Abstract

Background identification of modifiable risk factors is crucial in the prevention of dementia, given its limited treatment options. Studies on increased body mass index (BMI) as a risk factor for dementia show conflicting results.

Methods we systematically retrieved and reviewed longitudinal population-based studies on increased BMI and dementia using a standard protocol. We searched Medline (1966–2006), Ageline (1978–2006), PsychInfo (1966–2006), CINAHL (1982–2006), and other relevant databases, including the reference lists of theeligible aricles for review. Included studies were subjected to a quality assessment protocol.

Results we identified eight studies that met our selection criteria. These studies covered 1,688 cases of dementia from 28,697 participants. After adjustment for age, smoking, comorbidities, and other confounders, four studies presented significantly increased risk of dementia with elevated BMI.

Conclusion this systematic review supports the hypothesis that increased BMI is independently associated with increased risk of dementia. Long-term studies to examine the mechanisms underlying the relationship between obesity and dementia are needed.

Introduction

The prevention of dementia through identification and management of risk factors is of public health importance because of the growth in the elderly population. Dementia affects at least 6–10% of persons older than 65 years [1]. It increasingly causes significant morbidity and mortality with Alzheimer's disease ranking as the eighth leading cause of death among the elderly in the USA [2]. Despite therapeutic advances, current treatment modalities for dementia have limited success [3].

Dementia, in general, is increasingly recognised as a complication of cardiovascular risk factors [4, 5, 6]. For vascular-type dementia, this association is generally accepted [7, 8, 9]. Current trends in research also implicate the same risk factors in the development of Alzheimer's type dementia [10, 11]. It is possible that there is a common contributory pathway for the development of both Alzheimer's dementia and vascular-type dementia [12, 13]. It is possible that obesity contributes to both these major forms of cognitive impairment [14].

Obesity occurs in association with hypertension and diabetes [7, 8], but it is unclear whether obesity predisposes to dementia independent of other cardiovascular risk factors. The rationale for hypothesising this causal link is supported by experimental data suggesting the development of Alzheimer's type dementia with leptin dysregulation [15]. Similarly, cross-sectional studies have presented that measures of obesity, such as elevated waist–hip ratio [16] and increased body mass index (BMI) [17, 18], are associated with temporal lobe atrophy and decreased hippocampal volume.

Alarmingly, obesity affects 25–30% of adults in industrialised countries [19, 20]. Among older adults, aged 65–74 years, the number of overweight and obese adults has been increasing from 60 and 27%, respectively, in 1988, to 70 and 39%, respectively, in 2000 [21]. The prevalence of both obesity and dementia is expected to increase further with the growth of the aging population. Epidemiological studies on increased BMI as a risk factor for dementia show conflicting results. Some of these studies are limited by variable follow-up, study design and inadequate adjustment for medical comorbidities. In this context, we systematically reviewed all the available longitudinal studies investigating the association of increased BMI and dementia.

Methods

Data sources

We searched Medline (1966–March, 2006) using Ovid, EMBASE (1974–March, 2006), CINAHL (1982–March, 2006), Digital Dissertations Online (1997–March, 2006), PsychInfo (1966–March, 2006), and Ageline (1978– March, 2006) for papers published in English, using the combinations of medical subject heading (MeSH) terms obesity, BMI, body weight, and dementia, as per the Cochrane Collaboration recommended search strategy. We systematically searched the reference list of the retrieved articles and consulted experts in the field for pertinent unpublished data.

Selection criteria

We included longitudinal population-based studies that probed the association between increased BMI and dementia. Included studies ascertained the outcome of dementia on the basis of accepted clinical guidelines. We excluded studies reporting other cognitive abnormalities such as cerebral atrophy, cognitive decline, decreased functional capacity, and studies conducted on patients who already had psychiatric or neurologic impairments at baseline. We used the National Heart, Lung, and Blood Institute (NHLBI) BMI categories, with overweight defined as a BMI of 25–29.9 kg/m2 and obesity as a BMI >30 kg/m2 [22]. We independently searched for and identified potentially relevant studies using a uniform search strategy and a standard screening protocol (Figure 1). We initially found 532 articles of which 26 studies were considered relevant on the basis of the review of abstracts. We also sought unpublished [23], submitted, and additional data [26] from experts in the field.

