Get Newsletter
AlzRisk Risk Factor Discussion
Back

Risk Factor:
Risk Factor Type: Behavior, Metabolic
Current Understanding:
The tables below present a number of studies whose results, taken collectively, provide evidence suggesting that obesity may be associated with AD and total dementia. Many studies indicate that midlife obesity is associated with elevated risk of AD and dementia. Studies of adiposity in late-life, however, have found an inverse association, with higher risk for AD and dementia in underweight individuals and a reduced risk in obese participants. Much or all of the latter association could be explained by reverse causation, complex survival issues, and limitations in exposure measures. These results are consistent with results from related studies on adiposity and cognitive decline and from neuroimaging studies. However, it is still unclear whether there is a critical age at which obesity has a stronger influence on the development of AD and dementia, and whether the association is with overall adiposity or central adiposity. Thus, studies with longer follow-up times that include measures of both types of adiposity are needed to better understand this association. Despite these uncertainties, however, the association of obesity with AD and dementia provides an additional reason to maintain a healthy weight throughout life. For a more detailed interpretation of study findings and putative mechanisms, please view the Discussion.
Literature Extraction: Search strategy  * New *
Last Search Completed: 08 September 2012 - Last content updated on 19 Nov 2012.

Risk Factor Overview

Cite as:

Goonesekera S, Li S, Weuve J, Blacker D. "Obesity." The AlzRisk Database. Alzheimer Research Forum. Available at: http://www.alzrisk.org. Accessed [date of access]*.

* * *

Introduction

The tables in the Risk Factor Overview summarize a group of observational studies that describe a suggestive association between obesity and incident Alzheimer disease (AD) and total dementia. We did not meta-analyze the results in the tables (in whole or in part), because there were too few comparable results from independent populations. In spite of variations in how obesity was conceptualized and in cut-points used to define categories, many of the studies suggest that being overweight as opposed to normal weight in midlife may be associated with elevated risk of AD and dementia. However, these and other studies also indicate that being underweight, especially later in life, is associated with increased AD and dementia risk. In addition, some studies suggest that being overweight later in life may be associated with lower risk. Taken together, these results may reflect an increased risk of AD with midlife adiposity, and a decreased risk with late-life adiposity. Of note, reverse causation, complex survival issues, and limitations in exposure measures, discussed in detail below, may explain much or all of the reverse association in late-life.

Potential Mechanisms of Action

Obesity may indirectly cause AD dementia via its contribution to cerebrovascular pathology (both as a general vascular risk factor and via increasing the risk for type 2 diabetes) (Luchsinger and Gustafson, 2009; Knopman and Roberts, 2010). In addition, it may also directly cause AD pathology via hyperinsulinemia and leptin resistance.

Pro-inflammatory adipocytokines produced by adipose tissue contribute to insulin resistance and peripheral hyperinsulinemia (Whitmer et al., 2007). With peripheral hyperinsulinemia, more insulin crosses the blood-brain barrier and impedes beta-amyloid clearance. The brain appears to produce insulin as well, which, in contrast to peripheral insulin, may promote beta-amyloid clearance, but this insulin production could be inhibited by the peripheral hyperinsulinemia of obesity. Via either of these mechanisms, higher levels of peripheral insulin found in obese individuals could result in increased levels of beta amyloid in the brain, and thus lead to the formation of beta amyloid plaques, one of the hallmark pathological features of AD (Luchsinger and Gustafson, 2009).

Leptin resistance, common in obese individuals, may also contribute to the effect of obesity on risk for AD. Leptin improves hippocampal plasticity and synaptic function in rodents (Harvey et al., 2006). It is hypothesized that leptin resistance in humans contributes to impaired cognitive function in later life, but thus far the data in humans are limited to an inverse association between circulating leptin levels and incident AD and dementia (Lieb et al., 2009), which could simply be a reflection of the association with adiposity.
The J-shaped association between body mass index (BMI; weight in kilograms divided by height in meters squared) and AD seen in some studies suggests that underweight or weight loss may be related to increased dementia risk. As discussed in detail below, the most likely explanation for this is reverse causation – that chronic disease causes weight loss and increased risk of AD. One possible biological explanation is that individuals with lower BMI may have lower levels of insulin-like growth factor-1, which is associated with impaired cognitive function (Aleman et al., 2000; Okereke et al., 2007).

