Epidemiology/Health Services Research ORIGINAL ARTICLE Circulating Estrone Levels Are Associated Prospectively With Diabetes Risk in Men of the Framingham Heart Study 1,2 6 GUNEET KAUR JASUJA, PHD ALAN L. ROCKWOOD, PHD Epidemiologic studies (6,7) and ran- 3,4 6 THOMAS G. TRAVISON, PHD WAYNE MEIKLE, MD – 3 3,7 domized trials (8 10) in women have MAITHILI DAVDA, MPH ANDREA D. COVIELLO, MD, MS 2,5 1, 2 suggested that hormone therapy reduces ADAM J. ROSE, MD RALPH D’AGOSTINO, PHD 3 7,8 the risk of T2DM in postmenopausal ANQI ZHANG, PHD RAMACHANDRAN S. VASAN, MD 6 3 women. Furthermore, genetic disruption MARK M. KUSHNIR, PHD SHALENDER BHASIN, MD of estrogen receptor a (ERa)inmiceis associated with adiposity and insulin re- sistance (11). Only a few cross-sectional OBJECTIVEdIn postmenopausal women and preclinical murine models, estrogen adminis- tration reduces diabetes risk; however, the relationship of estradiol and estrone to diabetes in studies in older men have addressed the men is poorly understood. We determined the relationship between circulating estradiol and relationships between estradiol and estrone levels and diabetes risk in community-dwelling men of the Framingham Heart Study T2DM, and the data are conflicting; (FHS). some studies have shown a positive cor- relation of estradiol levels with T2DM METHODSdCross-sectional relationships of estradiol and estrone levels with diabetes were – (12,13), whereas others have found no assessed at examination 7 (1998 2001) in FHS generation 2 men (n = 1,458); prospective significant association (5,14). The rela- associations between hormone levels at examination 7 and incident diabetes were assessed 6.8 tionship between estrone and T2DM has years later at examination 8. Type 2 diabetes mellitus was defined as fasting glucose .125 mg/dL, medication use, or both. Estradiol, estrone, and testosterone levels were measured with liquid not been studied in men. Most studies chromatography tandem mass spectrometry, and free estradiol and estrone were calculated. used immunoassays for the measurement of estradiol levels, for which accuracy in RESULTSdIn cross-sectional models, men with elevated estrone and estradiol had 40% and the low range has been questioned (15– 62% increased likelihoods of existing diabetes per cross-sectional doubling of estrone and es- 17). – tradiol levels, respectively. Free estrone (cross-sectional odds ratio 1.28 [95% CI 1.02 1.62], P = By using data from the Framingham 0.04) was associated with impaired fasting glucose at examination 7. There was an increase in risk Offspring Study, we determined whether of existing diabetes with increasing quartiles of total and free estrone and estradiol and an in- circulating estrone and estradiol levels crease in risk of incident diabetes with increasing quartiles of estrone levels. In multivariate longitudinal analyses, a twofold increase in total or free estrone levels at examination 7 are associated with T2DM or IFG in was associated with 77 and 93% increases, respectively, in odds of incident diabetes at community-dwelling older men. In longi- examination 8. tudinal analyses restricted to nondiabetic men, we evaluated whether these hormones CONCLUSIONSdAlthough both estradiol and estrone exhibit cross-sectional associations were predictive of incident T2DM during a with diabetes in men, in longitudinal analyses estrone is a more sensitive marker of diabetes risk follow-up period of approximately 7 years. than estradiol. This analysis is among the first population- based assessments of the association between estradiol and estronedhere measured with liquid chromatography ging is associated with a decline in in the development of T2DM. Age-related tandem mass spectrometry (LC-MS/ Aglucose tolerance, resulting in higher decline in testosterone levels has been as- MS), widely considered the reference prevalence of type 2 diabetes mellitus sociated with an increased risk of T2DM method with the highest specificity and (T2DM) and impaired fasting glucose (IFG) in older men (2–5); however, the effects sensitivitydwith T2DM risk in men (18). in older adults (1). Previous studies have sug- of low or high estrone and estradiol levels gested a role of endogenous sex hormones on T2DM risk in men are not clear. RESEARCH DESIGN AND ccccccccccccccccccccccccccccccccccccccccccccccccc METHODS From the 1Department of Mathematics, Boston University, Boston, Massachusetts; the 2Center for Health Quality, Outcomes, and Economic Research, Bedford VA Medical Center, Bedford, Massachusetts; the Study sample 3Section of Endocrinology, Diabetes, and Nutrition, Boston University School of Medicine, Boston, Mas- The Framingham Heart Study (FHS) de- sachusetts; the 4Department of Biostatistics, Boston University School of Public Health, Boston, Massa- 5 sign and methods have been described chusetts; the Section of General Internal Medicine, Boston University School of Medicine, Boston, (19). Briefly, the original cohort was re- Massachusetts; 6ARUP Laboratories University of Utah, Salt Lake City, Utah; and the 7Section of Pre- ventative Medicine