Search tips
Search criteria 


Logo of acnLink to Publisher's site
Arch Clin Neuropsychol. 2010 February; 25(1): 49–59.
Published online 2009 November 25. doi:  10.1093/arclin/acp091
PMCID: PMC2809552

Cranial Volume, Mild Cognitive Deficits, and Functional Limitations Associated with Diabetes in a Community Sample


Diabetes is associated with dementia in older adults, but it remains unclear whether nondemented adults with type 2 diabetes show subtle abnormalities across cognition, neuroanatomy, and everyday functioning. Using the Aging, Brain Imaging, and Cognition study sample of 301 community-dwelling, middle-aged and older adults, we conducted a secondary analysis on 28 participants with and 150 participants without diabetes. We analyzed brain magnetic resonance imaging data, cognitive test performance, and informant ratings of personal and instrumental activities of daily living (PADL/IADL). Relative to controls, participants with diabetes had lower brain-to-intracranial volume ratios (69.3 ± 4.5% vs. 71.7 ± 4.6%; p < .02), and performed more poorly on measures of working memory, processing speed, fluency, and crystallized intelligence (all p <.05). Decrements in working memory and processing speed were associated with IADL limitations (p < .01). Nondemented adults with diabetes exhibit neuroanatomic and cognitive abnormalities. Their cognitive deficits correlate with everyday functional limitations.

Keywords: Diabetes, Endocrine disorders, Cognition, Neuropsychological testing, MRI, Function, Behavior


Diabetes is associated with higher prevalence of dementia in older adults (Ott et al., 1999; Whitmer, 2007). The increased risk for both Alzheimer's disease and vascular dementia has been found across population subgroups with diabetes, and may be higher in racial/ethnic minorities with diabetes (Luchsinger, Tang, Stern, Shea, & Mayeux, 2001). A large body of literature elucidates vascular, biochemical, and metabolic mechanisms linking diabetes with increased dementia risk (Biessels, Staekenborg, Brunner, Brayne, & Scheltens, 2006). However, it remains unclear whether nondemented adults with type 2 diabetes show abnormalities across cognition, neuroanatomy, and everyday functioning that precede the manifestation of dementia.

Despite the identification of neuropathological pathways and mechanisms, relatively little research has addressed the need for comprehensive investigation of neuropsychological and behavioral functioning in this population (van den Berg, Kessels, de Haan, Kappelle, & Biessels, 2005; Whitmer, 2007). Memory and executive function have been studied most frequently, and have been shown to be areas of decrement in older adults with diabetes (Biessels et al., 2006; Whitmer, 2007). The two existing major published reviews of cognition studies in diabetes (Stewart & Liolitsa, 1999; Strachan, Dreary, Ewing, & Frier, 1997a) have highlighted several limitations in cognitive assessment methods, including widespread differences in measure selection, frequent use of only screening examinations, quality of instruments, and limited domains examined. Moreover, previous studies generally have not examined cognitive performance decrements in a clinically interpretive manner. That is, although studies may report findings regarding statistically significant differences in the cognitive scores of the diabetic and nondiabetic participants, whether these decrements reach a level of clinical impairment relative to the nondiabetic controls is seldom explored. Despite a statement of need for consensus in cognitive testing in diabetes (Strachan, Dreary, Ewing & Frier, 1997b), the field has not as yet addressed these limitations systematically.

Neuroimaging studies in diabetes have included numerous magnetic resonance imaging (MRI) studies, mainly investigating white matter lesions (WMLs), lacunar infarcts, and cerebral atrophy. The studies have shown a relationship between diabetes and increased cerebral atrophy and more lacunar infarcts, but have not shown any consistent relationship with WMLs (van Harten, de Leeuw, Weinstein, Scheltens, & Biessels, 2006). Importantly, very few studies to date have investigated both MRI findings and neuropsychological functioning. None have presented detailed analyses on the association of MRI findings with neuropsychological performance in participants with diabetes (van Harten et al., 2006).

Finally, studies of personal activities of daily living (PADL) and instrumental activities of daily living (IADL) in diabetes are mainly limited to studies of severely disabled individuals. In a study of homebound older people with diabetes, executive function deficits were found to be associated with impaired PADL performance (Qui et al., 2006). However, few studies have studied PADL and IADL functioning in a normal aging community sample, where it would be expected that PADL function would be intact. In a study of geriatric clinic patients, IADL performance was impaired in persons with diabetes when compared with persons without diabetes, but PADL performance was intact (Munshi et al., 2006).

