Obesity Linked to Dementia Risk
Gray matter atrophy tied to BMI and other obesity metrics.
Lower gray matter volume was tied to higher levels of all obesity measures, a cross-sectional study in Great Britain found.
Body mass index (BMI), waist-hip ratio, and fat mass were all related to low gray matter volume, and the combination of overall obesity (BMI ≥30) and central obesity (waist-hip ratio >0.85 for women; >0.90 for men) was tied to the greatest risk of gray matter atrophy, reported Mark Hamer, PhD, of Loughborough University in Leicestershire, England, and G. David Batty, DSc, of University College London.
Obesity was also related to smaller putamen, caudate, pallidal, and nucleus accumbens volumes, they wrote in Neurology. There were no significant differences in white matter brain volume.
“Existing research has linked brain shrinkage to memory decline and a higher risk of dementia, but research on whether extra body fat is protective or detrimental to brain size has been inconclusive,” Hamer said in a statement. “While our study found obesity, especially around the middle, was associated with lower gray matter brain volumes, it’s unclear if abnormalities in brain structure lead to obesity or if obesity leads to these changes in the brain.”
These findings conflict with a large-scale study that showed obesity was tied to lower dementia rates, although other studies have shown different data. That weight loss is common in the preclinical phase of dementia — up to a decade before diagnosis — may help explain these seemingly paradoxical findings, Hamer and Batty noted.
In this study, the researchers looked at 9,652 individuals (average age about 55; 47.9% men) who were assessed from 2014 to 2016 as part of the U.K. Biobank study. Measures included BMI, waist-hip ratio, and total fat mass assessed by bioimpedance; brain images were produced with structural MRI.
The researchers determined 18.7% of the sample to be obese. These people were less likely to have a college degree and to be physically active, and had a higher prevalence of hypertension (31.7%), diabetes (7.2%), and heart disease (4.8%).
After adjusting for covariates including age, sex, physical activity, smoking, hypertension, education, major depression, alcohol, and heart disease, people (n=1,291) who had a high BMI and central obesity had the lowest average gray matter brain volume of 786 cm3. In contrast, people of healthy weight (n=3,025) had an average gray matter brain volume of 798 cm3, and people with a high BMI without central obesity (n=514) had an average gray matter brain volume of 793 cm3.
Higher levels of all obesity measures were related to lower gray matter volume. Beta coefficient per 1 SD increase was:
- BMI: -4,113 (95% CI -4,862 to -3,364)
- Waist-hip ratio: -4,272 (95% CI -5,280 to -3,264)
- Fat mass: -4,590 (95% CI -5,386 to -3,793)
Participants with diabetes had lower gray matter volume than those without the condition. When diabetes was added to the models, associations between BMI and gray matter, and between waist-hip ratio and gray matter, were only partly attenuated, by about 6% to 8%.
Obesity was associated with a lower volume of the caudate (only waist-hip ratio), putamen (only BMI and total fat mass), pallidum, and nucleus accumbens regions (P<0.001).
“This is an important and exciting paper,” said Sudha Seshadri, MD, of University of Texas Health Sciences Center in San Antonio, who was not involved with the study.
“Although it is simple in design, it adds data about the potential huge impact of obesity on brain structure, and thus risk of age-related brain diseases such as dementia. That the effect was seen in persons who were only 55 years old on average makes it more important, and reiterates that it is not too early to think of dementia prevention in middle-aged adults,” she told MedPage Today.
The study extends earlier work done in the U.S. with the Framingham Offspring cohort by looking at cortical gray matter volume, not just total cerebral volume, and assessing subcortical structures that control mood, motivation, and movement, Seshadri observed.
“They did not find BMI or waist-hip ratio associated with white matter volumes, but in white matter the early changes are typically in microstructure — such as changes in fractional anisotropy or free water on diffusion tensor imaging — or white matter hyperintensity volumes and these were not looked at,” she said.
Other study limitations included possible selection bias, race-ethnic differences were not explored, and risk factors such as hypertension and heart disease were self-reported, Seshadri added. Previous research has shown that people with both hypertension and high waist-hip ratio had lowest cognitive function, and some differences in gray matter volume could be due to that, she said.
Only about 5% of the target population took part in U.K. Biobank, and study members typically showed more favorable risk profiles than non-responders, noted Hamer and Batty. And, as in any observational study, residual confounding may have occurred, they added.