BACKGROUND: The effect of circumstances and exposures early in life on the risk of developing ankylosing spondylitis (AS) is largely unknown. The purpose of this study was to determine whether perinatal characteristics predict development of AS.

METHODS: AS cases (n = 1960; 59 % men) were defined as listed with a diagnosis of AS at least once in the Swedish National Patient Register and registered in the Swedish Medical Birth Register (born ≥1973). Population controls were retrieved from the Swedish Population Register (n = 8378; mean 4.3 controls/case), matched on birth year, sex and county. Odds ratios (OR) for developing AS were determined through conditional logistic regression, with regard to: birth weight, birth order, season of birth, maternal age, gestational length, size for gestational age, type of birth, mode of delivery, congenital malformations, mothers' country of birth, mothers' civil status and size of delivery unit.

RESULTS: In the univariate analyses statistically significant increases in risk for developing AS were observed for having older siblings (OR 1.18; 95 % Cl 1.06-1.30). No association was observed for the remainder of analysed exposures, although there was a weak association with birth weight below 3000 g (OR 1.19; 95 % CI 1.04-1.37), though not for "low birth weight" <2500 g (OR 0.90; 95 % CI 0.70-1.16). The increase in risk associated with having older siblings was consistent in a multivariate analysis adjusting for possible confounders (OR 1.23; 95 % Cl 1.09-1.39). The direction and magnitude of the point estimates were also consistent in several sensitivity analyses and when stratifying by sex.

CONCLUSIONS: Having older siblings was associated with an increased risk for developing AS. These results need to be repeated and confirmed in other cohorts.