PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Alzheimer Dis Assoc Disord. Author manuscript; available in PMC 2011 January 1.
Published in final edited form as:
PMCID: PMC2837096
NIHMSID: NIHMS127302

Metabolic Syndrome, Inflammation, and Non-Amnestic Mild Cognitive Impairment in Older Persons: A Population-Based Study

Rosebud O. Roberts, MB ChB, MS,* Yonas E. Geda, MD, MSc,* David S. Knopman, MD, Ruth H. Cha, MS,§ Bradley F. Boeve, MD, Robert J. Ivnik, PhD, LP, V. Shane Pankratz, PhD,§ Eric G Tangalos, MD, and Ronald C. Petersen, MD, PhD*

Abstract

The metabolic syndrome (MetS) is more strongly associated with cognitive impairment in the presence of inflammation. This suggests that the association of MetS with mild cognitive impairment (MCI) may vary with the etiology and the subtype of MCI. This study investigated the association between MetS with or without inflammation and MCI (amnestic [a-MCI] and non-amnestic [na-MCI]). We studied a randomly selected sample of 1969 subjects (ages 70 to 89 years) from Olmsted County, MN, using the Clinical Dementia Rating Scale, a neurological evaluation, and neuropsychological testing. Data for participants were reviewed for a diagnosis of normal cognition, MCI, or dementia. Clinical components of MetS were ascertained by interview and confirmed from the medical records; biochemical measurements were assayed from a blood draw. We compared 88 na-MCI cases and 241 a-MCI cases with 1640 cognitively normal subjects. MetS was not associated with either na-MCI or a-MCI. High C-reactive protein (CRP highest tertile vs lowest tertile) was associated with na-MCI (odds ratio [OR] = 1.85; 95% confidence interval [CI] = 1.05, 3.24) but not with a-MCI, after adjusting for sex, age, and years of education. The combination of MetS and high CRP (compared to no Mets and lowest CRP tertile) was associated with na-MCI (OR = 2.31; 95% CI = 1.07, 5.00), but not with a-MCI (OR = 0.96; 95% CI = 0.59, 1.54). The combined presence of MetS and high levels of inflammation is associated with na-MCI in this elderly cohort, and suggests etiologic differences in MCI subtypes.

Keywords: metabolic syndrome, insulin resistance, mild cognitive impairment, C-reactive protein, inflammation, cross-sectional study

The metabolic or insulin resistance syndrome is a predictor of cardiovascular events including coronary heart disease and stroke.1,2 The components of the MetS, impaired fasting blood sugar, hypertension, abdominal obesity, hypertriglyceridemia, and low high density lipoprotein, have all been associated with dementia in several studies.3-5 The MetS may accelerate cognitive ageing through mechanisms related to micro and macrovascular damage, insulin resistance, inflammation, and adiposity.6 Consistent with this hypothesis, the MetS has been associated with Alzheimer's disease,5 vascular dementia,7 and with cognitive decline.8,9 However, other investigators have reported no associations of the MetS with cognition.10,11

The inconsistency in study findings may relate to differences in assessment of cognitive impairment, criteria used for cognitive impairment, etiologic differences in MCI subtypes, or age at which MetS is assessed. To our knowledge, the association between the MetS and MCI (defined by specified criteria at the time of evaluation) has not been assessed in the population-based setting. Our previous studies have shown associations of vascular risk factors including diabetes,12 coronary heart disease,13 and stroke14 with na-MCI, but not with the amnestic subtype (a-MCI) which has been shown to have a stronger association with the Apolipoprotein ε4 allele, and with degenerative disease. Thus, we hypothesize that the MetS may be associated with the non-amnestic MCI subtype (na-MCI). This association of the MetS with na-MCI may occur through an inflammatory process with associated endothelial dysfunction. Finally, age at assessment of MetS, may influence the associations observed with MCI. In this study, we investigate the association of MetS with MCI overall and with MCI subtypes, and the potential interaction with inflammation in a population-based sample of elderly subjects who have been well characterized for MCI using published criteria at the time of evaluation.

