Trends in preterm births in Flanders, Belgium, from 1991 to 2002

Abstract

Changes in the preterm birth rate have been attributed predominantly to increases in multiple pregnancies, associated with advanced maternal age and assisted reproduction, and to obstetric intervention. We examined their contribution to the frequencies of preterm (<37 weeks), very preterm (<32 weeks) and severely preterm (<28 weeks) birth among 700 383 singleton and twin births in Flanders from 1991 to 2002. We examined changes across four 3-year periods (triennia) with confidence interval [CI] analysis and yearly incremental rates using linear and logistic regression analyses. Over the 12 years, twin pregnancies increased from 1.5% to 2.0%, averaging 1.6% [95% CI 1.54, 1.66] in 1991-93 and 1.9% [95% CI 1.81, 1.94] in 2000-02 (P < 0.001). The proportion of women aged 35 years or more increased from 6.8% [95% CI 6.69, 6.92] in 1991-93 to 11.3% [95% CI 11.2, 11.5] in 2000-02 (P < 0.001) and those aged under 20 from 1.9% [95% CI 1.81, 1.93] to 2.3% [95% CI 2.26, 2.41] (P < 0.001). Assisted reproduction increased from 2.6% [95% CI 2.48, 2.62] to 4.2% [95% CI 4.11, 4.30] (P < 0.001) and obstetric intervention to end pregnancy from 36.2% [95% CI 36.0, 36.4] to 40.3% [95% CI 40.1, 40.6] (P < 0.001). These increases related to an annual increase of 0.23% in the preterm birth rate from 5.5% [95% CI 5.4, 5.6] in 1991-93 to 7.2% [95% CI 7.1, 7.3] in 2000-02 (P < 0.001). The proportions of very and severely preterm births also increased by nearly a third, but their contribution to the total preterm birth rate remained stable at 15% and 5%, respectively. Odds ratios for the increases per year were 1.035 [95% CI 1.032, 1.038] for preterm birth, 1.024 [95% CI 1.018, 1.031] for very preterm and 1.028 [95% CI 1.017, 1.040] for severely preterm births after adjusting for other changes in the population. Overall, the data show, first, marked increases in the frequency of known contributors to the preterm birth rate, including twin pregnancies, advanced maternal age, assisted reproduction and obstetric intervention. Second, the preterm birth rate further increased significantly within subgroups of women with one or more of these characteristics. Third, the preterm birth rate also rose, from 4.4% [95% CI 4.2, 4.5] in 1991-93 to 5.6% [95% CI 5.5, 5.8] in 2000-02 (P < 0.001), in women with none of these contributing factors. This indicates that changes in the frequency of these known predictors are insufficient to explain the steady increase in preterm, very preterm and severely preterm births over more than a decade.