Objective

Our objective was to investigate cross-sectional and longitudinal associations of sex hormone concentrations with ankle-brachial index (ABI) and peripheral arterial disease (PAD).

Methods and results

We used data from 3034 (1612 women) participants of the Framingham Heart Study. ABI was measured and PAD defined as ABI below 0.90, intermittent claudication, or lower extremity revascularization. Sex hormone concentrations were measured by liquid chromatography-tandem mass spectrometry [total testosterone (T), total estradiol, and estrone], immunofluorometric assay (SHBG), or calculated (free T). Sex-specific multivariable linear and logistic regression models were conducted for each sex hormone separately. Cross-sectional multivariable analyses revealed that men with lower free T and higher estrone (E1) concentrations had a significantly lower ABI [for free T, lowest vs. higher quartiles, β = -0.02, with 95% confidence interval (CI) = -0.04 to -0.001; and for E1, highest vs. lower quartiles, β = -0.02, with 95% CI = -0.04 to -0.002, respectively). Lower total T and SHBG concentrations were also associated with prevalent PAD in age-adjusted [odds ratio (OR) = 2.24, 95% CI = 1.17-4.32; and OR = 2.06; 95% CI = 1.07-3.96, lowest vs. highest quartile, respectively), but not in multivariable logistic regression models. Longitudinal multivariable analyses showed an association of lower SHBG with ABI change (decline ≥ 0.15; n = 69) in men [OR for SHBG quartiles 1, 2, and 3 as compared with quartile 4 were 2.56 (95% CI = 1.01-6.45), 2.28 (95% CI = 0.98-5.32), and 2.93 (95% CI = 1.31-6.52), respectively]. In women, none of the investigated associations yielded statistically significant estimates.

Conclusion

Our investigation of a middle-aged community-based sample suggests that sex hormone concentrations in men but not in women may be associated with PAD and ABI change.