Figure 1

Selection process for relevant studies

Figure 1

Selection process for relevant studies

For the eight included studies [7, 14, 23–28], we independently abstracted information on study design, participant characteristics, measurement of BMI and dementia, adjustment of potential confounders, and estimates of association. Both authors independently rated each eligible paper using a quality assessment checklist (Table 1). This tool was developed by the authors on the basis of recommendations for assessing longitudinal observational studies [29, 30]. We assessed the reporting of the baseline cognitive function of the participants, measurement of BMI, and adjustment for confounders like smoking, medical comorbidities, and age. Disagreements in the selection and quality assessment were resolved by discussion and repeated assessment of the articles.

Table 1

Quality assessment checklist

Criteria Quality assessment score 
1. Was the study design explicitly stated? 
2. Was baseline cognitive status reported? 
3. Was the average age of participants below 65 years? 
4. Was BMI recorded at baseline (check only one item)  
 By standardized clinical measurement? 
 Or by self-reporting? 
 Did not specify 
5. Was BMI reported in categories (check only one item)  
 Using NHLBI cut-off values? 
 Or by differentiation between obese & non-obese? 
 Other classification scheme/Not specified? 
6. Was Dementia diagnosed by  
 Using DSM, ICD or NINCDS-ADRDA criteria? 
 Not specified 
7. Was follow-up time reported? 
8. Were the following covariates adjusted? (check all that applies)  
 Age? 
 Smoking? 
 Medical comorbidities? 
 APOE-ɛ4 carrier status? 
9. Were dropouts from the cohort documented? 
10. Was risk of Dementia analyzed by (check only one item)  
 Time to event/ Hazards ratio/Relative risk? 
 Linear regression/Means 
 Others 
TOTAL 16 
Criteria Quality assessment score 
1. Was the study design explicitly stated? 
2. Was baseline cognitive status reported? 
3. Was the average age of participants below 65 years? 
4. Was BMI recorded at baseline (check only one item)  
 By standardized clinical measurement? 
 Or by self-reporting? 
 Did not specify 
5. Was BMI reported in categories (check only one item)  
 Using NHLBI cut-off values? 
 Or by differentiation between obese & non-obese? 
 Other classification scheme/Not specified? 
6. Was Dementia diagnosed by  
 Using DSM, ICD or NINCDS-ADRDA criteria? 
 Not specified 
7. Was follow-up time reported? 
8. Were the following covariates adjusted? (check all that applies)  
 Age? 
 Smoking? 
 Medical comorbidities? 
 APOE-ɛ4 carrier status? 
9. Were dropouts from the cohort documented? 
10. Was risk of Dementia analyzed by (check only one item)  
 Time to event/ Hazards ratio/Relative risk? 
 Linear regression/Means 
 Others 
TOTAL 16 

Results

Description of included studies

We identified eight cohort studies on BMI and dementia, which included 28,697 participants with 1,688 cases of dementia representing study populations from Finland, France, Sweden, Japan, and the USA (Table 2). All the included studies presented outcomes of all-cause dementia, except for the study by Yoshitake et al. [7] who reported separate relative risks for Alzheimer's disease and vascular dementia. Included studies used the DSM III-R, ICD, or the National Institute of Neurological and Communication Disorders and Stroke/Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria for the diagnosis of dementia. Four studies [24–26, [28] evaluated the BMI-associated risk in middle-aged individuals while the rest [7, 14, 23, 27] included adults aged 65–70 years. Two studies [25, 28] included only men while the rest included both men and women [7, 14, 23, 24, 26, 27].

