Higher BMD, fewer fractures reported for women with PCOS at high risk for hyperandrogenism
Click Here to Manage Email Alerts
Women with a high genetic risk score for hyperandrogenism have a higher bone mineral density and a lower risk for fracture than those with a lower risk score, according to study results published in Bone.
“Genetic risk for polycystic ovary syndrome predisposes women to higher BMD and reduced risk of fractures,” Thozhukat Sathyapalan, MD, FRCP, chair in academic diabetes, endocrinology and metabolism at Hull York Medical School in the U.K., told Healio. “These findings support the current recommendations for not screening women with PCOS for osteoporosis and fractures.”
Sathyapalan and colleagues conducted a Mendelian randomization analysis to test the association between the genetic risk for hyperandrogenism with BMD and fractures. Data were obtained from 221,086 women (mean age, 56.7 years) and 187,816 men (mean age, 57.1 years) in the UK Biobank cohort who attended one of the 22 assessment centers between 2006 and 2010. A quantitative ultrasound of both heels was performed to assess mean BMD and BMD T-scores, with the heel with the lower BMD used in the analysis. Participants self-reported fractures over the past 5 years.
The single nucleotide polymorphisms identified in the study were obtained from the largest genome-wide association study for PCOS. Three novel loci and 11 previously reported loci with PCOS were tested for their association in women with PCOS to construct a genetic risk score. Linear regression analysis was performed with BMD as the dependent variable and the weighted genetic risk score, age, BMI and top five principal components as independent variables.
Of the cohort, 11% of women and 10% of men reported a fracture over the past 5 years. The top single nucleotide polymorphisms associated with BMD in women were rs11031005 near the FSHB gene, rs7563201 near the THADA gene and rs7864171 near the FANCC gene. The top single nucleotide polymorphisms associated with fracture were rs1784692 near the ZBTB16 gene and rs9696009 near the DENND1A gene.
Each one standard deviation increase in the genetic risk for hyperandrogenism was associated with a higher BMD (beta = 0.0007; P = .001) and a lower risk for fractures (OR = 0.97; 95% CI, 0.96-0.99; P = .003) in women. There were no associations with BMD or fracture risk for men at high genetic risk for hyperandrogenism.
“There are already some observational studies that show excess BMD in women with PCOS,” Sathyapalan said. “However, our study establishes causality as it uses the Mendelian randomization approach.”
In sensitivity analysis using effect estimates for PCOS rather than testosterone levels, higher BMD (P = .07) and a reduced risk for fractures (OR = 0.98; 95% CI, 0.96-0.99; P = .007) remained for women with a high genetic risk for PCOS. Genetic risk score was not associated with BMI in women or men.
“This to date is the most extensive Mendelian randomization study looking at BMD and risk of fractures in women with a genetic predisposition to PCOS, whilst also providing some preliminary results in men,” the researchers wrote. “Like all the Mendelian randomization studies, our study is subject to weak instrument bias. Another limitation of the study is that we included all self-reported all-cause fractures for the analysis to ensure adequate power. Further studies will be needed to ascertain if the protective effect of the genetic risk of hyperandrogenism in women living with PCOS is restricted to different types of traumatic and nontraumatic fragility fractures.”
For more information:
Sathyapalan Thozhukat, MD, FRCP, can be reached at thozhukat.sathyapalan@hyms.ac.uk.
Perspective
Back to Top
This is the first report using Mendelian randomization to dissect the relationship between polycystic ovary syndrome, bone mineral density and fracture risk, conducted in a very large sample size. However, the selection of a very specific risk factor, “excess testosterone in PCOS,” raises important analytic limitations.
To be valid, Mendelian randomization analyses must meet three criteria. First, the instrument variables — single nucleotide polymorphisms — are a strong representation of the risk factor.Second, the single nucleotide polymorphisms must not be associated with potential confounders of the risk factor-outcome association. Finally, the single nucleotide polymorphisms affect the outcome by acting only through the risk factor.
Although this study includes sensitivity analyses addressing the latter two criteria, the first was not addressed in detail, other than an admission that the study is subject to weak instrument bias. This is indeed the case. While the 14 single nucleotide polymorphisms used in this study are robust genome-wide association study signals for PCOS, they were only weakly associated with “clinical and biochemical hyperandrogenism,” with P values ranging from 0.5 to 3 x 10-7, only four of which were deemed significant in the source report. Therefore, the risk factor — hyperandrogenism in PCOS — is poorly represented by the genetic risk score based on these single nucleotide polymorphisms. As such, the single nucleotide polymorphisms are a better representation of PCOS itself, as evidenced by similar effects on BMD and fracture observed when the authors used effect estimates for PCOS.
Another caveat is that the study suggests that the genetic risk score represents testosterone levels in PCOS. Rather, they represent clinical and biochemical hyperandrogenism, which includes hirsutism and elevated circulating androgen levels. Instrument variables are best applied in Mendelian randomization analyses to represent a single risk factor, rather than a heterogeneous mixture of features. That these phenotypes were assessed in women further weakens the instrument strength in men. Had the study utilized robust single nucleotide polymorphisms for testosterone, a positive signal for increased testosterone to increase BMD would have been observed in men, as previously reported in UK Biobank data (Mohammadi-Shemirani P, et al. Elife. 2020;doi:10.7554/eLife.58914).
Given the above issues, the most robust analysis in the current report is that assessing whether PCOS itself affects bone outcomes, with a suggestive effect to increase BMD (P = .07) and a significant reduction in fracture risk (OR = 0.98 95% CI, 0.96-0.99; P = .007). That PCOS may be beneficial for bone represents a hypothesis that should be tested in future studies.
Collapse
Click Here to Manage Email Alerts