Methodological Issues

Exposure

The studies reported in the Risk Factor Overview measured adiposity using waist circumference, waist-to-hip ratio (WHR), BMI, change in BMI, change in waist circumference, and change in weight– with some of these measured directly and others based on self-report. Adiposity is generally conceptualized as either total adiposity (percent body fat), or as distributed in specific compartments, such as central adiposity (abdominal body fat). It is not clear which of these is most closely linked to AD, although central adiposity is thought to involve more adverse metabolic activity and may be a more specific risk marker. Obesity is typically defined based on BMI, which is influenced by lean bone and muscle mass. In young highly muscular males, for instance, BMI may be high when adiposity is low. On the other hand, with age and conditions that decrease muscle and bone mass, BMI may stay constant or fall while adiposity rises; this is called sarcopenic adiposity and is associated with increased risk for chronic diseases such as diabetes and cardiovascular disease. Central adiposity is defined as greater amounts of intra-abdominal fat and is generally assessed by measurements of waist circumference or WHR. These measures are also imperfect as they depend to some extent on skeletal shape.

Reliability and validity of the exposure measurement. The gold standard measures of dual energy X-ray absorptiometry (DEXA) and bioelectrical impedance used to measure total adiposity are infeasible in epidemiologic studies (Luchsinger and Mayeux, 2007), as are the imaging studies used to measure abdominal adiposity (Yoshizumi et al., 1999, Thomas et al., 2012). Instead, such studies typically use body weight or BMI as a measure of total adiposity, and waist circumference or WHR as a measure of central adiposity. The limitations of BMI and circumference measures are offset to a substantial extent by the increased participation conferred by the convenience of these measures. However, these measures may be more problematic in some groups of participants than others. For instance, in older participants, WHR or waist circumference may provide more accurate assessments of adiposity than BMI (Whitmer et al., 2007; Luchsinger and Gustafson, 2009) due to changes in lean body mass. This may be particularly true for women, who typically have a greater amount of adipose for a given BMI than men (Gallagher et al, 1996). Similarly, a given BMI generally reflects more adiposity among Asians than among Caucasians (Luchsinger and Gustafson, 2009).

Weight and particularly circumference measures are subject to measurement error. In addition, some studies of obesity relied on self-reported weight, which is subject to additional error if participants misreport their weight. However, in most instances these types of errors would not be expected to be associated with AD diagnosis and would result in underestimates of the strength of the association between adiposity and AD.

Timing of exposure measurement. In the studies included in this review, some measurements of adiposity occurred in “midlife” when participants were in their 50s or younger, while others occurred in participants' 70s and later. The critical age at which adiposity has the greatest effect on the development of AD and dementia is unclear (Luchsinger and Mayeux, 2007). Evidence from some studies indicates that midlife BMI may have a J-shaped association with AD and dementia, with the lowest risk in the normal-weight category, and the highest in the obese category (e.g., Fitzpatrick et al., 2009). On the other hand, it appears that low body mass, or rapid decline in body mass, in late life may increase risk, and some of the studies reported here suggest that those who are overweight have a reduced risk compared to those who are underweight (e.g. Fitzpatrick et al., 2009; Luchsinger et al., 2007). These findings must be interpreted with caution (see Reverse Causation and Survival Bias below).