Further research clarifying this overall subclinical pattern with respect to brain imaging parameters, neuropsychological performance across a comprehensive range of domains, and functional behaviors (e.g., PADL and IADL) is needed. Unique to our study is the analyses of the association between neuropsychological test performance, MRI, and functional independence (PADL/IADL).

The purpose of this study was to examine, in a community aging sample without clinically diagnosed dementia, whether persons with diabetes, compared with those without diabetes, exhibit (1) abnormalities on brain MRI and neuropsychological testing, and (2) patterns of association among MRI parameters, neuropsychological performance, and PADL and IADL functional behaviors.

Materials and Methods

Setting and Participants

Data for the current study were drawn from the Aging, Brain Imaging, and Cognition (ABC) Study, a National Institutes of Health (NIH)-funded study of normal aging. Participants were recruited from Baltimore, MD, through random digit dialing or written invitation sent to Medicare beneficiaries of age 65 and older. Altogether, 301 participants enrolled in the ABC study. Fig. 1 shows participant selection for inclusion in this study. From these 301 participants, diabetes cases were identified by self-report and confirmed or disconfirmed through medical record review and physician-documented notes. Any person who did not report diabetes, but whose medical record indicated diabetes diagnosis or current treatment with diabetes medications (e.g., insulin, oral anti-diabetic agents) was classified as diabetic. Twenty-eight persons (22 in the nondiabetic group and 6 in the diabetic group) were excluded due to unconfirmed diagnosis of diabetes in the medical record or possible undiagnosed diabetes (i.e., no self-reported diabetes and no diabetes diagnosis in the medical chart, but presence of hyperglycemia on laboratory testing for the study). We then included only those nondiabetic participants within the age range of the diabetic participants, yielding 32 persons with diabetes and 175 without diabetes, a 15.4% prevalence rate, which is consistent with an identified diabetes prevalence rate of 15% among Medicare beneficiaries in Maryland (Merrill & Limpa-Amara, 2005). Further exclusions were made for probable severe cognitive impairment or dementia, based on screening Mini-Mental State Exam (MMSE) < 24) or documentation of dementia in the medical record (n = 3 in the diabetic group and n = 9 in the nondiabetic group); and the absence of MRI data (n = 1 in the diabetic group and 16 in the nondiabetic group). A total of 150 nondiabetic and 28 diabetic participants comprised the final study sample. All participants had a blood glucose value of >65 mg/dl at the time of neuropsychological testing, ensuring that cognitive performance did not reflect transient changes in cognition due to hypoglycemia (Holmes, Hayford, Gonzalez, & Weydert, 1983).

Fig. 1.
Participant selection.


The ABC Study participants attended a 1-day study visit at The Johns Hopkins Hospital. After providing informed consent, they underwent sociodemographic and medical history interview, full physical and neurological examinations, psychiatric interview, neuropsychological testing, phlebotomy for testing of blood glucose, and brain MRI, in which the presence of prior infarcts was noted by an expert rater. Neuropsychological tests were administered by trained examiners, following standard administration procedures for each test. The Johns Hopkins School of Medicine Institutional Review Board approved the study. All participants gave written, informed consent.

MRI Protocol

All participants underwent brain MRI on the same 1.5 T GE Signa scanner (Milwaukee, WI, USA). We acquired 124 contiguous 1.5 mm 3-D SPGR slices in the coronal plane. Parameters were: Repetition time = 35 s, echo time = 5 s, flip angle = 45°, image matrix = 256 × 256. Images were preprocessed with statistical parametric mapping software (SPM2, Wellcome Department of Cognitive Neurology, London, UK), using the Good and colleagues (2001) optimized method.

Images were first transformed into standard stereotactic space and global shape differences were removed with spatial normalization using the application of linear and nonlinear transformations of the study images to a customized template based on the T1-weighted images of 86 relatively healthy normal control subjects who took part in a study of normal aging and went through a screening process identical to that of the current study participants (mean age = 50.93 years, 41.9% men). Prior probability maps estimated the likelihood of a given tissue type at each voxel and, based on this information, the images were segmented into volumes of grey matter, white matter, and cerebrospinal fluid (CSF).

Magnetic resonance imaging data were displayed and white matter hyperintensities (WMH) were measured using image analysis software developed in our laboratory, ‘Measure’ version 0.7 (Barta, Dhingra, Royall, & Schwartz, 1997), which allows for outlining WMH regions of interest on each axial slice using a mouse-controlled cursor. After all WMH observed on T2 images throughout the brain were traced, their total volume was calculated via the Measure program. Blinded to participant characteristics and using a region-of-interest approach, all observable WMH were manually traced by a single rater who was trained by a neuroradiologist. In order to be considered an area of hyperintensity, the region of interest must have appeared hyperintense on both T2 and proton density-weighted images (Schretlen et al., 2007; Vannorsdall, Waldstein, Kraut, Pearlson, & Schretlen, 2009). Total cerebral volume was calculated as the sum of grey and white matter. Total intracranial volume was calculated as the sum of grey matter, white matter, and CSF. Brain-to-intracranial volume ratio was defined as the ratio of total cerebral volume to total intracranial volume times 100.

