Contraceptive use and pregnancy outcome in three generations of Swedish female teenagers from the same urban population.

Background: The primary objective of the study was to describe contraceptive use, reasons for discontinuation of contraception and pregnancy outcome in three generations of female teenagers over a period of 20 years. The secondary objective was to describe the relationship between contraception, smoking, body mass index (BMI) and socioeconomic status (SES).

Study Design: A cross-sectional comparison of 19-year-old women born in 1962, 1972 and 1982 and living in the city of Gothenburg, Sweden, in 1981, 1991 and 2001 was conducted. Contraceptive use, pregnancy outcome, smoking and weight/height were assessed by a postal questionnaire.

Results: Current contraceptive use was unchanged between the 62 (60%) and 72 cohorts (62%) but had increased (p<.01) in the 82 cohort (78%); there was no difference in contraceptive use between SES groups at any time. Condom use alone increased over time (p<.01), and the use of oral contraception and a condom together had increased in the 72 and 82 cohorts compared to the 62 cohort (p<.01). Reasons given for using and discontinuing oral contraceptives in three generations of teenagers were studied over 20 years. In addition to contraception, oral contraception was used to reduce dysmenorrhea and heavy bleeding. Discontinuation due to bleeding disturbances decreased (p<.01) over time, whereas discontinuation due to mental side effects increased (p<.01). The percentage of women who had been pregnant at < or =19 years of age in the 82 cohort (7%) was lower (p<.01) than in the 1962 (11%) and 1972 (13%) cohorts. However, there was a successive increase (p<.001) in the percentage of women who had been pregnant more than once at < or =19 years of age (1962/1972/1982: pregnant more than once, 8%/21%/31%). Smoking decreased over time (p<.01) and was no longer related to SES in the 82 cohort. BMI increased (p<.01) over time. There was no difference in BMI between SES groups in the 62 and 72 cohorts but was higher in the low-SES group in the 82 cohort compared to the middle (p<.01) and high (p<.05) SES groups.

Conclusions: Contraceptive use was higher in the 82 cohort where there was a corresponding reduction in the percentage of women who had been pregnant at < or =19 years of age compared with the 62 and 72 cohorts. Discontinuation of oral contraception due to mental side effects increased over time. The prevalence of smoking decreased and BMI increased, and there were changes in smoking prevalence and BMI in the different SES groups over time.