Methods

Identification of Study Participants

The study was conducted as part of a population based study. The details of the study design and participant selection have been previously published.15 Briefly, we used the medical records linkage system16 to construct a sampling frame of all Olmsted County, Minnesota, residents aged 70 to 89 years old on October 1, 2004. Of the eligible 4392 subjects, 2719 (61.9%) agreed to participate in the study by telephone (n = 669) or in person (n = 2050), and 1673 declined participation. All participants provided informed, written consent prior to participation. Study participants underwent a blood draw; a nurse evaluation that included a risk factor assessment and administration of the Clinical Dementia Rating Scale to the participant and to an informant; a physician evaluation; and a neuropsychological assessment using 9 cognitive tests to assess impairment in 4 cognitive domains: memory, executive function, language, and visuospatial skills.15

Criteria for MCI, Dementia, and Normal Cognition

All the data for each participant were reviewed by an expert panel of physicians, neuropsychologists, and nurses who evaluated the subject, for a diagnosis of MCI, normal cognition, or dementia by consensus. A diagnosis of MCI (cases) was made according to published criteria: cognitive concern by physician, nurse, or participant; impairment in one or more of the four cognitive domains; essentially normal functional activities; and absence of dementia.17 Participants were classified as having amnestic MCI (a-MCI) if the memory domain was impaired or non-amnestic MCI (na-MCI) if there was no impairment in memory. A diagnosis of dementia was made according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition.18 Subjects who did not receive a diagnosis of MCI or dementia were considered cognitively normal (controls) in accordance with published criteria.17,19

Assessment of the Metabolic Syndrome

Self-report of a physician diagnosis of diabetes and hypertension or treatment for these conditions was ascertained by interview. Participants were requested to bring bottles of all their current medications to the evaluation and from this, pharmacologic treatments for diabetes or hypertension were assessed. The diagnoses were also confirmed in the medical records linkage system.16 Weight was measured using an electronic balance, height was measured using a stadiometer, and blood pressure was measured in duplicate using a sphygmomanometer and averaged. Blood glucose levels at baseline were determined at the time of the blood draw after an overnight fast (for subjects evaluated in the morning) and after a 4-hour fast for subjects seen in the afternoon. Samples were aliquoted and stored at -70°C. Serum triglycerides, high-density lipoprotein, and total cholesterol were measured from thawed frozen plasma or serum samples from the blood draw.15

MetS was defined according to criteria proposed by the National Cholesterol Education Program (NCEP) Adult Treatment panel III (ATP III),2 the International Diabetes Federation,20,21 and the American Heart Association.22 The criteria include presence of 3 or more of the following conditions: 1) use of anti-diabetic agents (oral agents or insulin) or impaired fasting glucose ≥ 110 mg/dl;2,22 2) treatment for hypertension, systolic blood pressure ≥ 130 mmHg, or diastolic blood pressure ≥ 85 mmHg; 3) serum triglycerides ≥ 150 mg/dl or treatment; 4) serum high density lipoprotein < 40 mg/dl for men or < 50 mg/dl for women; 5) Obesity, defined as a body mass index (BMI; weight in kilograms divided by the height in meters squared) ≥ 30 kg/m2 based on the criterion used by the World Health Organization.21,23 This measure was a substitute for abdominal adiposity, the criterion used by NCEP ATP III (waist circumference ≥ 50 inches in men and ≥ 40 inches in women)2 because abdominal adiposity was not measured at baseline. Since subjects evaluated in the afternoon only had a 4-hour fast, a total cholesterol ≥ 200 mg/dl was considered abnormal instead of triglyceride level ≥ 150 mg/dl,22 and a blood glucose level ≥ 140 mg/dl was considered impaired.24

Covariates

Date of birth, sex, educational status, marital status, and current cigarette smoking were assessed by interview.15 A history of coronary heart disease (CHD) and stroke were also assessed by interview and confirmed from the records-linkage system.16 The presence of depressive symptoms was assessed by administration of the Neuropsychiatric Inventory Questionnaire to a spouse or to an informant of the participant.25 C-reactive protein was measured from the thawed frozen samples from the baseline blood draw, and apolipoprotein (Apoe) ε2, ε3, ε4 genotyping was performed using standard methods.26 All study protocols were approved by the Mayo Clinic and Olmsted Medical Center Institutional Review Boards.

