Patients classified as metabolically unhealthy generally had higher dementia risk estimates regardless of obesity status, whereas metabolically healthy obesity was not associated with increased dementia risk in primary analyses, according to a prospective cohort study published in BMC Medicine.
Researchers analyzed data from 11,482 patients in the US Health and Retirement Study and 13,068 patients in the Swedish Twin Registry. All patients were free of dementia at baseline. Patients were classified as metabolically healthy without obesity, metabolically healthy with obesity, metabolically unhealthy without obesity, or metabolically unhealthy with obesity.
Obesity was defined as a body mass index of at least 30 kg/m². In the Health and Retirement Study, metabolically unhealthy status was defined as having at least 2 of 3 criteria: hyperglycemia, hypertension, or dyslipidemia based on high-density lipoprotein cholesterol. In the Swedish Twin Registry, the definition used at least 2 of 4 criteria, with triglycerides also included.
Investigators evaluated dementia risk separately for measurements obtained in midlife, defined as age 65 years or younger, and late life, defined as older than 65 years. Cox proportional hazards models were adjusted for age, sex, smoking status, and education level.
Mean follow-up ranged from approximately 8 to 17 years across cohorts and age groups. In the Health and Retirement Study, 9% of patients in the midlife sample and 24% in the late-life sample developed dementia. In the Swedish Twin Registry, corresponding rates were 6% and 23%.
Most primary associations did not reach statistical significance. Among women in the Health and Retirement Study, however, metabolically unhealthy status without obesity during midlife was associated with a 62% higher adjusted hazard of dementia compared with metabolically healthy status without obesity.
In the Swedish Twin Registry, metabolically unhealthy status without obesity in late life was associated with a 13% higher adjusted hazard of dementia in the overall sample and a 22% higher adjusted hazard among men compared with metabolically healthy status without obesity.
Across both cohorts, metabolically unhealthy status without obesity showed the clearest pattern of higher dementia risk estimates. Findings for metabolically unhealthy obesity were less consistent, particularly in late life, where estimates suggested higher dementia risk in the Health and Retirement Study but lower dementia risk in the Swedish Twin Registry.
By contrast, metabolically healthy obesity was not associated with increased dementia risk in either age group in the primary analyses. Estimates generally suggested no association or lower dementia risk compared with metabolically healthy patients without obesity. In late-life analyses, metabolically healthy obesity was associated with lower dementia risk estimates in both cohorts, but these findings were not statistically significant.
Secondary analyses examining obesity and metabolic health separately supported the same general pattern. In midlife, metabolically unhealthy status was associated with higher dementia risk estimates in both cohorts, but the associations were not statistically significant. In late life, metabolically unhealthy status was associated with higher dementia risk in both cohorts. Obesity alone showed little evidence of increased dementia risk; in late life, obesity was associated with lower dementia risk in the Swedish Twin Registry but not in the Health and Retirement Study.
Sensitivity analyses underscored the complexity of the findings. Results varied when researchers applied alternative definitions of metabolic health, reflecting the lack of consensus on how metabolically healthy obesity should be defined. Under a stricter definition, estimates for metabolically healthy obesity shifted toward higher dementia risk, although these findings were not statistically significant and relatively few patients were classified as metabolically healthy.
When researchers accounted for death as a competing event, late-life associations between metabolically unhealthy status without obesity and dementia were attenuated. Late-life metabolically healthy obesity and metabolically unhealthy obesity were associated with lower dementia risk estimates in competing-risk models. The investigators noted that these findings may reflect selective survival and competing mortality, in which patients with obesity or poorer metabolic health may die of other causes before dementia is diagnosed.
The study has several limitations. Dementia ascertainment differed between cohorts. The Health and Retirement Study used the Langa-Weir cognitive classification rather than clinical dementia diagnoses, while the Swedish Twin Registry used national health registers, dementia medication records, and in-person assessments in some substudy cohorts. The researchers noted that Swedish register-based dementia diagnoses have high specificity but moderate sensitivity, meaning some dementia cases may have been missed. The study also assessed all-cause dementia and could not evaluate dementia subtypes in both cohorts.
Triglyceride measurements and cholesterol-lowering medication data were unavailable in the Health and Retirement Study, which may have affected metabolic-health classification. Follow-up duration may also have been insufficient to fully account for the long preclinical phase of dementia, during which unintentional weight loss can occur before diagnosis. The observational design precludes conclusions about causality.
The findings add nuance to prior evidence on obesity, metabolic health, and dementia risk. The researchers noted that their findings were broadly consistent with a recent meta-analysis but contrasted with some prior midlife cohorts, including Whitehall II and analyses of the UK Biobank, Atherosclerosis Risk in Communities study, and Framingham Offspring Study, in which metabolically healthy obesity showed higher dementia risk estimates. They suggested that differences across studies may reflect variation in age at measurement, follow-up time, metabolic-health definitions, and transitions between metabolically healthy and unhealthy states over time.
Overall, the findings suggest that metabolic health status may be associated with dementia risk independently of obesity status, but the results should be interpreted cautiously because most primary associations were not statistically significant and several findings varied across sensitivity analyses.
“Being metabolically unhealthy, especially in midlife, may be associated with increased dementia risk, regardless of obesity status. Mid- and late-life [metabolically healthy obesity] showed no increased risk and suggested potential inverse associations,” wrote lead study author Martin Nakash, of Karolinska Institutet, and colleagues.
Disclosures: The researchers reported no competing interests.
Source: BMC Medicine
