Sociodemographic profile of primiparous mothers from nine birth cohorts in three Brazilian cities Perfil sociodemográfico de mulheres primíparas de nove coortes de nascimentos em três cidades brasileiras Perfil sociodemográfico de mujeres madres primíparas de nueve cohortes de nacimiento en tres ciudades brasileñas

Fertility reduction is a phenomenon observed in demographic transition. The demographic changes noted in female fertility represent a need for adjustment on health services regarding female health and family planning support. Thus, this study aimed to perform a descriptive analysis by tracing the sociodemographic profile of primiparous mothers belonging to nine Brazilian birth cohorts, in three cities from different states. Standardized questionnaires were applied to assess reproductive characteristics and covariables. Primiparous mothers were defined as women whose child included in birth cohorts was their firstborn child. Sample description was performed using analysis of variance (continuous variables) and chi-square (categorical variables). In total, 44,615 women were included in the analyses and 41.8% (95%CI: 41.3; 42.2) were categorized as primiparous. The primiparity rates were the lowest in Ribeirão Preto (São Paulo State) 1978 (32%; 95%CI: 30.9; 33.1) and the highest in most recent cohorts, reaching up to 50% of the participants (São Luís – Maranhão State 2010: 47.2%; 95%CI: 45.8; 48.6; Ribeirão Preto 2010: 50.2%; 95%CI: 49.1; 51.4); Pelotas (Rio Grande do Sul State) 2015: 49.4% (95%CI: 47.9; 50.9). Primiparous mothers’ age and schooling increased over the years in all cohorts. Maternal age at the first childbirth behaved similarly in the three studied cities. There was an increase in the proportion of firsttime mothers that were older, higher educated and belonged to richer income groups. Also, the proportion of teenage mothers (aged 15 years or younger) increased until the early 2000’s and started decreasing around the years 2010, especially among women in the poorer income groups.


Introduction
The reduction of fertility is one of the transformations observed in demographic transition -along with mortality reduction and changes in morbidity patterns (from infectious to chronic diseases). This phenomenon has developed over the last 150 years, incurring changes in the population's structure with an increase in the absolute number of older adults as well as in the proportion of older adults over children in the population 1 . Women no longer reach an average of six or seven childbirths but rather less than two in most economically advanced countries (The World Bank. Fertility rate, total (births per woman), 2018. https://data.worldbank.org/indicator/SP.DYN.TFRT.IN, accessed on Mar/2020).
In Brazil, fertility control was a heterogeneous event first observed, in the early 1940s, among the richer and higher educated women from urban areas in the country's most developed region (Southeast). The North and Northeast regions showed a noted decrease in fertility 20 years later, being this difference observed not exclusively in some regions but also among rural and urban areas 2 . These regional differences can still be observed. When compared to women living in the North, women from more developed regions show lower fertility levels, being the fertility profile influenced by education, socioeconomic status, age, macro-region, marital status, age of first sexual intercourse, and age at first birth 3 .
Since the mid-2000s, fertility rates in Brazil have been below the replacement level. This, along with an increase in the mean age of fertility between 2000 and 2010, suggests that a fall in fertility levels in the country could have been intensified by the postponement of births 4 . Postponing the first pregnancy could also result in a larger proportion of women reaching the end of the reproductive period without or with fewer children than planned.
Although an increase in the proportion of women aged over 30 years giving birth has been observed, the persistence of deliveries from girls aged under 15 years is still noted, representing 0.7% of total births, reaching up to 3.7% among indigenous women (Brazilian Health Informatics Department. http://www2.datasus.gov.br/DATASUS/index.php?area=0205&id=6936, accessed on 30/Sep/2020). An increasing tendency of unwanted pregnancies is also observed. Approximately three-quarters of pregnancies of single women aged from 15 to 19 years are unwanted 5 . In addition to the social, economic, and family constraints this entails, many of these pregnancies are interrupted by unsafe abortion, a major cause of maternal mortality and morbidity 6 .
Primiparous mothers' sociodemographic and health characteristics are especially relevant since first-time pregnancies are associated with a three times higher risk of pre-eclampsia 7 along with other complications during pregnancy and childbirth such as prolonged labor 8,9 , and emergency cesarean section 8,10 . Advanced maternal age -shown to be negatively associated with fecundability and the ability to conceive 11 -also increases the risk of developing hypertension-related disorders during pregnancy 12,13,14 .
The changing scenario in childbearing will result in new demands to be met in healthcare services, including pregnancy and further family planning assistance. Considering this, this study aimed to perform a descriptive analysis by tracing the sociodemographic profile of primiparous mothers belonging to nine Brazilian birth cohorts, dating from 1978 to 2015 (1978, 1982, 1993, 1994, 1997, 2004, 2010, and 2015), in three different cities from three states.