Statistical Analyses

We compared subjects with and without MCI using the non-parametric Wilcoxon rank sum test for continuous measures and a chi-squared test for categorical variables. We used logistic regression models to estimate associations (odds ratios [OR]; 95% confidence intervals [CI]) between MetS and MCI, adjusting for age in years, sex, and number of years of education. We repeated the analyses for MCI subtypes (na-MCI and a-MCI). In additional multivariable analyses we assessed potential confounding by Apoe genotype, coronary heart disease, depression, and stroke (fully adjusted model). We evaluated effect modification in stratified analyses across strata of age (70 to 79 years vs 80 to 89 years), sex, and years of education (< 12 years vs ≥ 12 years), and by explicitly incorporating interaction terms between MetS and the variable in the logistic regression models. Because of the reported interaction between MetS and inflammation in their association with cognitive decline,8 and the observed associations of vascular risk factors with na-MCI among subjects in the present study, we investigated the potential interaction between inflammation and MetS in their associations with MCI and its subtypes, with CRP as the measure of inflammation. To further explore this interaction, we categorized subjects into 6 groups based both on presence or absence of MetS and on CRP tertile. We then assessed the OR (95% CI) of MCI for 5 of the categories compared to the reference group which comprised subjects with no MetS and CRP in the lowest tertile. In regard to criteria for MetS, we also defined an impaired fasting glucose as ≥ 100 mg/dl as suggested in more recent literature and repeated the analyses using this cutpoint for impaired fasting glucose. However, the results were essentially the same and only the data using the cutpoint of ≥ 110 mg/dl are reported. We also performed analyses to assess the association of MetS with MCI, adjusting for use of statins (yes or no) to assess confounding, and stratified by use of statins to assess possible interaction (effect modification). This is because subjects with MetS are typically more likely to be receiving treatment with statins, which in turn, may reduce the burden of MetS and the impact of inflammation.27 All analyses were performed using SAS version 8.0 software (SAS Institute, Cary, NC).

Results

Of the 2050 who were evaluated in person, 1969 completed the face-to-face evaluation; 67 subjects who were diagnosed with dementia based on the evaluation and 14 who did not complete the evaluation are not included in the present study. The characteristics of the 1969 participants are described in Table 1; 51% were male, 48% were < 80 years, and 53% had more than 12 years of education. The frequency of the MetS was 48% overall and did not differ between MCI cases and controls: 46% in subjects with any MCI, and 48% in cognitively normal subjects. Age, sex, years of education, history of CHD, stroke, depression, and ApoE ε4 allele status were associated with MCI in unadjusted analyses; there was no difference in the use of statins (Table 1). The frequency of a BMI ≥ 30 kg/m2 was lower in MCI cases than in controls (P = 0.04; Table 1), but there were no significant differences in regard to the remaining 4 individual components of the MetS.

Table 1
Characteristics of study subjects at baseline

The frequency of MetS was higher among subjects with na-MCI (51.1%) than among subjects with a-MCI (44.4%), but neither differed significantly from cognitively normal subjects (48.3%; Table 2). Compared to subjects with a-MCI, subjects with na-MCI had a non-significantly higher frequency of an impaired fasting glucose (P = 0.26), elevated blood pressure (P = 0.10), and BMI ≥ 30 kg/m2 (P = 0.19). Compared to cognitively normal subjects, subjects with na-MCI had a higher frequency of an elevated blood pressure (P = 0.06), whereas subjects with a-MCI had a significantly lower frequency of a BMI ≥ 30 kg/m2 (P = 0.02). Subjects with na-MCI had higher median CRP levels than subjects with a-MCI (P = 0.02) and cognitively normal subjects (P = 0.003; Table 2).

Table 2
Frequency of MetS criteria and CRP distribution by mild cognitive impairment subtype

Subjects with MetS had higher CRP levels than subjects without MetS; this difference was statistically significant among cognitively normal subjects, but not among subjects with either MCI subtype (Figure). CRP levels were highest among subjects with na-MCI; the median CRP (25th, 75th percentile mg/L) was 2.44 (0.9, 5.89 mg/L) in na-MCI subjects with MetS and 1.98 (0.93, 4.48 mg/L) in na-MCI subjects without MetS. In contrast, the median CRP (25th, 75th percentile) levels were lower in both cognitively normal subjects with MetS (1.73 [0.8, 3 3.58] mg/L) and without MetS (1.35 [0.68, 2.93] mg/L).