Neuropsychological Tests

The MMSE (Folstein, Folstein, & McHugh, 1975) was used as a screening instrument. A battery of 28 additional neuropsychological tests was administered. The 28 tests were grouped into eight domains of cognitive function according to the classification outlined by Lezak, Howieson, and Loring (2004). Cronbach's alpha was calculated to determine internal reliability of each of the eight domains.

Working memory

The working memory domain included the Digit Span subtest of the Wechsler Adult Intelligence Scale (WAIS-R) (Wechsler, 1981), the Brief Test of Attention, letters and numbers scales (BTA; Schretlen, 1997), and yielded a Chronbach's alpha of 0.73.

Processing speed

The processing speed domain comprised of the Digit Symbol subtest of the WAIS-R (Wechsler, 1981), the Speed subtest of the Salthouse Perceptual Comparison Test (Salthouse, 1991), and the Trail Making Test Parts A and B (Reitan, 1958). Cronbach's alpha for the processing speed domain was 0.90.


The fluency domain had a Chronbach's alpha of 0.74 and included letter fluency (Lezak et al., 2004), category fluency (Lezak et al., 2004), and design fluency (Jones-Gotman & Milner, 1977).

Crystallized intelligence

The crystallized intelligence domain was comprised of the Information and Similarities subtests of the WAIS-R (Wechsler, 1981) and the National Adult Reading Test IQ portion (Blair & Spreen, 1989). Cronbach's alpha for the crystallized intelligence domain was 0.86.

Fluid intelligence

The fluid intelligence domain included the Rey–Osterrieth Complex Figure Test Copy Integrity (Rey, 1941; Rey & Osterrieth, 1993), the Benton Facial Recognition Test (Benton & Van Allen, 1969), and the Block Design and Picture Completion subtests of the WAIS-R (Wechsler, 1981), and had a Chronbach's alpha of 0.76.

Verbal memory

The verbal memory domain included the Hopkins Verbal Learning Test, revised, subtests of Learning Trials (1–3) and Delayed Recall Trial (Brandt & Benedict, 2001) and the Logical Memory I and II subtests of the WMS-R (Wechsler, 1987). Cronbach's alpha for this domain was 0.84.

Visual memory

The visual memory domain had a Chronbach's alpha of 0.89 and was comprised of the Brief Visuospatial Memory Test, revised, subtests of Learning Trials (1–3) and Delayed Recall Trial (Benedict, 1997) and the Visual Reproduction I and II subtests of the WMS-R (Wechsler, 1987).

Executive functioning

The executive function domain was comprised of the categories and perseverative errors scores of the Modified Wisconsin Card Sorting Test (mWCST) (Heaton, 1981; Nelson, 1976), and yielded a Cronbach's alpha of 0.85.


Ratings of PADLs and IADLs were obtained using the Lawton PADL/IADL Questionnaire Form R (Lawton & Brody, 1969), a commonly used instrument for assessing functional tasks in research and clinical practice. The questionnaire assesses the following PADLs: Eating, dressing, grooming, ambulating, getting into/out of bed, bathing, and frequency of incontinence. The questionnaire assesses the following IADLs: Using a telephone, traveling outside of walking distance, shopping, cooking, doing housework, handyman work, and laundry, taking medications, and managing money. Participants and their informants were asked to grade an individual's ability to perform each task independently on a 0–2 scale, with 0 indicating no assistance needed, 1 indicating some assistance needed, and 2 indicating complete inability to do task without assistance. Raw scores were summed. The scale has demonstrated acceptable psychometric properties across a range of studies in middle-to-older-aged populations (Mangen & Peterson, 1984).

