Analysis of European data offers no support for obesity paradox in cognition

The “obesity paradox”—a few more pounds being linked with positive health outcomes—did not hold up in terms of better cognition in older adults, researchers in Germany reported.

Based on an analysis of longitudinal data from the Survey of Health, Ageing, and Retirement in Europe (SHARE), there was a positive association between BMI change and change in cognitive performance that was “dominantly driven by BMI decrease,” according to Judith M. Kronschnabl, MA, of the Munich Center for the Economics of Aging, and co-authors.

“Weight gain was, on average, not significantly related to cognitive performance; only respondents with preceding weight loss profited from small increases in BMI,” they wrote in PLoS ONE. “Our analyses provide no support for an ’obesity paradox’ in cognition… The association between weight change and cognitive performance in older age is based on weight changes being related to illness and recovery.”

A brief timeline of recent research on the potential links between obesity and cognition shows a fluctuating connection. “The complex associations between late-life adiposity and physical and cognitive health have often been explained in the context of the obesity paradox [which] describes an association between excess adiposity and favorable health outcomes,” explained Consuelo H. Wilkins, MD, of Vanderbilt University Medical Center in Nashville, and co-authors in Ethnicity & Disease.

While in their 2017 study, Wilkins’ group found that obesity was tied to better executive function, they also cautioned that “obesity paradox remains controversial despite mounting evidence of its existence.”

In 2019, researchers looked at elevated BMI, cognitive dysfunction, and heart failure (HF) to see if the first two factors interact to predict clinical HF endpoints and reported that “lower BMI demonstrated a negative association between cognitive performance and hospitalizations, such that lower cognitive function, is associated with a greater number of hospitalization in the prior six months.”

However, they also warned that it was not clear if “the interaction of BMI and cognitive dysfunction influences risk for hospitalization, or if hospitalization influences subsequent cognitive function, depending on BMI.”

Years of data from Italy’s Mugello study, which investigated the association between BMI and functional cognition in people in their 90s, reported better cognition for those with higher BMI, although the Mugello investigators stopped short of suggesting possible mechanisms of action.

More recently, researchers in Canada took a stab at providing that answer and evaluated whether physical fitness moderated the relationship between leptin and adiposity, as “lower leptin independent of adiposity may be confounding the obesity paradox.” They said that “the link between obesity and mortality may not be dependent on total adiposity, but rather on endocrine function and adipocyte leptin secretion.”

Finally, in a 2020 review article, Lorenzo M. Donini, MD, PhD, of Sapienza University in Rome, and co-authors pointed out that the “actual paradox seems to [keep] defining obesity using BMI,” and that is problematic, as BMI is “not able to quantify body fat percentage and adiposity distribution, nor the degree of metabolic disturbances that it can underlie.”

They called for “correct phenotyping of patients” that takes into account body fat, lean mass, metabolic and functional variables, duration of obesity and normal weight would better “capture… the trajectories of mortality in a wealth of diseases…The obesity paradox is just a lesson to be learned.”

Kronschnabl’s group looked at four waves of SHARE data, between 2006 and 2016, with 58,389 participants from 15 countries, all ages ≥50. “We estimated asymmetric fixed effects models by gender, adding possible confounding variables such as age, grip strength, health conditions, and physical activity. Additionally, we investigated possible heterogeneity in the BMI-cognition relation,” they explained.

They found that for women and men, “grip strength and observed diseases accounted for a sizeable and significant part of the association of weight loss and cognitive decline—over and above age, which might have already partially absorbed any (unobserved) age-related health deterioration, and physical activity.”

Also, to answer the question, “Is weight gain beneficial for the oldest old?,” the authors divided participants into two groups: One with a mean age of about 76 and the other with a mean age of 60. They highlighted several noteworthy findings:

  • Estimates from the models exploiting between-variation (pooled sample and between-individuals) were significantly higher in the older group versus the younger group.
  • The negative association of BMI and cognition found for women in the full sample could not be replicated for the older group.
  • Estimates from models exploiting within-variation only, including fixed-effects individual slopes, were more pronounced, but not statistically significant, in the older group versus the younger group.
  • For the confounding effects of grip strength and illness, “patterns largely prevailed but were again more pronounced among the older, except for physical activity, which appeared to be more important for the younger.”

Finally, they assessed if weight gain might be meaningful in terms of cognition for people with an earlier episode of weight less and found a “positive interaction for women: weight gain without preceding weight loss had no effect on cognitive performance but became beneficial at relatively high levels of preceding weight loss.”

Study limitations included the cross-sectional nature of the data and the mixed findings from different countries, both of which contributed to heterogeneity. Echoing Donini’s group, the authors called for “[m]ore precise measures of body composition” such as bioelectric impedance analysis (BIA) or dual energy x-ray absorptiometry,” although they conceded those tools may be difficult to use in a large-scale study. However, “waist-to-hip ratio could be employed as a complementary measure of abdominal fatness,” they wrote.

  1. Analysis of data from a large, long-term European survey providds no support for an “obesity paradox” in cognition in older age, with the association between weight change and cognitive performance based on weight changes being related to illness and recovery.

  2. There was a positive association between BMI change and change in cognitive performance, mostly because of BMI decrease.

Shalmali Pal, Contributing Writer, BreakingMED™

SHARE data collection is supported by the European Commission/FP5, the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, and the National Institute on Aging.

Kronschnabl and co-authors reported no relationships relevant to the contents of this paper to disclose.

Cat ID: 130

Topic ID: 82,130,404,494,730,130,361,192,255,925