Table 2

Description of included studies

Authors Country/publication year Sample size No. of dementia cases Participants' ages at baseline Average follow-up years Quality rating (0–16) 
Whitmer et al. [24USA/2005 10,276 713 40–45 27 14 
Rosengren et al. [25Sweden/2005 7,402 254 47–55 25 14 
Luchsinger [23USA/unpublished 980 183 77 ± 5.7 15 
Kivipelto et al. [26Finland/2004 1,449 61 50.6 ± 6.0 21 15 
Gustafson et al. [14Sweden/2003 382 93 70 18 12 
Nourhashémi et al. [27France/2003 3,646 66 ≥65 
Kalmijn et al. [28USA/2000 3,734 215 45–68 25 
Yoshitake et al. [7Japan/1995 828 103 ≥65 12 
Authors Country/publication year Sample size No. of dementia cases Participants' ages at baseline Average follow-up years Quality rating (0–16) 
Whitmer et al. [24USA/2005 10,276 713 40–45 27 14 
Rosengren et al. [25Sweden/2005 7,402 254 47–55 25 14 
Luchsinger [23USA/unpublished 980 183 77 ± 5.7 15 
Kivipelto et al. [26Finland/2004 1,449 61 50.6 ± 6.0 21 15 
Gustafson et al. [14Sweden/2003 382 93 70 18 12 
Nourhashémi et al. [27France/2003 3,646 66 ≥65 
Kalmijn et al. [28USA/2000 3,734 215 45–68 25 
Yoshitake et al. [7Japan/1995 828 103 ≥65 12 

Quality assessment

Despite their variability in methodology and analytic models, all studies scored well in the quality assessment tool with a mean of 12.25 ± 2.73 points from a possible total of 16 points (Table 2). Overall, most of the studies scored well in terms of their adjustment for confounders such as age, smoking, and other medical comorbidities. Four studies [7, 23, 14, 27] explicitly stated that they have performed baseline cognitive assessment of their participants. Rosengren et al. [25] excluded participants with previous stroke. BMI was computed on the basis of clinical weight and height measurements in all studies except for the study by Nourhashemi et al. [27], who utilised self-reported data for participants when they still had normal cognitive function at baseline. Nourhashémi et al. [27] excluded subjects with dementia diagnosed within the first 3 years of follow-up in order to exclude subjects who might have preclinical dementia during their baseline assessment.

Association between BMI and dementia

Table 3 presents the findings of the eight studies and a summary of their conclusions. In their paper, Kivipelto et al. [26] had three models for the analysis of midlife BMI and risk of dementia. We presented their third model, which was adjusted for apolipoprotein E-ɛ4 (APOE-ɛ4) and history of diabetes mellitus, stroke, and myocardial infarction. The unpublished data [23] from the Washington Heights-Inwood Columbia Aging Project (WHICAP) also had three models in their analysis; none of them showed statistically significant relationships between the increasing BMI categories and dementia. In this review, we chose to present the model of the WHICAP study that adjusted for age, gender, years of education, ethnicity, APOE-ɛ4 carrier status, smoking, and other cardiovascular risk factors.