Design and Analysis

Confounding and mediation. As all studies included in the tables are observational, there could be unmeasured and unknown confounding by variables not accounted for in the analyses. All results shown were from analyses adjusted for age, and nearly all accounted for gender and level of education, which are important sources of confounding in estimating the association between adiposity and AD. Few studies adjusted for physical activity, however, which reduces obesity risk and is also associated with reduced AD risk. As physical activity is strongly linked with obesity, adjusting for this variable to overcome confounding may not provide a satisfactory solution. More broadly, obesity often occurs as part of the metabolic syndrome, with hypertension, diabetes, and hypercholesterolemia (Ford et al, 2002; Ford et al, 2010). Each of these appears to a greater or lesser extent to contribute to risk, whether directly or via contributing to risk for cerebrovascular disease. It is difficult to control for these because of the degree of correlation, and because any of these factors, or associated cerebrovascular disease, may also act as intermediates between obesity and AD. Interestingly, one study found that further adjustment for potential intermediates did not produce substantially attenuated effect estimates, suggesting that these variables are not solely intermediates between obesity and AD (Whitmer et al., 2007).

Reverse causation. Assessments of late-life adiposity typically precede AD assessments by only a few years (< 5 years) (e.g. Hayden et al., 2006; Luchsinger et al., 2012; Raffaitin et al., 2009; Muller et al., 2007; Forti et al., 2010). Thus, pre-clinical disease may have been present at the time the exposure was measured and we cannot exclude the possibility that obesity status is a consequence of AD rather than vice versa. Those with AD can experience a loss of sense of smell resulting in reduced appetite, and an altered sense of satiety and changes in chewing and swallowing as a consequence of loss of limbic function. These phenomena may account for the inverse association between late-life adiposity and AD (Knopman et al., 2007). It is also possible that individuals with prodromal dementia may lose weight by simply forgetting to have regular meals (Knopman et al., 2007). In addition, other chronic diseases such as heart failure can be associated with AD risk and can lead to frailty and weight loss in late life (Minotti et al., 1993; Rosca et al., 2012).

Some studies compared the rate of weight loss in participants who subsequently developed dementia with the rate of weight loss in those who did not (Stewart et al. 2005; Johnson et al., 2006; Gao et al., 2011; Ogunniyl et al., 2011). One such study found that, compared with healthy participants, the participants who were diagnosed with dementia exhibited a steeper decline in BMI a few years prior to the diagnosis of dementia, and showed an accelerated rate of weight decline with disease progression (Stewart et al., 2005). A similar study found that the rate of weight loss in the year preceding the diagnosis among participants who developed AD was nearly twice the rate of weight loss in those who did not (Johnson et al., 2006). Attempting to eliminate the influence of weight loss that occurred as a result of pre-clinical disease, one of these studies re-examined the association between change in BMI and AD excluding cases that occurred during the first four years of follow-up (Buchman et al., 2005). They still found a significant decrease in risk of AD with late life obesity, suggesting that reverse causation may not fully explain the change in the risk from mid-life to late life or that the effects of AD or associated disorders on body mass occur over a longer time frame.

Survival bias. The J-shaped relationship observed in some studies, with individuals who are underweight in midlife having an increased risk of AD compared with normal-weight individuals, could be due in part to leaner individuals being more likely to survive until the later decades of life (Berrington de Gonzales et al, 2010). In addition, this could partially explain the inverse association between late-life adiposity and AD if obese individuals who do survive until old age have characteristics that protect them from cardiovascular disease, despite their adiposity, and also protect them from dementia.

Effect modification. Evidence from some studies suggests that the association between adiposity and AD/total dementia varies by gender (e.g. Beydoun et al., 2008; Hayden et al., 2006; Gustafson et al., 2003) and by smoking status (e.g. Chiang et al., 2007). In these studies, the deleterious association between overweight and AD was generally stronger among women than among men (e.g. Beydoun et al., 2008; Hayden et al., 2006, Whitmer et al., 2007). There is also evidence for stronger associations among smokers than among non-smokers (e.g. Chiang, 2007).