Figure
Distribution of C-reactive protein levels in subjects with and without the metabolic syndrome (MetS) in cognitively normal subjects, and in subjects with amnestic (a-MCI) and non-amnestic (na-MCI) MCI, with y-axis on a log scale. CRP levels were highest ...

In multivariable models, MetS was not significantly associated with MCI (combined a-MCI and na-MCI), a-MCI or na-MCI; the point estimates were all close to the null (Table 3). The point estimates and 95% CIs remained essentially the same in the fully-adjusted models. With additional adjustment for use of statins, the ORs were unchanged (Table 3). There were essentially no differences across strata based on statin use for association of MetS with all MCI, a-MCI, and na-MCI (data not reported). However, we did not have sufficient power to detect a significant effect modification for the association of MetS with na-MCI. High C-reactive protein (CRP highest tertile vs lowest tertile) was associated with na-MCI (OR = 1.85; 95% CI = 1.05, 3.24) but not with a-MCI (OR = 1.07; 95% CI = 0.76, 1.52), after adjusting for sex, age, and years of education. When CRP (tertiles) and MetS were simultaneously assessed in the same model, the highest CRP tertile was significantly associated with na-MCI (OR = 1.84; 95% CI = 1.05, 3.25) but not with a-MCI (OR = 1.09; 95% CI = 0.77, 1.54); MetS was not associated with either na-MCI (OR = 1.03; 95% CI = 0.66, 1.61) or a-MCI (OR = 0.90; 95% CI = 0.68, 1.20). The interactions between MetS and inflammation, as measured by tertiles of CRP were not statistically significant for all MCI (P = 0.872), for a-MCI (P = 0.877), or na-MCI (P = 0.435). When associations of MCI with the combined categories of inflammation (based on CRP tertiles) and MetS were examined, the combination of MetS and high CRP (compared to no MetS and lowest CRP tertile; the reference group) was associated with na-MCI (OR = 2.31; 95% CI = 1.07, 5.00; P = 0.03), but not with a-MCI (OR = 0.96; 95% CI = 0.59, 1.54; Table 4). However, given that the interaction between MetS and CRP did not reach significance, we cannot exclude the possibility that this difference predominantly reflects that main effect of CRP that we have observed.

Table 3
Associations of metabolic syndrome with all MCI, amnestic and non-amnestic MCI
Table 4
Associations of metabolic syndrome and tertiles of C-reactive protein with any MCI, amnestic and non-amnestic MCI

Discussion

In our population-based sample of 70 to 89 year olds, subjects with MetS and high levels of inflammation had a 2.3-fold elevated OR of na-MCI. Given several prior reports of an interaction between levels of inflammatory response markers and MetS, our finding was not completely unexpected.8,9,28 Although we found no overall association between MetS and MCI, we have previously observed in this cohort, that vascular phenomena such as diabetes,12 stroke,14 and CHD13 are associated with na-MCI rather than with a-MCI, the MCI subtype that is typically associated with AD. Thus, we have extended prior observations by others, in demonstrating that in the presence of inflammation, the MetS may be more likely to be associated with the form of cognitive impairment, na-MCI, that may be more closely linked to cerebrovascular disease.29

The observation that the combined presence of MetS and a high level of inflammation is associated with na-MCI but not with a-MCI, is novel. Other investigators have reported associations of MetS with cognitive impairment among subjects with high levels of inflammation.8,9,28 However, these studies have typically not distinguished between cognitive impairment subtypes. In two prospective cohort studies, subjects with MetS and high levels of inflammation had a greater cognitive decline during follow-up8 or an increased risk of cognitive impairment9,28 compared to subjects with low levels of inflammation. Thus, the effects of the MetS on cognition in elderly persons may be mediated directly through inflammation, or indirectly through effects of inflammation on cerebral atherosclerosis and endothelial dysfunction.6,22 Consistent with the observed modifying role of inflammation, it has been suggested that the effect of including CRP in criteria for MetS in predicting atherosclerotic vascular disease should be investigated.30

