AbstractObjectiveAddition of androgen receptor (AR) blockade (flutamide) to insulin-sensitising therapy (metformin) may confer synergistic benefits in girls with hyperinsulinaemic androgen excess. We hypothesised that girls with shorter AR gene CAG repeat alleles, and thus greater receptor sensitivity, might benefit more from the addition of low-dose flutamide.DesignOpen randomised crossover study.MethodsIn this study, 32 post-menarcheal girls (mean age 12.1 years) with a history of low birth weight and precocious pubarche were subgrouped by CAG genotype (‘short’: CAG mean length ≤20, n=14; ‘long’: CAG >20, n=18). Within each subgroup, girls were 1:1 randomised to metformin alone (850 mg/day) or in combination with flutamide (62.5 mg/day) for 12 months. To allow comparisons with no treatment, long-CAG girls randomised to flutamide–metformin, and short-CAG girls randomised to metformin alone were observed for 12 months before treatment. Body composition by absorptiometry, fasting lipid profiles and levels of insulin, glucose and androgens were measured during the first 12 months on each treatment.ResultsIn all girls, 12 months flutamide–metformin lowered body fat and improved lipid profiles when compared with no treatment. Compared with metformin alone, flutamide–metformin achieved greater reductions in the percentage of body fat and abdominal fat mass in the short-CAG subgroup (P=0.001 to P<0.0001). In contrast, in the long-CAG subgroup, flutamide–metformin produced no further improvements when compared with metformin alone.ConclusionsIn young post-menarcheal girls with preclinical androgen excess, low-dose flutamide–metformin improved body composition and key endocrine–metabolic abnormalities. However, only those girls with genetic markers of greater AR sensitivity may benefit from the addition of flutamide above metformin alone.