and Epidemiology, Boston University School of Medicine, Boston, Massachusetts; and cruited from Framingham, Massachu- 8NHLBI’s Framingham Heart Study (FHS), Framingham, Massachusetts. setts, in 1948 to identify risk factors for Corresponding author: Guneet Kaur Jasuja, [email protected]. cardiovascular disease. In 1971, the study Received 28 November 2012 and accepted 8 February 2013. enrolled a second-generation cohort (Gen DOI: 10.2337/dc12-2477 d ’ © 2013 by the American Diabetes Association. Readers may use this article as long as the work is properly 2) 5,124 of the original participants cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/ adult children and their spouses. The licenses/by-nc-nd/3.0/ for details. men of this Framingham Offspring Study care.diabetesjournals.org DIABETES CARE 1 Diabetes Care Publish Ahead of Print, published online May 20, 2013 Estrone and estradiol and T2DM risk in men cohort who attended examination 7 offspring examination 5 in 1991–1995, Unadjusted estimates of the relative (1998–2001) were eligible for the current the stability of these FHS samples in stor- risk quantifying the cross-sectional rela- study (n = 1,625). Men with missing es- age was evaluated by measuring the tionship between hormones, divided into trone and estradiol measurements (n = concentrations of cholesterol, HDL cho- quartiles, and outcomes were generated 159), those with prostate cancer undergo- lesterol, and triglycerides before freezing with the modified Poisson regression ing androgen deprivation therapy (n =5), and storage at 2808C and then repeating approach, which uses the robust variance and those with missing diabetes data at the measurement in 2007 (20). estimator to avoid bias in interval estima- examination 7 (n = 3) were excluded, re- Serum estradiol and estrone were tion and corresponding significance tests sulting in a sample size of 1,458 for the measured with a highly sensitive LC-MS/ (25). Covariate-adjusted cross-sectional cross-sectional analyses. MS assay. Derivatization of estrone and associations were analyzed with separate For analyses of incident T2DM, the estradiol was performed with dansyl chlo- polytymous logistic regression models for subset of men who attended examination ride. Estrone-d4 and estradiol-d5 (20 mL total and free estrone and estradiol. These 8 (2005–2008) were examined. The me- each) were added to 200-mLserumsam- models simultaneously assessed the odds dian time between examination 7 and 8 ples, extracted with methyl t-butyl ether of IFG and T2DM in comparison with assessments was 6.8 years. For this anal- (21,22), derivatized with dansyl chloride normal glucose levels. ysis, we excluded men who had existing (3.7 mmol/L) in sodium carbonate (10 The longitudinal associations be- T2DM at examination 7 (n = 226), those mmol/L, pH10.5) at 608C for 10 min, di- tween baseline hormone levels at exami- who did not attend examination 8 (be- lutedinacetonitrileandwater,andthe nation 7 with the cumulative incidence of cause of death or loss to follow-up), and samples were analyzed on API 4000 triple- T2DM at examination 8 were assessed those who lacked a T2DM assessment at quadrupole mass spectrometer (Applied with separate regression models for each examination 8 (n = 201). The longitudinal Biosystems/MDS Sciex) with turbo ion of the total and free hormones. In this analysis was therefore restricted to 1,031 spray HPLC pumps series 1200, and au- analysis, the men who had T2DM at men. tosampler HTC PAL (LEAP). The mobile examination 7 were excluded. Again, the phase and mass spectrometry method modified Poisson regression approach Ascertainment of outcomes in FHS used has been described previously (22). was used to determine unadjusted asso- Subjects were considered having T2DM if The limit of quantitation for both hor- ciations between hormone quartiles and their fasting glucose levels exceeded 125 mones was 2 pg/mL. Interassay coeffi- T2DM status. Multivariate models used mg/dL or they reported use of medication cients of variation (CVs) for estrone were multiple logistic regression. to control T2DM. Subjects were consid- 4.5, 7.7, and 6.9% at estrone concentra- Both cross-sectional and longitudinal ered to have normal glucose levels if they tions of 8, 77, and 209 pg/mL, respec- models considered the roles of age, BMI, had fasting blood glucose ,100 mg/dL tively; interassay CVs for estradiol were smoking, total testosterone, and SHBG without medication; they were consid- 6.9, 7.0, and 4.8% at estradiol concentra- (for total hormone levels). ered to have IFG if fasting blood glucose tions of 8, 77, and 206 pg/mL, respec- To enhance clarity, results on log- was between 100 and 125 mg/dL in the tively. transformed values were back trans- absence of T2DM treatment. A subject Free estradiol and estrone concentra- formed and thus may be interpreted in was deemed to have cardiovascular dis- tions were calculated from a previously terms of relative rather than absolute ease if he had coronary artery disease (an- published law of mass action solution differences in hormone values.