The lack of an association of MetS with a-MCI is inconsistent with the purported role of a-MCI as a prodromal stage for AD and the positive association of vascular risk factors with AD. This inconsistency may, in part, be due to a certain reasons. The association of Apoe ε4 allele with a-MCI may be stronger and evident earlier in the pathologic disease process than the effect of vascular diseases. This is confirmed by the significant association between a-MCI and presence of an Apoe ε4 allele (OR = 1.68; 95% CI = 1.23, 2.30) in the present study.14 In addition, Apoe ε4 allele has been associated with lower levels of inflammation.31,32 These factors may mitigate against observing an association of MetS with a-MCI in the earlier stages of the disease or until there is more severe vascular disease as in AD. Subjects with a-MCI in this study were at the very mild end of the disease spectrum; the mean CDR sum of boxes score was only 0.5. Thus, as with other studies that have not observed a predictive role of MetS for vascular outcomes in the short term,22,33 a positive association of MetS or vascular risk factors with a-MCI, or a-MCI progression to AD may be observed in the longer term. Thus, although vascular risk factors may be associated with both a-MCI and na-MCI, the association with a-MCI may be observed at a more severe stage of MCI.

We did not observe an association between MetS and MCI overall. This lack of an association when MCI subtype or inflammation is not taken into account may be related to the older age of subjects in the present study, and is consistent with similar studies conducted among older subjects. In the Leiden 85-Plus study (all age 85 years at baseline), there were no differences between subjects with and without the MetS for any of the cognitive tests at baseline;34 however, during follow-up there was a decelerated decline in cognition among subjects with MetS at baseline compared to subjects with no MetS.34 Similarly, among subjects with a mean age of 79 years, subjects with MetS had significantly better cognitive function compared to subjects without MetS (OR = 0.37; 95% CI = 0.15, 0.91).35

Alternately, significant positive associations between MetS and cognitive impairment have been reported in cohorts that were younger than subjects in the present study.8,9,28,36 Typically, these studies also did not distinguish between MCI subtypes or specific cognitive domains. In the Longitudinal Aging Study Amsterdam (65 to 88 year olds), MetS was associated with cognitive impairment cross-sectionally.28 In another prospective cohort (mean age 71 years), subjects with MetS at baseline had a greater cognitive decline than subjects without MetS during follow-up.8 In the Health, Aging and Body Composition Study (mean age 74 years), there were no significant differences in cognitive scores between subjects with and without MetS at baseline, but the MetS was associated with a 20% increased risk of cognitive impairment during follow-up.9 In the Singapore Longitudinal Aging Studies (mean age approximately 55 years), the MetS was significantly associated with a decline in cognition.37

Together, these studies suggest that the association between the MetS and cognitive function may vary with age at ascertainment of the MetS. Components of the MetS in mid-life are a risk factor for late-life dementia or accelerated cognitive impairment,38,39 yet, when assessed in elderly subjects, these conditions are associated with either a decelerated cognitive decline or have no association with late-life cognitive impairment.10,11,40 The decline in weight prior to dementia onset may represent adverse health conditions that may promote cognitive impairment or dementia.41 In contrast, high BMI in elderly persons may reflect well-being, consistent with the hypothesis that some forms of obesity may not be associated with insulin resistance and therefore, may not increase the risk of cognitive impairment.37 Similarly, elevated blood pressure in older persons may prevent cerebral hypoperfusion and thereby, may protect against cognitive impairment.40 It is also possible that the perceived beneficial effects of MetS or its components on cognition in the elderly may be due to survival bias wherein subjects with MetS or its components who are at greater risk for cognitive impairment have greater mortality, and do not survive to an older age.

There are several strengths of our study. The population-based design reduces the potential for selection bias. The prospective characterization of subjects for MCI using previously published specified criteria reduces the potential for misclassification of the MCI diagnosis, as could occur from retrospective application of MCI criteria to previously collected data on cognition, or from the use of non-established criteria for MCI. The medical records-linkage system allowed us to validate self-report of diabetes and hypertension, thereby minimizing misclassification regarding presence of the MetS. These findings in an older population-based cohort, add to the current literature suggesting that among the elderly, MetS or its components may not necessarily have an adverse impact on cognition. Future follow-up of the cohort will enable us to determine whether the combined presence of inflammation and MetS at baseline is associated with cognitive decline during follow-up, incident na-MCI, or whether in contrast, MetS without inflammation has no association with cognitive decline or contributes to a decelerated cognitive decline as suggested.34

Limitations to our study that should be noted include the cross-sectional assessment of the association of MetS with MCI; this precludes our ability to assess causality. There is potential for non-participation bias if subjects with MetS and cognitive impairment were less likely to participate in the study. To address potential non-participation bias, we used a propensity scores approach and medical record information abstracted from the record linkage system for participants and non-participants, to assess potential non-participation bias. This approach takes into account the characteristics of non-participants42 and assigns higher weights for participants who have characteristics similar to non-participants in logistic regression models. The propensity-adjusted estimates for this study were the same as the unadjusted estimates, suggesting an absence of non-participation bias on our findings (these results are not reported). Finally, the study was conducted among an older cohort of primarily Caucasian subjects; therefore caution should be exercised in applying the findings to subjects not represented in the study.