Data Analysis

For the diabetic and nondiabetic participants, raw scores for each cognitive test were z-transformed in order to standardize the scores for the tests comprising each domain. Means and standard deviations of the entire sample (n = 178) were used to compute the z-score transformations. Each participant's domain score consisted of the sum of the z-transformed scores on each test in the domain. Z-scores for timed measures were inverse coded as needed so that for all domains, higher scores indicate better performance. An independent samples t-test compared diabetic with nondiabetic participants on MMSE performance and WMH burden. Chi-squares assessed group differences in sex, race, and prevalence of stroke. Analyses of covariance (ANCOVA), adjusting for age, sex, race, and education, were used to assess group differences in each cognitive domain and on MRI measures. A series of hierarchical multiple regressions was used to test the hypothesis that MRI measures would explain significant variance in both cognitive functioning and PADL independence after accounting for age, education, race, and sex (entered first en bloc). IADL discrepancy score was calculated as self-rating minus informant rating and was used in a series of hierarchical multiple regressions to test the hypothesis that IADL discrepancy would predict decrement in cognition and cerebral reserve. Finally, for each cognitive test score, we determined the percentage of diabetic participants scoring one or more standard deviations below the mean ± standard deviations of the nondiabetic control group. This yielded the percentage of diabetic participants performing an impaired range relative to the control group. All data were analyzed using SPSS 15.0 (SPSS, Chicago, IL, USA).


Sample Characteristics

The diabetic group consisted of 25 participants with type 2 diabetes, two participants with type 1 diabetes, and 1 participant with diabetes type unspecified. Participant characteristics, stratified by diabetes status, are given in Table 1. The groups did not differ in age or sex; however, the diabetic group reported fewer years of education when compared with the nondiabetic group. Consistent with national trends, African Americans comprised a larger proportion of the diabetic group. For each participant, blood glucose at the time of testing was >65 mg/dl and did not fall within a range of values associated with acute, transient hypoglycemia-related cognitive impairment (Holmes, Hayford, Gonzalez, & Weydert, 1983).

Table 1.
Demographic and clinical characteristics of a community sample of adults without and with diabetesa

MRI and Neuropsychological Test Findings in Diabetic Participants

On MRI, this sample of nondemented diabetic participants had significantly lower brain-to-intracranial volume ratios than the nondiabetic participants (Table 1). Volumes of gray matter, white matter, CSF, and WMH did not differ between groups (all p > .05).

MMSE scores for the nondiabetic group (mean = 28.17, SD = 1.43) and the diabetic group (mean = 27.79, SD = 1.69) did not differ (t = 1.26, p = .21). With regard to neuropsychological test performance, in adjusted analyses, diabetic participants performed significantly more poorly than nondiabetic participants in the domains of working memory, processing speed, fluency, and crystallized intelligence (Table 2). Fluid intelligence, verbal memory, visual memory, and executive function scores did not reveal statistically significant differences between the diabetic and nondiabetic participants. Lower brain-to-intracranial volume ratio in diabetic participants was not associated with the neuropsychological test results (all p > .05).

Table 2.
Cognitive domain z-scores and mean (SD) individual neuropsychological test scores in a community sample of adults without and with diabetesa

For each neuropsychological test score, percentage of diabetic participants scoring in an impairment range relative to the nondiabetic control group was determined (Table 3). Between 11% and 46% of diabetic participants scored at least one standard deviation below the mean of the nondiabetic control group on each of the 28 neuropsychological tests. Notably, 4%–11% of diabetic participants performed three or more standard deviations below the mean of the nondiabetic participants on Brief Test of Attention numbers, Trail Making Test parts A and B, Similarities, Benton Facial Recognition, Picture Completion, and the mWCST, categories and perseverative errors.

Table 3.
Percentage of diabetic participants exhibiting test score decrements relative to the nondiabetic control group (raw scores)

Association of Cognition with PADLs and IADLs in Diabetic Participants

Self-ratings of PADL and IADL performance did not differ between diabetic and non-diabetic participants. In diabetic participants, self-ratings of PADL behaviors did not differ from informant-ratings of PADLs (p = .16 self and p = .36 rater), and no associations were found between PADLs and either brain-to-intracranial volume ratio or neuropsychological performance in any domain (all p > .05).

With regard to IADL behaviors, informants rated diabetic participants as needing a higher level of assistance, whereas the diabetic participants rated themselves as needing a lower level of assistance (discrepancy score mean = −0.42, SD = 1.76; β = 0.67, p = .001). Discrepancy between the IADL ratings of diabetic participants and their informants increased with greater decrements in the diabetic participants' working memory (β = 0.68, p = .001) and fluid intelligence (β = 0.50, p = .01).

In the diabetic group, MRI findings of decreased brain-to-intracranial volume ratio were not associated with informant-reported need for assistance performing IADLs (β = −0.23, p = .26). However, informant report of diabetic participants' ability to perform IADLs was associated with decrements in working memory (β = −0.60, p = .002) and processing speed (β = −0.53, p = .01). The association between need for IADL assistance and fluid intelligence approached significance (β = −0.54, p = .05).

In the non-diabetic participants, no associations were found between cognition, MRI, and functional activities.