Methodology
This was a cross-sectional population-based study using data from nine Brazilian birth cohorts conducted between 1978 and 2015 in the cities of Ribeirão Preto ( 15,16,17 . Pelotas is a city with approximately 350,000 inhabitants located in Rio Grande do Sul, the Southernmost state of Brazil. In 2010, its HDI was 0.739. In the years 1982,1993,2004, and 2015, all live births from January 1 to December 31, to mothers that lived within the urban area of Pelotas and in the Jardim América district (which currently belongs to the neighboring municipality of Capão do Leão), were deemed eligible for the study. The total number of participants was 5,914 in 1982, 5,265 in 1993, 4,287 in 2004, and 4,275 in 2015, representing 98.4%, 99.1%, 99.2%, and 97.2% of the city births, respectively. More information on the Pelotas birth cohort methodology has been described previously 18,19,20,21 .
The city of São Luís is the capital of the state of Maranhão and it is located in the Northeast region, one of the poorest states in Brazil. With a population of 958,545, the HDI of São Luís in 2010 was 0.768. Two birth cohorts were initiated in the city. The first included all births that occurred in the city maternity wards from March 1, 1997 to February 28, 1998, totaling 2,831 eligible children with a response rate of 96.3%. The second cohort was part of the BRISA study and evaluated all live births from January 1 to December 31, 2010. One-third of all 21,401 live births in this period were systematically sampled into the study. A total of 5,212 individuals were included, comprising 95.4% of all eligible births. Details of the São Luis cohort methodology can be found in previous publications 22,23 .
The BRISA study was developed in 2010 aiming to analyze perinatal outcomes, sociodemographic and behavioral factors, and use of health and perinatal services in a new birth cohort in the cities of Ribeirão Preto and São Luís and to compare these indicators to those from the previous cohorts in each site 22 . In 2018, Pelotas joined Ribeirão Preto and São Luís in a project intended to bring the three centers that manage the oldest and largest birth cohorts in Brazil (Ribeirão Preto, Pelotas, and São Luís) together to coordinate efforts ( Figure 1).
All cohorts included in this study applied standardized questionnaires to assess socioeconomic, demographic, reproductive, and behavioral information.
Primiparous mothers were defined as women whose pregnancy to the child included in any birth cohorts was their first. Only singleton births were included in the analysis.
This study included the following maternal characteristics as covariates: age (years), schooling (years), marital status (with or without a partner), skin color (white or other), and family income (quin-

Figure 1
Brazilian births cohorts included in the study.