Table 3

Summary of findings

Studies presenting positive & significant results 
Authors Results Conclusions 
Whitmer et al. [24BMI ≥30 25–29.9 18.5–24.9 HR (95% CI) 1.74 (1.34–2.26) 1.35 (1.14–1.60) 1.0 Adjusted for age, smoking, and other comorbidities; high BMI significantly increases the risk of dementia 
 ≤18.5 1.24 (0.70–2.21) BMI-associated risk of dementia higher among women than in men 
Rosengren et al. [25BMI ≥30 27.5–29.99 25–27.49 22.5–24.99 20–22.49 <20 HR (95% CI) 1.98 (1.10–3.56) 1.72 (1.03–2.88) 1.49 (0.92–2.41) 1.49 (0.92–2.41 1.0 2.43 (1.10–5.29) Adjusted for age, smoking, and other comorbidities; high BMI significantly increases the risk of dementia 
Gustafson et al. [14HR for total dementia among women Age with high BMI BMI at age 70 BMI at age 75 BMI at age 79 HR (95% CI) 1.13 (1.04–1.24) 1.13 (1.04–1.24) 1.15 (1.05–1.26) Adjusted for smoking, comorbidities, and socioeconomic status; women with high BMI at ages 70–79 have significant risk for dementia No association found in men 
Kalmijn et al. [28BMI (1 − SD increase) 2.9 kg/m2 RR 95% CI 1.21 (1.05–1.40) Adjusted for age and education only, high BMI significantly increases the risk of dementia 25 years later 
Studies presenting no significant association 
Authors Results Conclusions 
Luchsinger [23BMI >30 25–30 18.5–24.9 ≤18.5 HR (95% CI) 0.9 (0.6–1.3) 0.6 (0.4–0.9) 1.0 1.4 (0.7–2.9) Adjusted for age, education, smoking, APOE-ɛ4 carrier status, and other cardiovascular risk factors; no significant association found with increasing BMI and dementia 
Kivipelto et al. [26BMI >30 25–30 ≤25 OR (95% CI) 1.88 (0.76–4.63) 0.99 (0.47–2.15) 1.0 Adjusted for age, smoking, comorbidities, and other risk factors; midlife obesity has a higher but non-significant risk for dementia 
Nourhashemi et al. [27BMI ≥27 23–26 21–22 <21 RR (95% CI) 0.716 (0.429–1.195) 1.0 0.709 (0.401–1.254) 1.185 (0.716–1.960) Adjusted for age, smoking, and exclusion of early cases of dementia but not for medical comorbidities; no significant association with BMI and dementia found 
Yoshitake et al. [7For Alzheimer's disease RR 0.75 95% CI (0.54–1.03) Adjusted for age, smoking, and other comorbidities; BMI has no significant association for either Alzheimer's disease or vascular dementia 
 For vascular dementia  
 RR 1.31 (0.98–1.74)  
Studies presenting positive & significant results 
Authors Results Conclusions 
Whitmer et al. [24BMI ≥30 25–29.9 18.5–24.9 HR (95% CI) 1.74 (1.34–2.26) 1.35 (1.14–1.60) 1.0 Adjusted for age, smoking, and other comorbidities; high BMI significantly increases the risk of dementia 
 ≤18.5 1.24 (0.70–2.21) BMI-associated risk of dementia higher among women than in men 
Rosengren et al. [25BMI ≥30 27.5–29.99 25–27.49 22.5–24.99 20–22.49 <20 HR (95% CI) 1.98 (1.10–3.56) 1.72 (1.03–2.88) 1.49 (0.92–2.41) 1.49 (0.92–2.41 1.0 2.43 (1.10–5.29) Adjusted for age, smoking, and other comorbidities; high BMI significantly increases the risk of dementia 
Gustafson et al. [14HR for total dementia among women Age with high BMI BMI at age 70 BMI at age 75 BMI at age 79 HR (95% CI) 1.13 (1.04–1.24) 1.13 (1.04–1.24) 1.15 (1.05–1.26) Adjusted for smoking, comorbidities, and socioeconomic status; women with high BMI at ages 70–79 have significant risk for dementia No association found in men 
Kalmijn et al. [28BMI (1 − SD increase) 2.9 kg/m2 RR 95% CI 1.21 (1.05–1.40) Adjusted for age and education only, high BMI significantly increases the risk of dementia 25 years later 
Studies presenting no significant association 
Authors Results Conclusions 
Luchsinger [23BMI >30 25–30 18.5–24.9 ≤18.5 HR (95% CI) 0.9 (0.6–1.3) 0.6 (0.4–0.9) 1.0 1.4 (0.7–2.9) Adjusted for age, education, smoking, APOE-ɛ4 carrier status, and other cardiovascular risk factors; no significant association found with increasing BMI and dementia 
Kivipelto et al. [26BMI >30 25–30 ≤25 OR (95% CI) 1.88 (0.76–4.63) 0.99 (0.47–2.15) 1.0 Adjusted for age, smoking, comorbidities, and other risk factors; midlife obesity has a higher but non-significant risk for dementia 
Nourhashemi et al. [27BMI ≥27 23–26 21–22 <21 RR (95% CI) 0.716 (0.429–1.195) 1.0 0.709 (0.401–1.254) 1.185 (0.716–1.960) Adjusted for age, smoking, and exclusion of early cases of dementia but not for medical comorbidities; no significant association with BMI and dementia found 
Yoshitake et al. [7For Alzheimer's disease RR 0.75 95% CI (0.54–1.03) Adjusted for age, smoking, and other comorbidities; BMI has no significant association for either Alzheimer's disease or vascular dementia 
 For vascular dementia  
 RR 1.31 (0.98–1.74)  

