Negative impact of polycystic ovary syndrome on bone health: a systematic review and meta-analysis

Abstract

Abstract BACKGROUND Polycystic ovary syndrome (PCOS) has reproductive and metabolic aspects that may affect bone health. Controversial results from different studies regarding the risk of fractures, bone mineral density (BMD) or bone markers led to uncertainty whether PCOS might improve or deteriorate bone health. OBJECTIVE AND RATIONALE This study aimed to investigate the impact of PCOS on bone markers, BMD and fracture risk. SEARCH METHODS A systematic review and a meta-analysis were carried out. PubMed, EMBASE and Cochrane databases were searched for eligible studies from 1st of January of 1990 to 9th of October of 2018. Eligible studies enrolled women older than 18 years with PCOS, which should be diagnosed according to the Rotterdam Consensus, the Androgen Excess Society, the National Institutes of Health Consensus or the International Classification of Diseases. The studies were grouped according to patient mean BMI: <27 kg/m2 or ≥27 kg/m2. The results were polled as mean difference (MD), standardized MD (SMD) and hazard ratio (HR). OUTCOMES Overall, 921 studies were retrieved, and 31 duplicated studies were removed. After screening the titles and abstracts, 80 studies were eligible for full text reading. Of those, 23 studies remained for qualitative synthesis. With the exception of one study, all studies were considered high quality based on the Newcastle–Ottawa scale (NOS; score ≥6). Meta-analysis was performed in 21 studies, with a total of 31 383 women with PCOS and 102 797 controls. Women with PCOS with BMI <27 kg/m2 had lower BMD of the total femur (MD, −0.04; 95% CI, −0.07 to 0.00; I2 = 31%; P = 0.22) and spine (MD, −0.07; 95% CI, −0.13 to −0.01; I2 = 70%; P < 0.01) when compared with the control group, whereas for women with BMI ≥27 kg/m2 no difference was observed (femur: MD, 0.02; 95% CI, −0.02 to 0.05; I2 = 20%, P = 0.29; spine: MD, 0.02; 95% CI, −0.06 to 0.05; I2 = 0%; P = 0.84). Osteocalcin was remarkably reduced in women with PCOS with BMI <27 kg/m2 (SMD, −2.68; 95% CI, −4.70 to −0.67; I2 = 98%; P < 0.01), but in women with BMI ≥27 kg/m2, there were no differences between PCOS and controls. Few studies (n = 3) addressed the incidence of bone fractures in women with PCOS. The HR for total bone fractures did not identify differences between women with PCOS and controls. WIDER IMPLICATIONS On the basis of the available evidence, it is possible to assume that PCOS in women with BMI <27 kg/m2 is associated with reduced BMD in the spine and femur, and decreased bone formation, as manifested by lower levels of circulating osteocalcin. These findings suggest that bone parameters in PCOS may be linked, to some extent, to adiposity. These studies included premenopausal women, who have already achieved peak bone mass. Hence, further prospective studies are necessary to clarify the existence of increased risk of fractures in women with PCOS.