BRISA: Brazilian Birth Cohort Studies -Ribeirão Preto-São Luís.
Cad. Saúde Pública 2021; 37(4):e00057520 tiles). All cohorts collected maternal date of birth on the perinatal interviews enabling the estimation of maternal age at childbirth. Maternal schooling and, dwelling with or without a partner were collected with maternal self-report in all cohorts. The interviewed mothers answered questions about the latest grade they completed at school and this information was converted into years, considering the organization of the Brazilian schooling system. The mothers were also questioned about their relationship status (single, married, divorced, widowed, or living with a partner). These answers were categorized into living with (married or living with a partner) or without (single, divorced, widowed) a partner. Skin color was collected by the interviewer's perception in Pelotas 1982 and 1993, and Ribeirão Preto 1978, and maternal self-report in Pelotas 2004 and 2015, and 2010 BRISA Cohorts (Ribeirão Preto and São Luís). Also, the participants answered about family income, including the income anyone that lived in the household had received in the month preceding the interview.
A sample description according to the cohort of origin and socio-demographic characteristics was carried out using analysis of variance (continuous variables) and chi-square (categorical variables). Also, the frequency of primiparity and its respective confidence intervals in each cohort were described -considering a significance level of 0.05. Chi-square analysis and analysis of variance were used to describe the primiparous mothers belonging to each birth cohort. Furthermore, a detailed description of the primiparas according to the mother's age and family income in the perinatal period was performed.
Statistical analyses were performed using the Stata version 16 (https://www.stata.com/). The cohorts included in this study were approved by the Ethics Research Committees of the Universities at each city, associated with the National Research Ethics Committee (CONEP, in Portuguese). Written informed consent form was obtained from all subjects.

Results
In total, 44,615 mother/child dyads were included in this study. In Ribeirão Preto, there were 6,827 in      could be observed in the group of mothers aged 35 years or older at childbirth, approximately 7% in the former two cities and 4% in São Luís, in the same period (Table 4). When describing primiparous aged 15 or younger and 35 or older according to family income, in the cities of Pelotas and Ribeirão Preto, the poorer income quintile showed an expressive decrease in the 1990s followed by a progressive increase of young mothers in the late 2000s and 2010s. A similar pattern can be observed in Ribeirão Preto for the second poorer income quintile but not in Pelotas. For the richer income quintile, the proportion of mothers that were aged 15 years or younger behaved similarly to the poorer quintile, a decrease followed by an increase in later years can be observed.

Cohorts
Age ( Among mothers aged 35 or older during their first childbirth a decrease of proportion was observed among the poorest income quintiles and a significant increase in the richer quintiles in the cities of Ribeirão Preto and Pelotas. In the city of São Luís, the poorer income quintiles suffer little variation between the two cohorts for the primiparous aged 15 or younger, and the proportion of mothers aged 35 or older increased in the fourth income quintile but not in the richest one (Table 5).

