November 11, 2023

Preventing Chronic Disease

How does APOE status impact the relationship between obesity and dementia risk?

Obesity has been reported by some to increase risk of dementia, while other studies have reported that obesity is protective against dementia. Might the effects of APOE genotype help to explain the discrepancy?

Peter Attia

Read Time 7 minutes

As I’ve discussed in a recent “Ask Me Anything” episode on metabolic health, obesity (and the metabolic dysfunction that often accompanies it) is a risk factor for many of mankind’s deadliest diseases. But the relationship isn’t always straightforward – such as in the case of cognitive decline. While obesity has been reported by some to increase risk of dementia, other studies have reported just the opposite – that obesity is protective against dementia.

Of course, we’ve seen countless times that conflicting results can often be explained by biases in study population or other differences in study design. And although findings for obesity and dementia are certainly no exception, the existence of methodological explanations does not exclude the possibility that the apparent “paradox” might also reflect important gaps in our understanding of the biology underlying an association between two physiological variables. Indeed, in a recent study, authors Shinohara et al. investigated whether a third variable – namely, variants of the APOE gene – might influence the way in which obesity impacts dementia risk, explaining why it appears protective in some and detrimental in others.


As explained in detail on The Drive, the APOE gene, which codes for apolipoprotein E, exists in three different forms (known as “alleles”) in the human population: ε2, ε3, and ε4. Each individual has two copies of the APOE gene (one from each parent), and the combination of alleles they possess (i.e., their genotype) represents the strongest known genetic correlate of risk for late-onset Alzheimer’s disease (AD). (Late-onset AD is also known as “common AD,” differentiating it from the few cases of AD that are caused by very rare, highly penetrant gene variants, which, unlike APOE variants, nearly guarantee that one will develop AD.) Specifically, the ε4 allele, which is present in at least one copy in roughly 25% of the population, is associated with elevated risk of AD and other forms of dementia, while the ε2 allele is associated with modestly reduced risk.

The precise reasons for the differences in dementia risk across APOE alleles are not fully understood but are thought to relate to apolipoprotein E’s critical role in transporting cholesterol synthesized in astrocytes to neurons and in modulating neuroinflammation. Thus, APOE represents a link between lipid/lipoprotein biology and dementia, making it an attractive candidate for influencing the relationship between obesity and dementia.

What did the study show?

Authors Shinohara et al. used clinical and neuropathological records from the longitudinal U.S. National Alzheimer’s Coordinating Center (NACC) database to investigate how APOE status impacted the association between obesity and dementia risk. The investigators restricted their analysis to patients ≥60 years of age, with normal cognition, and for whom APOE status was known, classifying subjects with a BMI >30 at all visits as obese (n=3932) and subjects with a BMI from 20 to 30 at all visits as non-obese controls (n=16,116). Subjects were categorized based on APOE status as APOE2 carriers (two ε2 alleles or one ε2 and one ε3), APOE3 homozygotes (ε3/ε3) and APOE4 carriers (ε3/ε4 or ε4/ε4). Those with an ε2/ε4 genotype were excluded due to the opposing effects of these two alleles on dementia risk. Other exclusion criteria included having a BMI <20 at any visit or a BMI that changed between obese and non-obese across visits.

The investigators first assessed the relationship solely between obesity and dementia (i.e., without stratifying by APOE status), using a model adjusted for sex, race, education, APOE genotype, and diabetes. Not surprisingly, obesity was found to be associated with a higher risk of cognitive decline across all subjects (HR=1.17, 95% CI: 1.11-1.23, P<0.001) in terms of change over time in Clinical Dementia Rating (CDR) – a standardized scale assessing multiple domains of cognition. Specifically, the presence of obesity increased the likelihood of earlier cognitive decline, as stratification by age revealed that subjects with obesity were significantly more likely than non-obese subjects to exhibit cognitive decline prior to age 80, whereas this difference in risk vanished beyond this age mark. In contrast to results with cognitive decline, obesity was found to be associated with a lower overall frequency of dementia relative to the non-obese group (OR=0.710, 95% CI: 0.66-0.77, P<0.0001).

However, when the authors examined results separately for each APOE classification, they observed striking divergence in the relationship of obesity and dementia risk between subjects with different genotypes. While obesity was associated with earlier cognitive decline in APOE2 carriers (HR=1.42, 95% CI: 1.21-1.66, P<0.001) and APOE3 homozygotes (HR=1.24, 95% CI: 1.16-1.32, P<0.001), this was not the case in APOE4 carriers (HR=1.06, 95% CI: 0.99-1.14, P=0.350). Further, the apparent protective effect of obesity on dementia frequency was strongest in APOE4 carriers (OR=0.648, 95% CI: 0.57-0.73, P<0.001), moderate in APOE3 homozygotes (OR=0.722, 95% CI: 0.65-0.81, P<0.001), and nonexistent in APOE2 carriers (OR=0.910, 95% CI: 0.71-1.17, P=1.00).

Does obesity really reduce dementia risk?

Let’s first consider Shinohara et al.’s observation that obesity has a net effect of reducing lifetime risk of dementia, regardless of APOE status. Can this be true? Possibly, but if so, it is likely primarily because patients with obesity are more likely to die of other causes prior to the development of this degree of cognitive impairment. Obesity is associated with increased risk of cardiovascular disease, cancer, liver disease, and other pathologies that can cause death in relatively early decades of life (i.e., <70), whereas dementia incidence increases sharply after age 70. 