After adjusting for age, smoking, medical comorbidities, and other confounders, four studies [14, 24, 25, 28] reported significant association between increased BMI and dementia, which included the larger and more recent studies of Whitmer et al. [24] and Rosengren et al. [25]. Two other studies [7, 26] reported higher risk of dementia with increased BMI, but the association did not reach statistical significance. As a group, the four studies [14, 24, 25, 28] with statistically significant results had larger sample sizes (21,794 versus 6,903 participants), longer follow-up periods (23.75 ± 3.95 versus 10.25 ± 6.30 years), and younger participants at baseline. In our review, no study on dementia reported any protective effect of increased BMI.

Discussion

There is a prevailing assumption that obesity is not a major risk factor for age-related cognitive decline [31]. This review suggests that increased BMI is likely to be an independent risk factor for dementia. Studies with appropriate exclusion of patients with dementia at baseline, enough power, and appropriate adjustment of potential confounders support the association between increased BMI and dementia. Studies [24, 25] that have younger participants at baseline and longer follow-up periods showed even more significant association.

BMI is an acceptable measure of body fat or adiposity in the general population. In the Heritage Family Study, the correlation between BMI and per cent body fat was 0.73 [32]. However, among older individuals, BMI may underestimate adiposity because lean body mass is replaced by fat on aging [33]. The association of increased adiposity to dementia may be weaker among the older participants who may have more body fat in spite of their low body weight. This also implies that the association of BMI to dementia may be more accurately assessed at midlife than in the older years [24]. In addition, the study of younger participants may also exclude the older and possibly underweight subjects who may have undetected preclinical dementia. In this review, studies [24, 25, 28] that measured baseline BMI at midlife were the ones that showed stronger association.

Studies such as that by Rosengren et al. [25], which did not exclude patients diagnosed with dementia in their early follow-up period, and the WHICAP data [23], which only had a 5-year follow-up, both reported a J-shaped association between BMI and dementia. This observation requires further scrutiny as patients with dementia often lose weight. With dementia, previously obese patients may lose up to 50% of their pre-dementia body weight [34]. This weight loss could even precede diagnosis of dementia by several years [35]. The study of Nourhashemi et al. [27], which excluded participants with dementia during their first 3 years of follow-up, concluded that low BMI did not present any significant risk. It appears that failure to exclude subjects diagnosed early with dementia in a cohort and inadequate follow-up period may skew the risk of dementia towards those with borderline and low BMI.

In this review, women with increased BMI were at a greater risk for dementia than men [14, 24]. Data from the pooled European population-based cohorts also demonstrated increased incidence of dementia among women [36]. These observations propose a possible interaction between obesity and gender in the risk for dementia. Several hypotheses for this difference include hormonal changes, gender differences in body fat distribution [14, 24], or probable survival bias, as women with higher BMI are more likely to survive past the age of 70 than men [14, 37]. More studies will be needed to answer these questions.

The association of BMI to dementia is complex. Ascertaining obesity as an independent risk factor for dementia requires careful adjustment of confounding variables, adequate number of dementia cases, exclusion of cognitively impaired subjects at baseline, standardised measurement of body weight and height, and uniform BMI cut-off categories as suggested by the NHLBI guidelines [22]. Whenever possible, investigators assessing the outcome of dementia should be blinded to the baseline BMI measures of their participants.

Smoking may be a risk factor for dementia [38] and needs to be excluded. Smokers have been shown to be thinner, with higher risks for mortality and other comorbidities [39]. Failure to exclude smokers may produce a J-shaped correlation between BMI and other health outcomes such as mortality and dementia. Owing to smoking, underweight cohorts may show greater risk for dementia than those with normal or increased BMIs. More importantly, there should be adjustment for APOE-ɛ4 carrier status in the analyses of studies. APOE-ɛ4 carrier status is a strong risk factor for dementia [40, 41]. Although controversial, there is also growing evidence that low educational attainment [42] and decreased physical activity [43, 44] are potential risk factors for dementia.

