Hair loss, insulin resistance, and heredity in middle-aged women. A population-based study.

Abstract

Context: The association of androgenic alopecia (AGA) with insulin resistance, coronary artery disease and hypercholesterolemia has been previously reported in men, but no such association has been reported in women with female androgenic alopecia (AGA). Female AGA has usually been linked with hyper-androgenism and hirsutism and, most recently, also with polycystic ovarian syndrome (PCOS), even though epidemiological documentation of the latter association is scanty. Polycystic ovarian syndrome is quite common among Caucasian women, and its association with insulin resistance is well documented.

Objectives and design: The aim of this study was to obtain a more precise estimation of the prevalence on female AGA and to describe its possible connections with insulin resistance linked parameters and with paternal and maternal family history of alopecia. A cross-sectional population based cohort survey was carried out in the City of Oulu, Finland in 1998.

Setting and participants: As a part of a population based cohort study the hair status of 324 women aged 63 years was assessed by a modification of Ludwig's scale. The background data consisting of anthropometric measures (weight, height, body mass index, waist, hip and neck circumferences), smoking status, chronic diseases and their medication as well as the family history of AGA were collected by questionnaires and interviews made by study nurses and in clinical examination. Blood samples for laboratory tests were taken on the same occasion.

Results: The prevalence of extensive loss of hair (at least grade II or III on Ludwig's scale) was quite high (31.2%). The insulin resistance associated parameters, such as waist and neck circumferences, abdominal obesity measured by waist-to-hip ratio, mean insulin concentration (11.3 mU/l versus 9.95 mU/l, p=0.02) or urinary albumin-to-creatinine ratio (1.80 versus 1.58, p=0.01), were significantly higher in women with extensive hair loss compared to those with normal hair or only minimal hair loss (grade I on Ludwig's scale). The women belonging to the highest quintiles of neck or waist circumferences had significantly increased risk for extensive hair loss compared to those with normal hair or minimal hair loss, the unadjusted ORs being 2.25 (95% CI, 1.26-4.03) and 1.75 (95% CI, 1.00-3.07), respectively. Similarly in women with hyperinsulinemia (fs-insulin >10 mU/l), microalbuminuria (urinary albumin-to-creatinine ratio exceeding the highest microalbuminuria decile (>2.5 mg/mmol) and paternal history of AGA the ORs for alopecia were increased being 1.65 (95% CI, 1.02-2.67), 2.39 (95% CI, 1.21-4.73) and 2.08 (95% CI, 1.26-3.44). All of these ORs, except those for highest quintiles of waist and neck circumferences remained significant in multiple adjusted models.

Conclusions: According to the results of this study, female AGA (grade II or III on Ludwig's scale) was quite common among Finnish women aged 63 years. Our results support the hypothesis that women with some markers of insulin resistance have significantly increased risk for female AGA. Paternal history of alopecia seemed to be more common in female AGA compared to women with normal or minimal loss of hair.