Acknowledgments

The authors thank the National Institute on Aging (U01 AG 06786, K01 AG 028573, P50 AG 16574), the National Institute of Mental Health (K01 MH 68351), the National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01 AR30582), and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program for support of this study.

Study Supported By: National Institute on Aging (U01 AG06786, P50 AG16574, and K01 AG028573), the National Institute of Mental Health (K01 MH68351), the National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01 AR30582), and by the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program.

References

1. Rutter MK, Meigs JB, Sullivan LM, et al. Insulin resistance, the metabolic syndrome, and incident cardiovascular events in the Framingham Offspring Study. Diabetes. 2005;54:3252–3257. [PubMed]
2. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III) JAMA. 2001;285:2486–2497. [PubMed]
3. Gustafson D, Rothenberg E, Blennow K, et al. An 18-year follow-up of overweight and risk of Alzheimer disease. Arch Intern Med. 2003;163:1524–1528. [PubMed]
4. Kivipelto M, Helkala EL, Laakso MP, et al. Midlife vascular risk factors and Alzheimer's disease in later life: longitudinal, population based study. BMJ. 2001;322:1447–1451. [PMC free article] [PubMed]
5. Vanhanen M, Koivisto K, Moilanen L, et al. Association of metabolic syndrome with Alzheimer disease: a population-based study. Neurology. 2006;67:843–847. [PubMed]
6. Yaffe K. Metabolic syndrome and cognitive disorders: is the sum greater than its parts? Alzheimer Dis Assoc Disord. 2007;21:167–171. [PubMed]
7. Kalmijn S, Foley D, White L, et al. Metabolic cardiovascular syndrome and risk of dementia in Japanese-American elderly men. The Honolulu-Asia Aging Study. Arterioscler Thromb Vasc Biol. 2000;20:2255–2260. [PubMed]
8. Yaffe K, Haan M, Blackwell T, et al. Metabolic syndrome and cognitive decline in elderly Latinos: findings from the Sacramento Area Latino Study of Aging Study. J Am Geriatr Soc. 2007;55:758–762. [PubMed]
9. Yaffe K, Kanaya A, Lindquist K, et al. The metabolic syndrome, inflammation, and risk of cognitive decline. JAMA. 2004;292:2237–2242. [PubMed]
10. Weverling-Rijnsburger A, Blauw GJ, LA M, et al. Total cholesterol and risk of mortality in the oldest old. Lancet. 1997;350:1119–1123. [PubMed]
11. Verghese J, Lipton RB, Hall CB, et al. Low blood pressure and the risk of dementia in very old individuals. Neurology. 2003;61:1667–1672. [PubMed]
12. Roberts RO, Geda YE, Knopman DS, et al. Association of duration and severity of diabetes mellitus with mild cognitive impairment. Arch Neurol. 2008;65:1066–1073. [PMC free article] [PubMed]
13. Roberts RO, Knopman DS, Geda YE, et al. Coronary heart disease is associated with non-amenstic mild cognitive impairment. Neurobiol Aging. 2008 in press. [PMC free article] [PubMed]
14. Knopman DS, Roberts RO, Geda YE, et al. Association of prior stroke with cognitive function and cognitive impairment: a population-based study: complementary relationships of stroke and APOE on cognition. Arch Neurol. 2008 in press. [PMC free article] [PubMed]
15. Roberts RO, Geda YE, Knopman DS, et al. The Mayo Clinic Study of Aging: design and sampling, participation, baseline measures and sample characteristics. Neuroepidemiology. 2008;30:58–69. [PMC free article] [PubMed]
16. Melton LJ., 3rd History of the Rochester Epidemiology Project. Mayo Clin Proc. 1996;71:266–274. [PubMed]
17. Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med. 2004;256:183–194. [PubMed]
18. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM IV) 4th. Washington, D.C.: American Psychiatric Association; 1994.
