458. Ware, J.E., & Sherbourne, C. D. (1992). The MOS 36 item short form health survey (SF 36): I. Conceptual framework and item selection. Medical Care, 30, 473 483. Wechsler, D. (1981). Manual for the Wechsler Ad ult Intelligence Scale Revised. New York: The Psychological Corporation. West, R.L., Crook, T.H., & Barron, K.L.(1992). Everyday memory performance across the life span: Effects of age and noncognitive individual differences. Psychology and Aging, 7 72 82 Whitfield, K.E., Baker Thomas, T., Heyward, K., Gatto, M., & Williams, Y. (1999). Evaluating a measure of everyday problem solving for use in African Americans. Experimental Aging Research, 25 209 221. Willis, S.L. (1991). Cognition and everyday compet ence. In: K.W. Schaie (Ed.). Annual review of gerontology and geriatrics (Volume 11, pp. 80 109) New York: Springer. Willis S.L. (1996a) Everyday cognitive competence in elderly persons: Conceptual issues and empirical findings. Gerontologist, 51 11 17. Willis, S.L. (1996b). Everyday problem solving. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (4th edition, pp. 287 307). New York: Academic Press. Willis, S.L., Jay, G.M., Diehl, M., & Marsiske, M. (1992). Longitudinal change an d prediction of everyday task competence in the elderly. Research on Aging, 14 68 91. Willis, S. L., & Marsiske, M. (1991). Life span perspective on practical intelligence In D. E. Tupper & K. D. Cicerone (Eds.). The neuropsychology of everyday life: Issues in development and rehabilitation (pp. 183 198). Boston: Kluwer. Willis, S.L. & Marsiske, M. (1993). Manual for the Everyday Problems Test. University Park, PA: Pennsylvania State University.

PAGE 102

102 Willis S.L. & Schaie K.W. (1986). Practical intelligence in later adulthood. In: R.J. Sternberg, R.K. Wagner (Eds). Practical intelligence: Nature and origins of competence in the everyday world (pp. 236 268) New York: Cambridge University Press. Willis, S.L. & Schaie, K.W. (1993). Everyday cognition: Taxonomic and methodological considerations. In: Puckett JM, Reese HW, (Eds.). Mechanisms of everyday cognition (pp. 33 54). Hillsdale, NJ: Erlbaum. Willis, S.L., Tennstedt, S.L., Marsiske, M., Ball, Wright, E. (2006). Long term effects of cognitive training on everyday functional outcomes in older adults. Journal of the American Medical Association 296 2805 2814. Wilson B. A ., Cockburn, J., Baddeley, A. (1985) The Rivermead Behavioral Memory Test. Reading, England: Thames Valley Test Co.; Gaylord, MI: National Rehabilitation Services. Peterse n, R. C. (2004). Mild cognitive impairment Beyond controversies, towards a consensus: Report of the International Working Group on Mild Cognitive Impairment. Journal of Internal Medicine, 256 240 246. Wolinsky, F.D., Vander Weg, M.W., Martin, R., Unverza gt, F.W., Willis, S.L., Marsiske, control among older adults? Journal of Gerontology: Social Sciences, 65B(5) 591 598. Wood K. M., Edwards, J. D., Clay, O. J., Wadley, V. G., Roenker, D. L., & Ball, K. K. (2005). Sensory and cognitive factors influencing functional ability in older adults. Gerontology 51 (2), 131. Yaffe K, Petersen RC, Lindquist K, Kramer J, & Miller B. (2006). Subtype of mild cognitive impairment and progressi on to dementia and death. Dementia and Geriatric Cognitive Disorders, 22 312 319.

PAGE 103

103 BIOGRAPHICAL SKETCH Kelsey Thomas graduated from the University of California at San Diego in 2008 Following in San Diego on a study investigating a compensatory cognitive training pro gram and supported employment for Iraq/Afghanistan veterans with traumatic brain in juries. Ms. Thomas was accepted into the doctoral program in Clinical and Health Psychology at the University of Florida in 2010. She is currently there working toward her doctorate in clinical and health psychology, with a specialization in clinical neur opsychology. Her focus and research interests involve conducting applied neuropsychological research, with particular emphasis in cognitive training and rehabilitation interventions that can be easily implemented to attenuate or delay the cognitive declin e and subsequent functional decline associated with both normal and pathological aging.