Associations Between Cognitive Function and Sex Hormones in the Look AHEAD Cohort

Introduction

Levels of estradiol (E2) and total testosterone (T) in individuals of both sexes decline as they age. Many studies have linked decreases in these sex hormones with increased risk for developing Alzheimer’s Disease (AD). As the prevalence of obesity and type 2 diabetes (T2DM) rises it is important to understand the effect of the two on cognitive health: both have been linked to increased rates of cognitive decline and dementia. Weight gain in men with obesity has been found to increase total testosterone levels, whereas, in obese women weight loss decreases endogenous E2 and T levels. It remains unclear how weight loss in individuals with T2DM and obesity may impact brain health as they age. Additionally, prior studies have provided evidence that higher E2 among older women with T2DM increases risk of dementia. To better understand the relationship that T2DM, obesity, weight change, aging, and sex hormones have with cognitive function we used the data from the Action for Health in Diabetes (Look AHEAD) randomized controlled clinical trial.

Hypothesis

We hypothesized that levels of E2 and T in both women and men are associated with cognitive function and may influence the legacy that participation in a 10-year intensive lifestyle intervention featuring weight loss has on cognitive function.

Methods

579 women and 417 men with T2DM, overweight, or obesity, and mean age of 69 years were randomly assigned to a 10-year intensive lifestyle intervention (ILI) focused on weight loss or a control condition of diabetes support and education (DSE). Participants were followed with standardized measure of body mass index (BMI) and risk factors. Those using postmenopausal hormone or androgen therapy were excluded. Serum E2 and T were assayed centrally twice, during Epoch 1 (8-13 years post-randomization) and Epoch 2 (14-18 years post-randomization), an average of 4 years apart. (Times 1 and 2, an average of 4 years apart) centrally. Standardized measures of cognitive function (attention, executive function, memory, verbal fluency, and composite) were administered by trained and certified staff. Bayesian multiple imputation was used for E2 and T levels that were below the assay’s detectable level. Mixed effects models were used to describe relationships that log-transformed sex hormone levels had with current BMI and with intervention assignment.

Conclusions

•Despite prior reports, no evidence that higher endogenous E2 is linked to cognitive deficits or older women with T2DM

•Higher T may be associated with better verbal fluency in older men with T2DM, but not other cognitive domains

•Random assignment to 10 years of ILI is associated with relative deficits in cognitive function in women, but nit men, however these do not appear to be linked to E2 or T

Discussion

E2 and T levels were higher among men than women. Among men, greater BMI was associated with higher E2 (p<0.001) but lower T (p<0.001). Higher T was associated with better verbal fluency; however, the remaining cognitive functions were not associated with sex hormones levels. Among women, greater BMI was associated with higher E2 (p<0.001) and higher T (p<0.001). None of the cognitive function test scores were associated with E2 or T, in women. The 10-year ILI did not result in a legacy of reduced BMI among women at either epoch. Among men, it had a modest legacy of lower BMI at Time 1 (p=0.01), which attenuated by Time 2 (p=0.08). ILI had a legacy of poorer cognitive functioning among women that was most evident for executive function (mean z-score difference: -0.146+0.072, p=0.01). This effect, however, could not be explained by differences in current levels of sex hormones. The ILI did not have a legacy effect on cognitive functioning in men.