This study was designed to examine subtle abnormalities of cognition, neuroanatomy, and everyday functioning in persons with diabetes. In this community sample without dementia, persons with diabetes, relative to nondiabetic controls, showed evidence of decrements in working memory, processing speed, fluency, and crystallized intelligence and lower brain-to-intracranial volume ratio on MRI. Working memory and processing speed were associated with a greater need for assistance performing IADLs.

The present findings are consistent with other studies that report cognitive decrements in the domains of working memory and processing speed in adults with diabetes (Akisaki et al., 2006; Manschot et al., 2006; van Harten et al., 2007). Unlike previous studies, which have typically found an executive function decrement in diabetic participants (Manschot et al., 2006; Schillerstrom, Horton, & Royall, 2005; van Harten et al., 2007), the current study did not. This is likely an artifact of test selection and cognitive domain grouping. Prior studies have used the Trail Making Test Part B, Stroop Color-Word Test part III, or tests of verbal fluency to assess executive function (Akisaki et al., 2006; Schillerstrom et al., 2005; van Harten et al., 2007), whereas our executive function domain consisted of the mWCST. The grouping of these tests into cognitive domains is not orthogonal and there is an overlap with respect to the cognitive skills used in these different domains. We did find that the diabetic group performed more poorly on several measures thought to be reliant on frontal lobe functioning, such as tests of letter-cued verbal fluency, design fluency, Digit Span backwards, and the Trail Making Test Part B. Thus, although we did not find a decrement in our diabetic group on our executive function test per se, our findings for other tests that have been conceptualized as executive function measures do show the expected associations.

When we reported percentage of diabetic participants with test scores ≥ 1 SD below the mean of the nondiabetic control group (Table 3), a larger magnitude of cognitive impairment was detected. A pattern of impairment was found for each test score, with up to 43% of diabetic participants showing impairment on some tests. Previous studies to date have limited their reporting of comparisons between diabetic and nondiabetic participants to group-level analyses, which do not provide such clinical interpretation. And, it has been suggested that with group-level analyses, a 0.5 SD difference between diabetic and nondiabetic participants would be indicative of a medium effect size with regard to cognitive decrement (Strachan et al., 1997b). However, based on the methodology of comparing between-group means, it has been concluded, perhaps prematurely, that there are only a few decrements in cognition in nondemented diabetic patients, and that they are not of a magnitude that would be expected to have clinical impact (Biessels, Kerssen, de Haan, & Kappelle, 2007). Our finding of diabetic participants scoring ≥ 3 SD below our normal controls suggests evidence to the contrary.

Personal activities of daily living and IADL function have generally not been studied in populations with diabetes who are free from dementia. However, consistent with our findings, a study by Munshi and colleagues (2006) found relatively preserved PADL function and a greater need for assistance in performing IADLs in participants with diabetes. In our sample, it was expected that no decrements in PADL functioning would be found due to the community dwelling status of our participants. To our knowledge, no other studies have investigated an association between decreased brain-to-intracranial volume ratio and PADLs or IADLs in a sample of patients with diabetes. The lower brain-to-intracranial volume ratio observed in our diabetic group was not associated with cognitive decrements, PADLs, or IADLs.

Studies by Akisaki and colleagues (2006) and Manschot and colleagues (2006) have described an association between cerebral atrophy and cognitive decrement. But, in those samples, the diabetic participants exhibited increased cerebral infarcts and WMHs in addition to decreased brain-to-intracranial volume ratio, when compared with the nondiabetic participants. The investigators have, therefore, explained the findings as vascular in origin (Akisaki et al., 2006; Manschot et al., 2006). In contrast, in the present study, our diabetic and nondiabetic participants did not differ in history of stroke or WMH burden, indicating absence of advanced diabetes-related vascular pathology.

There is consensus that diabetes is associated with increased risk for both Alzheimer's disease and vascular dementia (Biessels et al., 2006; Whitmer 2007). The pathways are not clear because of the multifactoral nature of biochemical and metabolic etiologies (e.g., hyperinsulinemia, glucose toxicity leading to advanced glycation end-products, and oxidative stress, hypoglycemia, neurologic risk factors associated with the metabolic syndrome), in addition to the vascular mechanisms (Biessels et al., 2006; Whitmer 2007). Vascular mechanisms implicated in cognitive impairment in diabetes include atherosclerosis leading to stroke and microvascular disease leading to ischemia (Whitmer, 2007). Moreover, there appear to be specific vulnerabilities to diabetes-related cognitive dysfunction: (a) during the period of childhood brain development; (b) during the period of older adulthood, when the brain undergoes neurodegenerative changes associated with aging; and (c) in the setting of notable microvascular or macrovascular complications of diabetes (Taguchi, 2009; Biessels, Deary, & Ryan, 2008). Consequently, as intended in our exclusion of participants with known dementia, and as demonstrated by absence of increased stroke or WMH burden in our diabetic participants, our sample likely represents older persons with diabetes who have not yet experienced the extensive neurological damage that results from advanced diabetes-related vascular pathology.