Discussion
Data from the Pelotas birth's cohorts show that maternal average age and schooling years increased over time. This information is also evident when comparing the data from 1978 and 2010 from Ribeirão Preto. Childbearing seems to have been postponed among the rich and higher-educated women. Primiparity among women aged over 30 years increased among the rich but remained stable for the poor. Teenage childbearing decreased until the early 2000s followed by an increase, especially among poor women. The rise in the proportion of women who chose to delay childbearing reflects the stage of demographic transition Brazil is going through -where a low-level fertility equilibrium has been reached. The postponement of childbearing is associated with low fertility rates, as noted when women delay their first pregnancy to older ages, also reflecting on a reduced number of children per woman 25 . Other than a reduced number of children, delaying childbearing has a demographic effect in popula-Cad. Saúde Pública 2021; 37(4):e00057520 tions growth since there will be fewer women in the future and therefore fewer births. Also, it causes a sociological effect, because of the idea of an ideal family portrayed for future generations to follow 25 . Considering that there is a known association of age advance with fertility reduction, childlessness (reaching the end of the fertility stage without giving birth) is getting more common in countries that reached a low fertility equilibrium 4 . In Brazil, this low fertility pattern was not evenly reached among different socioeconomic groups. White women and those women with a higher level of education (4-7 years of study) were the first to postpone fertility compared to black women and those with lower education levels (0-3 years) in the years 1980s and 1990s 4 .
In our sample, the proportion of primiparous mothers aged 35 years or older was higher among the richer. This is likely a reflection of the association between education and economic achievement that can result in the delaying of childbearing among higher educated women, that prioritize career fulfillment when planning to start a family 26 .
Moreover, being a first-time mother when the woman is aged 35 years or older is related to a higher risk of complications due to the accumulation of risk factors. Primiparity and maternal older age are associated with the risk of developing hypertension and pre-eclampsia that can lead to preterm childbirth and complications to the infant 8,13 . Considering the known risks for first-time mothers who are aged 35 years or older, this group's antenatal care can be especially directed to prevent and to control them.
Teenage childbearing is directly associated with unplanned pregnancies and it is more commonly observed among the poorest population strata, those living in rural areas, and with lower schooling years. Pregnancy and labor represent a health risk for this population of young women, being one of the main causes of death among girls aged from 15 to 19 years worldwide 27 .
In this article, we observed that teenage pregnancies were rising when comparing the cohorts that date until 2004 and showed a decrease in the 2010 and 2015 cohorts, suggesting a patterns of growth followed by a decline that could be a reflection of the lack of continued investments in policies that focus in this age group when investing in women's health. Also, Brazilian economic policies that had been focusing on poverty reduction and human development [Brazil's HDI rising from 0.649 (2000) Cad. Saúde Pública 2021; 37(4):e00057520 to 0.755 (2015)] suffered a step back in recent years with drastic political changes that could make the scenario for teenage pregnancies in years to become even worse.
Reproductive health agenda should plan actions that focus on preventing possible complications associated with primiparity in mothers aged over 35, as well as, on younger girls under 19 years-old, in lower socioeconomic groups, aiming to reduce the vulnerability these women are placed in and the risk of unwanted pregnancies.
In Brazil, the female health agenda started being developed in the early 20th century, focusing exclusively on pregnancy and labor. These policies are characterized by not evaluating the population's health needs and establishing its goals on a general level. Since then, programs and policies have been trying to adjust to female health needs such as contraception and family planning. In 2003, the Woman's Integral Health Attention Policy was launched aiming to broadly cover female health needs. However, teenage female health is listed as one of the gaps still to be addressed 28 .
We also observed an increase in primiparous mothers with black or brown skin color, reaching up to about 30% of the mothers in the 2015 Pelotas cohort (compared to 18% in 1982) and 42% in Ribeirão Preto's 2010 cohort (compared to 30% in 1978). This could be due to how the variables were collected, with the interviewer in Pelotas 1982 and 1993, and Ribeirão Preto 1978, and self-report in Pelotas 2004 and 2015, and Ribeirão Preto 2010. Also, African ancestry consciousness has arisen over the years, possibly influencing participants' answers 29 .
The city of São Luís showed some results that did not agree with those observed in Ribeirão Preto and Pelotas (lower difference in maternal age and schooling over the studied periods, increase (and not decrease) in the proportion of primiparous living with a partner, lower proportion of mothers aged over 30 years old, smaller reduction among teenage mothers). Among the included cities, São Luís has the most disparities compared with the other included populations (about 80% of the black and brown-skinned population, seaside located, with an HDI near Pelotas' index). This could explain why the included covariates suffered lower improvements throughout the compared periods (1997 to 2010), for example, the lower increase of primiparous mothers aged 35 or over.
Regarding the study limitations, a clear disparity in the way information was collected in each cohort and especially through the years can be observed. Older cohorts tend to present greater proportions of missing information for the interviewed individuals turning the comparison within and between locations difficult to perform.
A strength of this study that should be highlighted is the high enrollment rates of all included cohorts -which are above 95%. This increases the representativeness of the findings of the included samples over the populations they were taken from.

Conclusion
In the 40 years covered by this study, primiparous mothers' profile has changed. Women who are delivering their first child tend to be older, higher educated, and to belong to richer income groups. A rise in the proportion of women having their first child after 30 years of age was also observed. Teenage childbearing showed a decrease in the whole population, however, an increase in the proportion of primiparous teenage mothers -notably among those in lower-income groups -in the late 2000s should be understood as an awakening sign to the direct political actions to be taken in order to break this rising phenomenon.  9; 50,9). La edad de las madres primíparas y escolaridad se incrementó a lo largo de los años en todas las cohortes. La edad maternal en el momento del nacimiento del primer hijo se comportó de forma similar en las tres ciudades estudiadas. Hubo un incremento en la proporción de las madres por primera vez, que eran más viejas, con más formación educativa, y pertenecientes a los grupos con ingresos más altos. Asimismo, la proporción de las madres adolescentes (