Sample characteristics
Overall, 1031 women born before 1950 with an average age of 70 years and 1007 after 1950 with a mean age of 48 years were included in the analysis. Table 1 displays the main characteristics of the sample. Compared with women born before 1950, women born after 1950 had a higher proportion of overweight and obesity, which is in line with a greater consumption of energy and alcohol per day. The proportion of smoker women was also higher in this group. Regarding reproductive factors, the proportion of women who had their menarche after 12 years old, had a more advanced age for their first birth, had a smaller number of children, and shorter breastfeeding periods was higher among women born after 1950. The use of hormonal contraceptives was also higher in this group and they also reported more fertility problems.
Association between SE indicators and reproductive factors
Figure 1 shows the relationship between socioeconomic level and age at first delivery. All four SE indicators were related with age at first delivery. Higher parent’s SE, education and occupational level and lower urban vulnerability were associated with higher age at first delivery. Although these differences are seen in all women, they are higher in women born after 1950 with a difference of about 5 years between the lowest and the highest educational and occupational levels. (Averaged age at first delivery in women with unfinished primary studies equals 23.4 years vs 29.8 in women with high education; averaged age in women with low occupational level equals 24.1 vs 28.7 in women with high occupational level.) Age at first delivery was also associated with urban vulnerability, with little changes in women born before / after 1950. In this way, although in Spain, the tendency is to increase the age at first child (supplementary figure 1), supplementary figure 2 shows that there are changes in the tendency when stratifying by educational level: the age at first child goes down all the time in women with unfinished primary or primary study levels, while the tendency is U-shaped in women with secondary studies. Supplementary figure 3 shows the percentages of women with university or secondary studies have largely increased in recent cohorts. Therefore, the average change in the age at first child is mostly due to changes in women’s educational level rather than changes within each educational level.
The association between socioeconomic level and the number of pregnancies is illustrated in Figure 2. In women born after 1950, but not in women born before 1950, the educational and occupational levels were negatively associated with the number of pregnancies: the higher the educational and occupational level, the lower the number of pregnancies (figure 2c and 2d).
The percentage of women who have taken hormonal contraceptives is nearly double in women born after 1950 compared to those born before 1950 and this ratio continue almost constant across the different SE levels. The proportion of women who took the pill increased with higher educational and occupational levels and with lower urban vulnerability index in women born before 1950; halving the consumption in the lowest educational levels with respect to the highest ones (25.6% vs 37.8 %). and 22.6% in urban vulnerability Q4 compared to 44.7% in Q1 (figure 3). In women born after 1950, however, there was no association between socioeconomic indicators and having taken hormonal contraceptives except in higher parent’s SE.
Figure 4 shows the proportion of women who took hormone replacement therapy in the two periods studied. Statistical significance is only reached in the group of those born before 1950, with more educated women near doubling the use of hormone replacement therapy (HRT) of less educated women (14.4% compared to 9.2%) and women living in the less vulnerable areas doubling the use of HRT of women living in more vulnerable areas (Q1: 14.1% vs. Q4: 9.3%).
Education level and urban vulnerability index were the SE indicators associated with fertility problems. Educational level displayed 6% more probability in university level for women born after 1950 compared to less educated women. Whereas, a lower urban vulnerability was associated with higher fertility problems, but only in the group of women born after 1950 (figure 5).
Association between socioeconomic scores and other quantitative variables (supplementary table 1)
Number of alive newborns and average time between pregnancies
The relationship between number of alive newborns and SE level mirrors that of number of pregnancies and the average time between pregnancies. In women born after 1950, education and occupational level was negatively associated with the number of pregnancies: a higher educational and occupational level was associated with fewer pregnancies.
Socioeconomic level and breastfeeding
The occupational level was negatively associated with months of breastfeeding (higher occupational level with less months of breastfeeding) in the whole sample and in women born before 1950. The cumulative number of months of breastfeeding was also negatively associated with education level in the whole sample. The relationship between the months of breastfeeding and SE level mirrors that of the average time of breastfeeding per child.
Association between individual and contextual socioeconomic scores with other dichotomic variables associated with reproduction (supplementary table 2)
Socioeconomic level and abortion
The percentage of women who have suffered at least an abortion was associated with higher educational and parent’s SE level in the whole sample and in women born before 1950, but all these associations disappeared in women born after 1950.
Socioeconomic level and dead newborn
Women with lower educational level were more prone to have a dead newborn, but this result was only reproduced in women born before 1950. No other SE indicator was associated with this event.
Socioeconomic level and diagnosis of fertility problems
The level of education was indicators of SE associated with fertility problems. The educational level showed a V-shaped pattern with higher probability at the university level for women born after of 1950.
Preterm and post-term newborn
The proportion of women with post-term newborns decreased with increasing educational level in women born after 1950. However, no SE factor was found associated with the preterm newborn.
No association could be found between SE level and age at menopause, age at menarche and average time between pregnancies.