Limitations of the study include small sample size, which may have limited our statistical power to detect between-group differences in test scores shown in Table 2, as a non-statistically significant trend toward lower scores was observed in the diabetic group across those domains that were not statistically different between groups. Strachan and colleagues (1997a) have, in fact, postulated that population-based samples may commonly be underpowered to detect between-group differences in cognitive test performance due to the mild decrements associated with diabetes. Our inclusion of Table 3, however, addresses the question of performance decrements in the diabetic participants, when compared with our nondiabetic controls, in a more clinically interpretive manner, by reporting percentage of diabetic participants ≥ 1 SD below the control group mean for each test score. Here, clear performance decrements are apparent even in domains that were perhaps not powered to detect differences between diabetic participants and controls at the group level of analysis. In our analyses, we co-varied for education and considered co-varying for IQ; however, in doing so we would extract all of the variance shared between intelligence and the domains of working memory, fluency, and processing speed. As a result, our models would be over corrected. Further, such an approach would treat verbal abstract abilities (Similarities) and general fund of knowledge (Information) as error variance, when these may in fact be key cognitive skills that contribute to the successful management of a chronic disease such as diabetes. Given this, we elected to retain crystallized intelligence as a dependent variable in our analyses.

The diabetic sample included two participants with type 1 diabetes, and one participant for whom type of diabetes was not specified. We included these three participants in our analyses rather than excluding them, because Brands and colleagues (2007) have shown patients with type 1 and type 2 diabetes have similar cognitive profiles and neuroanatomic changes. Finally, we were not able to examine the contribution of various diabetes-specific or glycemic control variables to the cognitive decrements found in this study, as the parent study was not specifically designed to investigate diabetes. Importantly, we did have blood glucose value at the time of testing in order to ensure that the test results did not reflect transient alterations in cognition due to hypoglycemia.

The study has important strengths. First, much of the research examining cognitive functioning in nondemented persons with type 2 diabetes has relied on screening instruments or a few selected tests. Here, we used a comprehensive cognitive test battery, which eliminates confirmation bias by examining the range of cognitive domains, not just those commonly thought to be impaired in diabetes (e.g., memory, executive function). Second, persons with dementia were excluded in order to detect the influence of early cognitive decline on functioning prior to clinical manifestation. This subclinical stage is identified as a key period for screening and early intervention (Belleville, 2008; Petersen & Negash, 2008). Finally, the inclusion of PADL and IADL measures facilitates greater appreciation for the significance of these cognitive decrements in the lives of adults with diabetes.

Our finding that diabetic persons with cognitive decrements rated themselves as needing significantly less assistance with IADLs than did their caregivers has clinical implications. IADLs are defined as the ability to plan, initiate, and properly execute cognitively complex tasks of daily living. Individuals with IADL limitations and cognitive dysfunction exhibit increased declines in physical function over time (Wang, van Belle, Kukull & Larson, 2002). Importantly, diabetes management involves a complex combination of diet, regular exercise and physical activity, blood glucose self-monitoring, oral medication, and/or insulin administration. Therefore, both decreased cognition and poor awareness of need for assistance with IADLs among diabetic persons with decreased cognition may have serious implications for the ability of diabetic patients to manage their diabetes effectively.

Further research is needed to determine whether the observed cognitive, functional, and neurological pattern in diabetes is a transitional period to dementia seen in the general population or whether the observed pattern of neurological change, cognitive decline, and IADL decrement represents a subclinical state unique to diabetes that may or may not follow the standard accelerated pattern of conversion to dementia (Luchsinger et al., 2007; Petersen & Negash, 2008).


This research was funded by a grant from NIMH (R01MH060504). FHB is supported by the NIDDK Diabetes Research and Training Center (P60 DK079637).

Conflict of Interest

Dr. Schretlen is entitled to a share of royalty on sales of a test used in the study described in this article. The terms of this arrangement are being managed by the Johns Hopkins University in accordance with its conflict of interest policies.


These data were presented in part as an oral abstract at the American Diabetes Association's 68th Scientific Sessions, June 2008.


