THE EFFECT OF GENETIC AL AND ENVIRONMENTAL FACTORS ON LIPIDS: A TWIN STUDY

To assess the relative contribution of genetic and environmental factors (particularly androgens) on circulating levels of lipid fractions, serum androgen and lipid fractions were measured in 34 pairs of female-female twins aged from 15- 45 years, some of whom were discordant for polycystic ovary syndrome (PCOS) diagnosed by ultrasound. Nineteen pairs were monozygotic twins (MZ) and 15 pairs were dizygotic twins (DZ). Five pairs of MZ and 6 pairs of DZ were discordant for scan-PC a. We measured serum concentrations of total cholesterol (TC), high density lipoprotein cholesterol (HDL-C), triglycerides (TRIG), lipoprotein (a) [LP(a)], and apolipoprotein B (apo B). Also, testosterone (T), dehydroepiandrosterone sulphate (DHEAS), sex hormone binding globulin (SHBG) and 3a-androstanediol glucuronide (3a-diol G) levels were measured. Transabdominal ultrasound was performed. Serum levels of TC, HDL-C, TRIG, LP(a) and apo B in the twins with PCO were not significantly different from the levels in their matched co-twins with normal ovaries. There were no significant correlations between androgen-hormones including T, DHEAS and 3a-diol G with any of the lipid measurements. Body mass index (BMI) was positively correlated with TRIG, LP(a) (both p<0.05) and negatively correlated with HDLC (p<0.001). SHBG was negatively correlated with TRIG and LP(a) and positively associated with HDL-C (p<0.05). Insulin was significantly correlated with TRIG (p<0.001) and negatively with HDL-C (p<0.01). The MZ intraclass correlation exceeded that of the DZ for all the lipid variables measured. The heritability estimates for LP(a), apo B, TC and HDL-C were 0.95, 0.56, 0.48 and 0.54, respectively. However, the intraclass correlation coefficient for TRIG was not significantly different between MZ and DZ but maximum likelihood analysis indicated that at least 10% of the variance of circulating TRIG concentration is determined by genetic factors. We conclude that twins discordant for pcas do not have significantly different lipid profiles. We were unable to show any significant effect of androgens on the lipid fractions measured. The results confirm that levels of LP(a), HDL-C, TC and apo B are under significant genetic control and that this is particularly so for LP(a). However, only 10% of the variation in TRIG levels could be attributed to genetic influences after controlling for age and obesity. Increase in BMI and insulin had a significant adverse effect on the lipid profile in these female twins, effects which may enhance coronary risk.