Elsevier

Social Science & Medicine

Volume 54, Issue 8, April 2002, Pages 1281-1298
Social Science & Medicine

Depression, parenthood, and age at first birth

https://doi.org/10.1016/S0277-9536(01)00096-XGet rights and content

Abstract

This study tests the hypothesis that the correlation between current depression and parenthood depends on the age at first birth for adults. An early first birth suggests a poor start in life. It may reflect a disordered transition from adolescence into adulthood and may itself disrupt that transition, with life long consequences that influence emotional well-being. To test the hypothesis we analyze data from a 1995 survey of 2592 US 18 through 95. Among the parents we find a generally negative association between age at first birth and recent feelings and signs of depression, adjusting for age, sex, minority status, and education of the respondent's parents. Respondents who had a first birth before age 23 report more feelings and signs of depression than do nonparents; those who had a first birth after age 23 report fewer than do nonparents. The association between depression and age at first birth is monotonic for males but parabolic for females. The age at first birth associated with the lowest predicted depression for females is around 30. Regression analyses indicate that several correlates account for the apparent psychological benefits of delaying first birth: later first marriages, higher educational attainment, lower risk of having had a prolonged period needing a job but not being able to find one, lower risk of having had periods lacking the money for household necessities, and better current physical health. For women, health consequences limit the emotional benefits of prolonged delay of the first birth.

Introduction

This study explores the possibility that the association between parenthood and current depression depends on age at first birth. Whether being a parent erodes emotional well-being or improves it may depend on when in life a person became a parent. Early parenthood may disrupt the transition from adolescence into adulthood, with persistent distressing consequences (Heath, McKenry, & Leigh, 1995; McLaughlin & Micklin, 1983; Vernon, 1991; Williams, McGee, Olaman, & Knight, 1997). Older age at first birth may switch the association of parenthood with current depression from unfavorable to favorable. However biological factors may limit the emotional benefits of delay, particularly for women.

In societies organized around individual competition for status a conflict may exist between social and biological realities. Individuals come out of childhood lacking personal wealth, power and prestige. They compete for status and reward through the educational and occupational system. Success brings a higher rate of attainment with self-reinforcing benefits akin to compounding interest. Whatever one thinks about the desirability of this system or the validity of its operating assumptions, it creates a likely conflict between biological and social realities. Humans in such a society mature reproductively a decade or more before maturing socially. Early parenthood may indicate a poor start in life (McElroy, 1996; Moore et al., 1994; Rindfuss, Morgan, & Swicegood, 1984). It may reflect a disordered transition from adolescence into adulthood and may itself disrupt that transition, with life long consequences that influence emotional well-being.

Depression is a mood of sadness, loneliness and hopelessness accompanied by sleep problems, fatigue, and mental turmoil. People generally find depression aversive and wish to avoid it, but when depressed they despair—succumbing to a sense of futility and defeat. Often the depressed person feels trapped in a miserable present, with no way back through time to find a different route through life, and no visible way forward to a better condition. Depression is a state of debilitation as well as of anguish. The depressed person needs strength and will, but finds the personal reserves depleted.

People who become parents with undeveloped resources and unseasoned relationships often get overwhelmed by the strains of parenthood. Those strains may further disrupt the needed relationships and attainments, creating a self-reinforcing accumulation of disadvantage. The impact of children in the household on well-being provides a clue to the impact of age at first birth. Any emotional benefit of having children in the household depends on the economic prosperity of the household, the mother's paid employment, the availability of affordable child care services, and a spouse's emotional support and shared participation in child care (Glass & Fujimoto, 1994; Gove & Geerken, 1977; Kandel, Davies, & Raveis, 1985; Lennon & Rosenfield, 1994; Lennon, Wasserman, & Allen, 1991; McLanahan & Adams, 1987; Ross, Mirowsky, & Huber, 1983; Ross & Huber, 1985; Ross & Mirowsky, 1988; Ross & Van Willigen, 1996; Seal, Wright, & Sheley, 1993; Umberson & Williams, 1999). Without these, the strains of raising children outweigh the emotional benefits. That pattern suggests that people who take longer to prepare for parenthood may be more likely to benefit emotionally from it.

