Reproducibility of self-reported menopause age at the 24-year follow-up of a population study of women in Göteborg, Sweden

Abstract

To study the reproducibility of self-reported age at natural menopause, reported by women in their 70s, compared with menopause age reported in the initial postmenopausal period. A prospective study conducted in Göteborg, Sweden, based on a random sample of the total female population, started in 1968-1969, with follow ups in 1974-1975, 1980-1981, and 1992-1993. A total of 1,009 women born in 1922, 1918, or 1914 (participation rate 90.3%), representative of women of the respective ages in the general population, participated in the initial examination in 1968-1969. Of those women reporting natural menopause some time between 1968 and 1981, 565 women recalled their menopause age in 1992. The mean interval between the first and second reports was 18.3 years. The mean difference between first-reported and recalled menopause was 0.05 years. When menopause age was recalled in 1992-1993, 55.6% of the women with a natural menopause recalled their age at menopause correctly within 1 year, 22.6% underestimated their menopause age by more than 1 year, and 21.8% overestimated their menopause age by more than 1 year. Women undergoing early menopause (<45 years) tended to overestimate menopause age, whereas women undergoing late menopause (>55 years) tended to underestimate menopause age. These differences could not be explained by age, interval since menopause, smoking, exercise, education, or socioeconomic status. Menopause ages reported at an interval of almost 20 years showed a significant correlation. However, a regression tendency of the values toward the mean was observed, suggesting that strong correlation at the group level does not imply precision at the extremes.