  • Akisaki T., Sakurai T., Takata T., Umegaki H., Araki A., Mizuno S., et al. Cognitive dysfunction associates with white matter hyperintensities and subcortical atrophy on magnetic resonance imaging of the elderly diabetes mellitus Japanese elderly diabetes intervention trial (J-EDIT) Diabetes Metabolism Research and Reviews. 2006;22:376–384. [PubMed]
  • Barta P. E., Dhingra L., Royall R., Schwartz E. Improving stereological estimates for the volume of structures identified in three-dimensional arrays of spatial data. Journal of Neuroscience Methods. 1997;75(2):111–118. [PubMed]
  • Belleville S. Cognitive training for persons with mild cognitive impairment. International Psychogeriatrics. 2008;20:57–66. [PubMed]
  • Benedict H. R. B. Brief visuospatial memory test—revised professional manual. Odessa, FL: Psychological Assessment Resources, Inc; 1997.
  • Benton A. L., Van Allen M. W. Impairment in facial recognition in patients with cerebral disease. Cortex. 1969;4:344–358.
  • Biessels G. J., Deary I. J., Ryan C. M. Cognition and diabetes: a lifespan perspective. Lancet Neurology. 2008;7(2):184–190. [PubMed]
  • Biessels G. J., Kerssen A., de Haan E. H., Kappelle L. J. Cognitive dysfunction and diabetes: implications for primary care. Primary Care Diabetes. 2007;1:187–193. [PubMed]
  • Biessels G. J., Staekenborg S., Brunner E., Brayne C., Scheltens P. Risk of dementia in diabetes mellitus: a systematic review. Lancet Neurology. 2006;5:64–74. [PubMed]
  • Blair J. R., Spreen O. Predicting premorbid IQ: A revision of the national adult reading test. Clinical Neuropsychologist. 1989;3:129–136.
  • Brands A. M., Biessels G. J., Kappelle L. J., de Haan E. H., de Valk H. W., Algra A., et al. Cognitive functioning and brain MRI in patients with type 1 and type 2 diabetes mellitus: A comparative study. Dementia and Geriatric Cognitive Disorders. 2007;23:343–350. [PubMed]
  • Brandt J., Benedict H. R. B. Hopkins verbal learning test—revised professional manual. Odessa, FL: Psychological Assessment Resources, Inc; 2001.
  • Folstein M. F., Folstein S. E., McHugh P. R. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research. 1975;12:189–198. [PubMed]
  • Good C. D., Johnsrude I. S., Ashburner J., Henson R. N., Friston K. J., Frackowiak R. S. A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage. 2001;14:21–36. [PubMed]
  • Heaton R. K. Wisconsin Card Sorting Test manual. Odessa, FL: Psychological Assessment Resources, Inc; 1981.
  • Holmes C. S., Hayford J. T., Gonzalez J. L., Weydert J. A. A survey of cognitive functioning at difference glucose levels in diabetic persons. Diabetes Care. 1983;6(2):180–185. [PubMed]
  • Jones-Gotman M., Milner B. Design fluency: the invention of nonsense drawings after focal cortical lesions. Neuropsychologia. 1977;15:653–674. [PubMed]
  • Lawton M. P., Brody E. M. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist. 1969;9:179–186. [PubMed]
  • Lezak M. D., Howieson D., Loring D. Neuropsychological assessment. 4th ed. New York: Oxford University Press; 2004.
  • Luchsinger J. A., Reitz C., Patel B., Tang M. X., Manly J. J., Mayeux R. Relation of diabetes to mild cognitive impairment. Archives of Neurology. 2007;64:570–575. [PubMed]
  • Luchsinger J. A., Tang M. X., Stern Y., Shea S., Mayeux R. Diabetes mellitus and risk of Alzheimer's disease and dementia with stroke in a multiethnic cohort. American Journal of Epidemiology. 2001;154:635–641. [PubMed]
  • Mangen D. G., Peterson W. A. Research instruments in social gerontology: Vol. 3. Health, program evaluation, and demography. Minneapolis, MN: University of Minnesota Press; 1984.
  • Manschot S. M., Brands A. M., van der Grond J., Kessels R. P., Algra A., Kappelle L. J., et al. Brain magnetic resonance imaging correlates of impaired cognition in patients with type 2 diabetes. Diabetes. 2006;55:1106–1113. [PubMed]
  • Merrill A., Limpa-Amara S. S. Washington, DC: Mathematica Policy Research, Inc; 2005. Diabetes prevalence, outcomes, and preventive services among Maryland Medicare beneficiaries (MPR Reference No. 6087–300)
