Abstract Body: Background: In people with type 2 diabetes (T2D), weight loss due to an intensive lifestyle intervention (ILI) reduces triglyceride (TG) levels, a risk factor for cardiovascular disease. However, efficacy of treatment varies between people. Both testosterone and estradiol are associated with TG level. We examined whether baseline androgenic state, i.e., the ratio of testosterone to estradiol (T/E), modifies the treatment effect and partially explains the TG variability.
Methods: The Look AHEAD Study was a randomized control trial of 5145 individuals with T2D and overweight/obesity to evaluate the effect of ILI compared to the control group on cardiovascular outcomes. We selected a random sample of 1166 postmenopausal females and 1166 males (mean age 60 years [SD 6.2]) to examine heterogeneity of treatment effect by baseline T/E ratio on changes in TG at years 1 (weight loss nadir) and 4. We created longitudinal mixed models stratified by sex, and modeled the interaction between log-transformed baseline T/E ratio, ILI (vs control), and year, with log-transformed TG as the outcome. We plotted the estimated treatment effects over time for 25^th and 75^th percentile of baseline T/E ratio.
Results: Figure, panel A, shows that males with T/E ratio at the 75^th percentile (vs 25^th percentile) had a larger reduction in TG due to ILI (vs control) at year 1 (p = 0.002), which is no longer statistically significant at year 4 (p = 0.14, overall heterogeneity for both years, p = 0.009). We did not detect any statistically significant heterogeneity in TG levels by T/E ratio for females (panel B, p = 0.87).
Conclusion: Greater baseline androgenic state (T/E ratio) in males, but not females, was associated with a larger reduction in TG due to ILI, compared to the control at year 1 but not year 4. Males with lower T/E ratio might need more aggressive lipid treatment to achieve TG-reducing benefits of ILI.