ASSOCIATION OF DIETARY FACTORS AND SELECTED PLASMA VARIABLES WITH SEXHORMONE-BINDING GLOBULIN IN RURAL CHINESE WOMEN

Sex hormone-binding globulin (SHBG) is an important regulator of plasm a sex steroids as well as a sensitive indicator of insulin resistance. SHBG may be an important diagnostic measure of risk for pathologies a ssociated with insulin resistance syndrome (IRS) such as non-insulin-d ependent diabetes mellitus (NIDDM), obesity, hypertension, dyslipidemi a. and atherosclerotic cardiovascular disease. In women, SHBG is also im plicated in diverse pathologies such as cancers of steroid-sensitiv e tissues and hirsutism. Data from an ongoing ecological study linking diet and health in rural China were analyzed to determine the relatio n of selected plasma variables and diet to plasma concentrations of SH BG. All data represent county mean values, pooled by age and sex, to a ssess the relation between biochemical and lifestyle characteristics a nd disease-specific mortality rates at the county level. The study sam ple consisted of 3250 Chinese women between the ages of 35 and 64 y li ving in 65 widely dispersed rural counties. Consumption patterns for 2 1 different food groups were derived from a food-frequency questionnai re and a 3-d dietary survey and subsequently compared. Correlation ana lyses of county mean values demonstrated a significant association bet ween SHBG and insulin, testosterone, triacylglycerols, body mass index , age at menarche, and several foods. In regression analyses, after ad justments, the strongest predictors of SHBG concentrations were the di etary intakes of rice (beta = 0.42, P < 0.01), fish (beta = 0.34, P < 0.05), millet (beta = -0.27, P < 0.01), and wheat (beta = -0.34, P < 0 .01). When insulin, testosterone, and triacylglycerols were added to t he model only triacylglycerols (beta = -0.26, P < 0.05) remained a sig nificant independent predictor of SHBG. Additional analyses suggested that the consumption of green vegetables was modestly positively corre lated with SHBG and negatively with insulin values. Consumption of ric e and fish in particular appeared to favorably influence the principle plasma variables associated with a reduction in the risk for IRS path ologies.