Modeling issues. The studies discussed in this review modeled the relation of their adiposity measures to AD risk in a variety of ways. Some studies compared obese individuals to non-obese; this approach loses information by dichotomizing the data, and could mask an association. Other studies modeled their measure of adiposity as a continuous linear variable (e.g., Fitzpatrick et al., 2009; Gustafson et al., 2003; Hughes et al., 2009). If the underlying association follows a non-linear pattern, such as a U shape, this modeling approach could misspecify that association, potentially finding very little association or none at all. Some studies evaluated AD risk across categories of adiposity, which avoids both of these problems. However, if the dose-response curve is not constant within these exposure categories – for example, if underweight and normal-weight individuals have different risks but are grouped together (e.g., Kivipelto et al., 2005; Hayden et al., 2006; Muller et al., 2007; Raffaitin et al., 2008), this could also mask an association. Moreover, the variability in thresholds defining these categories makes it difficult to compare results from different studies.

Results from other lines of research

Results of studies that reported only on total dementia are generally consistent with those reported here. In midlife, both high and low BMI seem to be associated with increased risk (Beydoun et al., 2008; Anstey et al., 2011), while in late life a high BMI is inversely associated with risk (e.g. Atti et al., 2008; Dahl et al., 2008; Nourashemi et al., 2003). In certain instances, this inverse association could be at least partly attributed to reverse causation (Nourashemi et al., 2003).
Similarly studies of obesity and cross sectional cognitive impairment or longitudinal cognitive decline also suggest an association with midlife obesity. Several cross-sectional studies suggest that increased BMI in midlife is associated with lower cognitive function later in life (Dahl et al., 2010; Hassing et al., 2010). On the other hand, a cross-sectional study that measured participants’ BMI in late-life found that higher BMI was associated with reduced risk of cognitive impairment or dementia (West et al., 2009). Interestingly, however, the same study found an increased risk of cognitive impairment for those with higher WHR. Longitudinal studies that have evaluated the association between midlife adiposity and cognitive decline have provided mixed results with a positive association between midlife adiposity and cognitive decline in some studies (Dahl et al., 2010; Debette et al., 2011) and little or no association in others (Hassings et al., 2010).

There is also evidence that the direction of the association between overall adiposity (measured using BMI) and cognitive impairment depends on whether central adiposity (measured using WHR) is present. A cross-sectional study found that among individuals with greater central adiposity (WHR ≥0.8), those who had a normal BMI had a higher risk of cognitive impairment than those who were underweight or obese. However, this association was reversed in those with a WHR<0.8: those who were normal weight had a lower risk than the underweight and obese groups (Kerwin et al, 2011). Thus, the measure of adiposity used is critical for accurate interpretation of the association between obesity and cognitive function.
Imaging studies conducted on healthy participants also suggest an association between obesity and structural and functional abnormalities in the brain that may represent the early manifestation of a dementing disorder. A longitudinal study that evaluated the effects of midlife risk factors for cognitive decline found a positive association between midlife obesity, measured by WHR, and decline in total brain volume (Debette et al., 2011). In a cross-sectional study conducted on 103 healthy participants, obesity was associated with reduced white matter tract integrity in the brain (Staneck et al, 2011).

Discussion and recommendations

Overall, the data reflect a suggestive association between obesity and AD that depends on when and how obesity is assessed. While many studies suggest a deleterious association between midlife obesity and dementia, some suggest a reduced risk among obese participants when the association is evaluated later in life. The latter finding may relate in part to the limitations of BMI as a measure of adiposity in older adults. It could also be due to survival issues and reverse causation. Thus, further studies with longer follow-up times utilizing more accurate assessment of adiposity may be required to better understand the association. Studies aimed at disentangling whether percent body fat or central adiposity carries the greatest risk are also needed. Despite these uncertainties, however, the association of obesity and AD and dementia provides an additional reason to maintain a healthy weight throughout life.

References cited

Aleman A, de Vries WR, de Haan EHF, et al. Age-Sensitive Cognitive Function, Growth Hormone and Insulin-Like Growth Factor 1 Plasma Levels in Healthy Older Men. Neuropsychobiology 2000; 41: 73–78.