Long-term well-designed studies are still needed to examine the mechanisms underlying the relationship between obesity and dementia. For now, it appears that elevated BMI in middle age and in the older years seems to have no protective effect against cognitive impairment. Therefore, the maintenance of normal weight throughout one's lifespan is a worthwhile intervention for the prevention of dementia, aside from the prevention of other established medical comorbidities.

Key points

  • This review suggests increased BMI as a likely independent risk factor for dementia.

  • Adjustment for age, medical comorbidities, APOE-ɛ4 carrier status, smoking, and adequacy of follow-up time can affect the results and quality of studies investigating the link between BMI and dementia.

  • There is a need to explore the role of increased adiposity for the aging brain and to elucidate reasons for gender differences in the risk for dementia.

Acknowledgements

The authors would like to thank Jose Luchsinger of the Taub Institute for Research of Alzheimer's Disease and the Aging Brain for sharing unpublished data, and Miia Kivipelto of the Karolinska Institutet Aging Research Center for her correspondence regarding published articles.

Conflict of interest

None declared.

Funding

None.

Ethical approval

Not required.

References

The long list of references supporting this review has meant that only the most important are listed here and are represented by bold type throughout the text. The full list of references is available on the journal website http://www.ageing.oxfordjournals.org/ as appendix 1.

1
Hendrie
HC
Epidemiology of dementia and Alzheimer's disease
Am J Geriatr Psychiatry
 , 
1998
, vol. 
6
 (pg. 
S3
-
18
Suppl. 1
2
Centers of Disease Control and Prevention
Death: preliminary data 2003
Natl Vital Stat Rep
 , 
2005
, vol. 
53
 pg. 
2
 
3
Wallin
AK
Gustafson
L
Sjogren
M
, et al.  . 
Five-year outcome of cholinergic treatment of Alzheimer's disease: early response predicts prolonged time until nursing home placement, but does not alter life expectancy
Dement Geriatr Cogn Disord
 , 
2004
, vol. 
18
 (pg. 
197
-
206
)
4
Breteler
MM
Bots
ML
Ott
A
, et al.  . 
Risk factors for vascular disease and dementia
Haemostasis
 , 
1998
, vol. 
28
 (pg. 
167
-
73
)
7
Yoshitake
T
Kiyohara
Y
Kato
I
, et al.  . 
Incidence and risk factors of vascular dementia and Alzheimer's disease in a defined elderly Japanese population: the Hisayama Study
Neurology
 , 
1995
, vol. 
45
 (pg. 
1161
-
68
)
10
Launer
LJ
Ross
GW
Petrovitch
H
, et al.  . 
Midlife blood pressure and dementia: the Honolulu-Asia aging study
Neurobiol Aging
 , 
2000
, vol. 
21
 (pg. 
49
-
55
)
12
Hofman
A
Ott
A
Breteler
MM
, et al.  . 
Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer's disease in the Rotterdam Study
Lancet
 , 
1997
, vol. 
349
 (pg. 
151
-
54
)
14
Gustafson
D
Rothenberg
E
Blennow
K
, et al.  . 
An 18-year follow-up of overweight and risk of Alzheimer disease
Arch Intern Med
 , 
2003
, vol. 
163
 (pg. 
1524
-
28
)
15
Fewlass
DC
Noboa
K
Pi-Sunyer
FX
, et al.  . 
Obesity-related leptin regulates Alzheimer's Abeta
FASEB J
 , 
2004
, vol. 
18
 (pg. 
1870
-
78
)
16
Jagust
W
Harvey
D
Mungas
D
, et al.  . 
Central obesity and the aging brain
Arch Neurol
 , 
2005
, vol. 
62
 (pg. 
1545
-
48
)
17
Ward
MA
Carlsson
CM
Trivedi
MA
, et al.  . 
The effect of body mass index on global brain volume in middle-aged adults: a cross sectional study
BMC Neurol
 , 
2005
, vol. 
5
 pg. 
23
 
19
Rennie
KL
Jebb
SA
Prevalence of obesity in Great Britain
Obes Rev
 , 
2005
, vol. 
6
 (pg. 
11
-
2
)
21
Centers for Disease Control and Prevention
 