19. Ivnik RJ, Malec JF, Smith GE. WAIS-R, WMS-R and AVLT norms for ages 56 through 97. Clin Neuropsychol. 1992;6:1–104.
20. International Diabetes Federation. The IDF consensus worldwide definiton for the metabolic syndrome. http://www.idf.org/webdata/docs/Metac_syndrome_def.pdf.
21. Alberti KG, Zimmet P, Shaw J. The metabolic syndrome--a new worldwide definition. Lancet. 2005;366:1059–1062. [PubMed]
22. Grundy SM, Brewer HB, Jr, Cleeman JI, et al. Definition of metabolic syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association Conference on Scientific Issues Related to Definition. Circulation. 2004;109:433–438. [PubMed]
23. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med. 1998;15:539–553. [PubMed]
24. The expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 2003;26:S5–S20. [PubMed]
25. Kaufer DI, Cummings JL, Ketchel P, et al. Validation of the NPI-Q, a brief clinical form of the Neuropsychiatric Inventory. J Neuropsychiatry Clin Neurosci. 2000;12:233–239. [PubMed]
26. Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. J Lipid Res. 1990;31:545–548. [PubMed]
27. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004;110:227–239. [PubMed]
28. Dik MG, Jonker C, Comijs HC, et al. Contribution of metabolic syndrome components to cognition in older individuals. Diabetes Care. 2007;30:2655–2660. [PubMed]
29. Mariani E, Monastero R, Ercolani S, et al. Vascular risk factors in mild cognitive impairment subtypes. Findings from the ReGAl Project. Dement Geriatr Cogn Disord. 2007;24:448–456. [PubMed]
30. Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005;112:2735–2752. [PubMed]
31. Eiriksdottir G, Aspelund T, Bjarnadottir K, et al. Apolipoprotein E genotype and statins affect CRP levels through independent and different mechanisms: AGES-Reykjavik Study. Atherosclerosis. 2006;186:222–224. [PubMed]
32. Laurin D, David Curb J, Masaki KH, et al. Midlife C-reactive protein and risk of cognitive decline: a 31-year follow-up. Neurobiol Aging. 2008 doi: 10.1016/j.neurobiolaging.2008.01.008. [PubMed] [Cross Ref]
33. Wilson PW. Estimating cardiovascular disease risk and the metabolic syndrome: a Framingham view. Endocrinol Metab Clin North Am. 2004;33:467–481. [PubMed]
34. van den Berg E, Biessels GJ, de Craen AJ, et al. The metabolic syndrome is associated with decelerated cognitive decline in the oldest old. Neurology. 2007;69:979–985. [PubMed]
35. Laudisio A, Marzetti E, Pagano F, et al. Association of metabolic syndrome with cognitive function: the role of sex and age. Clin Nutr. 2008;27:747–754. [PubMed]
36. Komulainen P, Lakka TA, Kivipelto M, et al. Metabolic syndrome and cognitive function: a population-based follow-up study in elderly women. Dement Geriatr Cogn Disord. 2007;23:29–34. [PubMed]
37. Stefan N, Kantartzis K, Machann J, et al. Identification and characterization of metabolically benign obesity in humans. Arch Intern Med. 2008;168:1609–1616. [PubMed]
38. Whitmer RA, Gustafson DR, Barrett-Connor E, et al. Central obesity and increased risk of dementia more than three decades later. Neurology. 2008;71:1057–1064. [PubMed]
39. Ronnemaa E, Zethelius B, Sundelof J, et al. Impaired insulin secretion increases the risk of Alzheimer disease. Neurology. 2008;71:1065–1071. [PubMed]
40. Qiu C, Winblad B, Fratiglioni L. The age-dependent relation of blood pressure to cognitive function and dementia. Lancet Neurol. 2005;4:487–499. [PubMed]
41. Knopman DS, Edland SD, Cha RH, et al. Incident dementia in women is preceded by weight loss by at least a decade. Neurology. 2007;69:739–746. [PubMed]
42. D'Agostino RB, Jr, Rubin DB. Estimating and using propensity scores with partially missing data. J Am Statistical Assoc. 2000;95:749–759.