Shinohara et al. did not report statistics on causes of death, so we cannot say for certain whether the obese group had higher rates of death from non-dementia causes. However, data on the age at death between groups support the explanation that the apparent lower incidence of dementia among the obese group was due in large part to mortality preceding dementia development. The average age at death for the obese group was significantly younger than the average age at death for the non-obese group (79.3±8.7 for obese vs. 82.9±9.3 for non-obese, P<0.0001). Further, the cohort size for the obese group displayed a dramatic drop relative to loss in the non-obese group (a 27% decline vs. 17% in the non-obese group) specifically between ages 60 and 70 – a decade in which cancer and cardiovascular disease account for more deaths than all other causes combined.

In summary, I wouldn’t put much stock in the idea that obesity offers blanket protection against dementia – it’s equivalent to saying that death during childbirth is protective against postmenopausal heart attacks. Yet although death before dementia might largely explain the reduced dementia frequency in the obese cohort overall, it comes up short when we consider reduced dementia frequency in the obese cohort of APOE4 carriers.

APOE4 carriers demonstrated the greatest apparent obesity-associated protection from dementia, yet if pre-dementia mortality were fully responsible for the evident positive effect of obesity, we would expect APOE4 carriers to be least protected. These individuals are more likely to develop cognitive impairment at younger ages than those without the ε4 allele, lowering the likelihood that mortality from other causes would occur prior to dementia onset. These facts imply that, at least for APOE4 carriers, there may be more to the apparent inverse relationship between obesity and dementia.

What might explain the “protective” effect of obesity in APOE4 carriers?

Recall that apolipoprotein E – the protein encoded by the APOE gene – is vital for lipid transport to neurons. The form of apolipoprotein E encoded by the ε4 allele is somewhat impaired in this function, transferring cholesterol less efficiently than the forms of the protein encoded by the ε2 or ε3 alleles. This impairment is believed to be the mechanism by which the APOE-ε4 allele elevates dementia risk, and because it relates to lipid homeostasis, it’s tempting to draw connections to obesity.

While obesity certainly correlates with dyslipidemia in the periphery, it’s less clear that it has any implications for lipid homeostasis in the brain. Astrocytes produce cholesterol on their own, and plasma lipid levels have no influence on in-house lipid synthesis and transport in the brain. So although a connection between obesity and brain cholesterol homeostasis might exist, we don’t currently have scientific evidence of such a link.

A more likely explanation for the apparent “protective” effect of obesity among APOE4 carriers is reverse causality, which the authors themselves note as a possibility. Dementia tends to reduce food intake and cause reductions in body weight and BMI, typically starting in the preclinical stage of cognitive decline. So those with dementia are likely to have lower body weights, on average, than those without dementia. Some have speculated that the psychiatric symptoms of cognitive impairment – such as depression – may be a direct cause of appetite loss, though it’s also conceivable that declines in food intake and in cognition represent two outcomes of a separate, shared cause.

The possibility of reverse causality would also explain why the inverse relationship between BMI and dementia frequency was strongest among APOE4 carriers. Consider all participants who had a BMI over 30 prior to the study and subsequently lost weight as a result of incipient dementia. Those who began this decline in cognition and body weight after the start of the study would have been excluded from analysis by virtue of their transition from the “obese” classification to the “non-obese” classification – an exclusion which, regardless of APOE genotype, biases results toward underrepresentation of dementia rates among subjects in the obese group. But now consider those who made this BMI transition prior to the start of the study – but who hadn’t yet reached a level of cognitive impairment as to be clinically detectable. These individuals would remain included in the analysis cohort, thus overrepresenting dementia rates among the group classified as non-obese. Because APOE4 carriers are more likely than those with other APOE genotypes to develop dementia at an early age, a higher percentage of APOE4 carriers likely started the study in this “false non-obese” state of preclinical cognitive decline.

Bottom Line

One of the great challenges in understanding the causes of disease – and therefore in developing effective treatment and prevention strategies – is the fact that the deadliest maladies rarely exist in isolation, and their biological underpinnings interact in intricate ways. Shinohara et al. conducted their study based on the premise that the existence of obesity affects risk of dementia in a more complex manner than previously suggested, with APOE genotype playing a key role in defining the relationship between the two conditions. The study certainly has its share of limitations, and their results offer little in terms of unequivocal conclusions, but the basic finding that the relationship between obesity and dementia risk changes with APOE status nevertheless serves as fodder for future hypotheses and investigations. 

Descriptive, observational studies such as this serve an important function in helping to generate new questions, which more robust research or complementary lines of investigation can (hopefully) address in the future. How does APOE expression change as a function of body weight? Do statins or other lipid-lowering medications alter dementia risk differently across APOE genotypes? More mechanistic insights as to the reason for this unexpected relationship might have enormous implications for therapeutic approaches to dementia based on APOE genotype, underscoring the importance of deciphering the nuance in disease pathogenesis.


For a list of all previous weekly emails, click here

podcast | website | ama

Disclaimer: This blog is for general informational purposes only and does not constitute the practice of medicine, nursing or other professional health care services, including the giving of medical advice, and no doctor/patient relationship is formed. The use of information on this blog or materials linked from this blog is at the user's own risk. The content of this blog is not intended to be a substitute for professional medical advice, diagnosis, or treatment. Users should not disregard, or delay in obtaining, medical advice for any medical condition they may have, and should seek the assistance of their health care professionals for any such conditions.
Facebook icon Twitter icon Instagram icon Pinterest icon Google+ icon YouTube icon LinkedIn icon Contact icon