  • Munshi M., Grande L., Hayes M., Ayres D., Suhl E., Capelson R., et al. Cognitive dysfunction is associated with poor diabetes control in older adults. Diabetes Care. 2006;29:1794–1799. [PMC free article] [PubMed]
  • Nelson H. E. A modified card sorting test sensitive to frontal lobe defects. Cortex. 1976;12:313–324. [PubMed]
  • Ott A., Stolk R. P., van Harskamp F., Pols H. A., Hofman A., Breteler M. M. Diabetes mellitus and the risk of dementia: The Rotterdam Study. Neurology. 1999;53:1937–1942. [PubMed]
  • Petersen R. C., Negash S. Mild cognitive impairment: an overview. CNS Spectrums. 2008;13:45–53. [PubMed]
  • Qui W. Q., Price L. L., Hibberd P., Buell J., Collins L., Leins D., et al. Executive dysfunction in homebound older people with diabetes mellitus. Journal of the American Geriatrics Society. 2006;54(3):496–501. [PubMed]
  • Reitan R. M. Validity of the trail making test as an indicator of organic brain damage. Perceptual and Motor Skills. 1958;8:271–276.
  • Rey A. L'examen psychologique dans les cas d'encephalopathie traumatique. (Les problems.) Archives de Psychologie. 1941;28:215–285.
  • Rey A., Osterrieth P. A. Translations of excerpts from Andre Rey's psychological examination of traumatic encephalopathy and P.A. Osterrieth's The Complex Figure Copy Test. Clinical Neuropsychologist. 1993;71:4–21.
  • Salthouse T. A. Mediation of adult age differences in cognition by reductions in working memory and speed of processing. Psychological Science. 1991;2:179–183.
  • Schillerstrom J. E., Horton M. S., Royall D. R. The impact of medical illness on executive function. Psychosomatics. 2005;46:508–516. [PubMed]
  • Schretlen D. J. Brief Test of Attention – Professional Manual. Odessa, FL: Psychological Assessment Resources, Inc; 1997.
  • Schretlen D. J., Inscore A. B., Vannorsdall T. D., Kraut M., Pearlson G. D., Gordon B., et al. Serum uric acid and brain ischemia in normal elderly adults. Neurology. 2007;69:1418–1423. [PubMed]
  • Stewart R., Liolitsa D. Type 2 diabetes mellitus, cognitive impairment and dementia. Diabetic Medicine. 1999;16:93–112. [PubMed]
  • Strachan M. W., Dreary I. J., Ewing F. M., Frier B. M. Is type II diabetes associated with an increased risk fo cognitive dysfunction? A critical review of published studies. Diabetes Care. 1997a;20:438–445. [PubMed]
  • Strachan M. W., Dreary I. J., Ewing F. M., Frier B. M. Cognitive assessment in diabetes: the need for consensus. Diabetic Medicine. 1997b;14:421–422. [PubMed]
  • Taguchi A. Vascular factors in diabetes and Alzheimer's disease. Journal of Alzheimer's Disease. 2009;16(4):859–864. [PubMed]
  • Vannorsdall T. D., Waldstein S. R., Kraut M., Pearlson G. D., Schretlen D. J. White matter abnormalities and cognition in a community sample. Archives of Clinical Neuropsychology. 2009;24(3):209–217. [PMC free article] [PubMed]
  • van den Berg E., Kessels R. P., de Haan E. H., Kappelle L. J., Biessels G. J. Mild impairments in cognition in patients with type 2 diabetes mellitus: The use of the concepts MCI and CIND. Journal of Neurology, Neurosurgery, and Psychiatry. 2005;76:1466–1467. [PMC free article] [PubMed]
  • van Harten B., de Leeuw F. E., Weinstein H. C., Scheltens P., Biessels J. G. Brain imaging in patients with diabetes. Diabetes Care. 2006;29(11):2539–2548. [PubMed]
  • van Harten B., Oosterman J., Muslimovic D., van Loon B. J., Scheltens P., Weinstein H. C. Cognitive impairment and MRI correlates in the elderly patients with type 2 diabetes mellitus. Age and Ageing. 2007;36:164–170. [PubMed]
  • Wang L., van Belle G., Kukull W. B., Larson E. B. Predictors of functional change: A longitudinal study of nondemented people aged 65 and older. Journal of the American Geriatric Society. 2002;50:1525–1534. [PubMed]
  • Wechsler D. Wechsler adult intelligence scale—revised. New York: Psychological Corporation; 1981.
  • Wechsler D. Wechsler memory scale—revised. San Antonio, TX: The Psychological Corporation; 1987.
  • Whitmer R. A. Type 2 diabetes and risk of cognitive impairment and dementia. Current Neurology and Neuroscience Reports. 2007;7:373–380. [PubMed]

Articles from Archives of Clinical Neuropsychology are provided here courtesy of Oxford University Press