Healthy weight, overweight, and obesity among persons 20 years of age and over, according to sex, age, race, and Hispanic origin: United States, 1960–62, 1971–74, 1976–80, 1988–94, and 1999–2000. National Center for Health Statistics Data 2002. Accessed March 29, 2006 from http://www.cdc.gov/nchs/data/hus/tables/2002/02hus070.pdf
22
National Heart Lung Blood Institute
The Practical Guide on Identification, Evaluation, and Treatment of Overweight and Obesity
2000
National Institutes of Health
23
Luchsinger
J
Washington Heights-Inwood Columbia Aging Project
 
Taub Institute for Research of Alzheimer's Disease and the Aging Brain
24
Whitmer
RA
Gunderson
EP
Barrett-Connor
E
, et al.  . 
Obesity in middle age and future risk of dementia: a 27 year longitudinal population based study
BMJ
 , 
2005
, vol. 
330
 pg. 
1360
 
25
Rosengren
A
Skoog
I
Gustafson
D
, et al.  . 
Body mass index, other cardiovascular risk factors, and hospitalization for dementia
Arch Intern Med
 , 
2005
, vol. 
165
 (pg. 
321
-
26
)
26
Kivipelto
M
Ngandu
T
Fratiglioni
L
, et al.  . 
Obesity and vascular risk factors at midlife and the risk of dementia and Alzheimer disease
Arch Neurol
 , 
2005
, vol. 
62
 (pg. 
1556
-
60
)
27
Nourhashemi
F
Deschamps
V
Larrieu
S
, et al.  . 
Body mass index and incidence of dementia: the PAQUID study
Neurology
 , 
2003
, vol. 
60
 (pg. 
117
-
9
)
28
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
, vol. 
20
 (pg. 
2255
-
60
)
29
Stroup
DF
Berlin
JA
Morton
SC
, et al.  . 
Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group
JAMA
 , 
2000
, vol. 
283
 (pg. 
2008
-
12
)
31
Elias
MF
Elias
PK
Sullivan
LM
, et al.  . 
Lower cognitive function in the presence of obesity and hypertension: the Framingham heart study
Int J Obes Relat Metab Disord
 , 
2003
, vol. 
27
 (pg. 
260
-
8
)
32
Jackson
AS
Stanforth
PR
Gagnon
J
, et al.  . 
The effect of sex, age and race on estimating percentage body fat from body mass index: The Heritage Family Study
Int J Obes Relat Metab Disord
 , 
2002
, vol. 
26
 (pg. 
789
-
96
)
33
Baumgartner
RN
Heymsfield
SB
Roche
AF
Human body composition and the epidemiology of chronic disease
Obes Res
 , 
1995
, vol. 
3
 (pg. 
73
-
95
)
34
Wang
SY
Weight loss and metabolic changes in dementia
J Nutr Health Aging
 , 
2002
, vol. 
6
 (pg. 
201
-
5
)
35
Barrett-Connor
E
Edelstein
SL
Corey-Bloom
J
, et al.  . 
Weight loss precedes dementia in community-dwelling older adults
J Am Geriatr Soc
 , 
1996
, vol. 
44
 (pg. 
1147
-
52
)
36
Andersen
K
Launer
LJ
Dewey
ME
, et al.  . 
Gender differences in the incidence of AD and vascular dementia: The EURODEM Studies. EURODEM Incidence Research Group
Neurology
 , 
1999
, vol. 
53
 (pg. 
1992
-
97
)
38
Ott
A
Slooter
AJ
Hofman
A
, et al.  . 
Smoking and risk of dementia and Alzheimer's disease in a population-based cohort study: the Rotterdam Study
Lancet
 , 
1998
, vol. 
351
 (pg. 
1840
-
43
)
39
Manson
JE
Willett
WC
Stampfer
MJ
, et al.  . 
Body weight and mortality among women
N Engl J Med
 , 
1995
, vol. 
333
 (pg. 
677
-
85
)
40
Kuusisto
J
Koivisto
K
Kervinen
K
, et al.  . 
Association of apolipoprotein E phenotypes with late onset Alzheimer's disease: population based study
BMJ
 , 
1994
, vol. 
309
 (pg. 
636
-
8
)

Supplementary data

Comments

0 Comments