Anstey K, Cherbuin N, Budge M and Young J. Body mass index in midlife and late-life as a risk factor for dementia: A meta-analysis of prospective studies. Obesity Reviews 2011; 12: e426-e437.

Atti A, Palmer K, Volpata S, et al. Late-life body mass index and dementia incidence: Nine-year follow-up of data from the Kungsholmen Project. J Am Geriatr Soc. 2008; 56:111-116. Abstract

Berrington de Gonzalez A, Hartge P, Cerhan J, et al. Body-mass index and mortality among 1.46 million white adults. NEJM 2010; 363: 2211-2219.

Beydoun M, Lhotsky A, Wang Y, et al. Association of adiposity status and changes in early to mid –adulthood with incidence of Alzheimer’s disease.Am J Epidemiol. 2008, 168(10):179-1189. Abstract

Beydoun M, Beydoun H, and Wang Y. Obesity and central obesity as risk factors for incidental dementia and its subtypes: A systematic review and meta-analysis. Obesity Reviews 2008; 9: 204-218. Abstract

Buchmann A, Wilson R, Biemas J, et al. Change in body mass index and risk of incident Alzheimer disease. Neurology 2005; 65: 892-897.

Chiang CJ, Yip PK, Wu SC, et al. Midlife risk factors for subtypes of dementia: A nested case-control study in Taiwan. Am J Geriatr Psyciatry 2007; 15(9):762-771. Abstract

Dahl A, Lopponen M, Isoaho R, et al. Overweight and obesity in old age are not associated with greater dementia risk. J Am Geriatr Soc. 2008; 56: 2261-2266.

Debette S, Seshadri S, Beise A et al. Midlife vascular risk factor exposure accelerate structural brain aging and cognitive decline. Neurology 2011; 77:461-468.

Fitzpatrick A, Kuller L, Lopez O, Diehr P, et al. Midlife and late-life obesity and the risk of dementia. Arch Neurol 2009; 66(3): 336-342. Abstract

Ford E, Li C, and Zhao G. Prevalence and correlates of metabolic syndrome based on a harmonious definition among adults in the US. Journal of Diabetes 2010; 2: 180-193.

Ford E, Giles W, Dietz W, et al. Prevalence of the metabolic syndrome among US adults: Findings from the third National Health and Nutrition Examination Survey. JAMA 2002; 287(3): 356-359.

Gallagher D, Visser M, Sepulveda D, et al. How useful is body mass index for comparison of body fat mass across age, sex and ethnic groups? Am J of Epidemiol. 1996; 143: 228-239.

Gao S, Nguyen J, Hendrie H et al. Accelerated weight loss and incident dementia in an elderly African-american cohort. J Am Geriatr Soc. 2011; 59(1): 652-657.

Gazdzinski S, Kornak J, Weiner M, et al. Body mass index and magnetic resonance markers of brain integrity in adults. Ann Neurol. 2008; 63(5):652-657.

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. Abstract

Harvey J, Solovyova N, and Irving A.Leptin and its role in hippocampal synaptic plasticity.ProgLipid Res. 2006; 45(5):369-378.

Hayden K, Sandi P, Lyketsos C, et al. Vascular risk factors for incident Alzheimer disease and vascular dementia: The Cache County Study. Alzheimer Dis. Assoc. Disord. 2006; 20(2): 93-100. Abstract

Hughs ST, Borenstein A, Schofield E, et al. Association between late-life body mass index and dementia: The Kame Project. Neurology.2009; 72: 1741-1746.

Johnson D, Wilkins C and Morris J. Accelerated weight loss may precede diagnosis in Alzheimer disease. Arch Neurol. 2006; 63(9):1312-1317. Abstract

Kerwin D, Gaussoin S, Chlebowsk R, Kuller L, Vitolins M, Coker L, Kotchen J, Nicklas B, Wassertheil-Smoller S, Hoffman R, Espeland M for the Women’s Health Initiative Memory Study. J Am Geriatr. Soc. 2011; 59:107-112.

Kivipelto M, Ngandu T, Fratiglioni L, et al. Obesity and vascular risk factors at midlife and the risk of dementia and Alzheimer Disease. Arch. Neurology 2005; 62(10): 1556-1560.

Knopman D, Edland S, Cha R, et al. Incident dementia in women is preceded by weight loss by at least a decade. Neurology 2007; 69; 739-746.

Knopman D and Roberts R. Vascular risk factors: Imaging and neuropathologic correlates. J Alzheimer’s Disease, 2010; 20(3): 699-709. Abstract

Lieb W, Beiser A, Vasan R, et al. Association of plasma leprin levels with incident Alzheimer disease and MRI Measurements of brain aging. JAMA 2009; 302(23): 2565-2572. Abstract

Luchsinger J, Patel B, Tang M, et al. Measures of adiposity and dementia risk in the elderly. Arch Neurol. 2007; 64(3): 393-398. Abstract

Luchsinger J, Cheng D, Tang M, et al. Central obesity in the elderly is related to late-onset Alzheimer Disease. Alzheimer Dis. AssocDisord. 2012; 26(2): 101-5.

Luchsinger J and Gustafson D. Adiposity and Alzheimer’s disease. Curr Opin Clin Nutr Metab Care. 2009; 12(1): 15-21. Abstract

Luchsinger J and Mayeux R.Adiposity and Alzheimer’s disease. Curr Alzheimer Res. 2007; 4(2):127-134. Abstract

Minotti J, Pillay P, Oka R, et al. Skeletal muscle size: relationship to muscle function in heart failure. J Appl Physiol. 1993; 75(1): 373-81.

Muller M, Tang H, Schupf H, et al. Metabolic syndrome and dementia risk in a multiethnic elderly cohort. Dement Geriatr Cogn Disord. 2007; 24(3): 185-192. Abstract

Nourshashemi F, Deschamps V, and Krrien S. Body mass index and incidence of dementia: The PAQUID study. Neurology 2003; 66:117-119.

Ogunniyi A, Gao S, Unverzagt F, et al. Weight loss and incident dementia in elderly Yoruba Nigerians: A 10-year follow up study. Int Psychogeriatr. 2011; 23(3): 387-394.

Okereke O, Kang J, Ma J, et al. Plasma IGF-1 levels and cognitive performance in older women. Neurobiol Aging 2007; 28: 135-142.

Raffaitin C, Fin H, Empana J, et al. Metabolic syndrome and risk for incident Alzheimer’s Disease or vascular dementia: The Three-City Study. Diabetes Care.2009; 32: 169-174. Abstract

Rosca M, and Hoppel C. Mitochondrial dysfunction in heart failure. Heart Fail Rev. 2012; [Epub ahead of print]

Stanek K, Grieve S, Brickman A, et al. Obesity is associated with reduced white matter integrity in otherwise healthy adults. Obesity 2011; 19: 500-504.

Stewart R, Masaki K, Xue Q, et al. A 32-year prospective study of change in body weight and incident dementia: The Honolulu Asia Aging Study. Arch Neurol. 2005; 62: 55-60. Abstract

Thomas E, Parkinson J, Frost G, et al. The missing risk: MRI and MRS phenotyping of abdominal obesity and ectopic fat. Obesity 2012; 20(1): 76-87.

West NA and Haan MA. Body adiposity in late-life and risk of dementia or cognitive impairment in a longitudinal community-based study. J Gerontol A BiolSci Med Sci. 2009; 64(1): 103-109.

Whitmer R, Gunderson E, Quesenberry Jr. C, et al. Body mass index in midlife and risk of Alzheimer Disease and vascular dementia. Current Alzheimer Research.2007; 4:103-109. Abstract

Whitmer RA. The epidemiology of adiposity and dementia. Current Alzheimer Research 2007; 4: 117-122. Abstract

Yoshizumi T, Nakamura T, Yamani M, et al. Abdominal fat: Standardized technique for measurement at CT. Radiology 1996; 211: 283-286.