There are basically three sets of reasons why the association between parenthood and depression may depend on age at first birth: family formation problems, socioeconomic problems, and health problems. All three suggest benefits of delay, but the third also suggests a limit to the benefits of delay, as described below.

Family formation: Early first birth may signify a poor start to family formation. It may be associated with things such as premarital parenthood, early and unstable marriage, or high fertility (Elster & Lamb, 1986; Heath et al., 1995; Milan, 1991; Nock, 1998; Teti, Lamb, & Elster, 1987; Williams et al., 1997). Such conditions may create persistent difficulties that increase depression (Heath et al., 1995; Ross & Mirowsky, 1999; Williams et al., 1997). Conditions like being a single or divorced parent or being locked in a strained and unsupportive marriage are distressing themselves and they can lead to further difficulties in the form of disrupted education or employment and periods of material or economic hardship (McLanahan & Sandefur, 1994; Moore et al., 1994; Williams et al., 1997).

Socioeconomic attainment: Early first birth may interfere with education, employment, and household economics. Low education, interrupted employment, low household income, and economic hardship increase depression (Mirowsky & Ross, 1986; Pearlin, Menaghan, Lieberman, & Mullan, 1981; Ross & Huber, 1985). Early first births occur at a critical period in the competition for position and status. Early births can delay or deter the completion of high school, discourage entry into college, or prevent the completion of college (Anderson, 1993; Elster & Lamb, 1986; Hofferth & Moore, 1979; Hoffman, Foster, & Furstenberg, 1993; Moore & Waite, 1977). The longer individuals go to school the longer they generally delay the first birth. The demands of caring for children may interfere with employment, particularly among women (Desai & Waite, 1991; Presser & Baldwin, 1980). The combination of added household needs, lowered educational attainment, and disrupted employment increase the risk of not having the money to pay for things such as food, clothing, and shelter (Hoffman et al., 1993; Moore et al., 1994). Individually and in combination, low education, interrupted employment, and economic hardship increase depression (Ross & Huber, 1985).

Health and physical status: Health consequences also may account for an association of early parenthood with depression. Some complications of pregnancy are most common among the youngest mothers. For example, in the US the mother's risk of anemia is 29.4 per 1000 live births for women under age 20. The rate declines to 15.4 per 1000 by age 30, where it remains through the 40–49 age bracket (the oldest bracket in standard US statistics) (Ventura, Martin, Curtin, & Mathews, 1997, Table 25). To some extent these statistics reflect the fact that younger mothers are more likely to be first time mothers. To some extent they may reflect incomplete reproductive development, or poor preparation for motherhood. Younger women smoke cigarettes during pregnancy more often than older women and get prenatal care less often (Ventura et al., 1997, Tables 28 and 33).

Perhaps more to the point, education and other forms of achieved status affect health throughout life, for men as well as for women (Ross & Wu (1995), Ross & Wu (1996)). First births that disrupt status attainment may thereby create later depressing health problems (Waldron, Weiss, & Hughes, 1998).

Although social factors generally may favor delaying the first birth, the physical demands of bearing and raising children also may limit the benefits of delay, particularly for women. Pregnancy and childbirth put enormous stress on the body. Risk for a number of complications are elevated among the oldest mothers as well as the youngest. For example, in the US, pregnancy-related hypertension and eclampsia rates are highest in the 40-and -older bracket (next highest in the under 20 bracket). (Eclampsia is coma and convulsions during or immediately after pregnancy, related to high blood pressure, accumulation of fluid in body tissues, and large amounts of protein in the urine signifying kidney problems.)

Tending and supporting children also can be physically challenging, while limiting the time available for exercise (Bird & Fremont, 1991). Individuals who delay the first birth too long may not have the physical capacity to meet the challenges. Physical fitness and health generally decline as people age. A late first birth may be exhausting in itself, and may precipitate developing health problems, with consequences for emotional well-being. Older women face higher than average risk of having a pre-existing chronic condition during the pregnancy and birth (Ventura et al., 1997, Table 25). They also face higher than average risk of many pregnancy or birth complications (Ventura et al., 1997, Table 37). Chronic conditions and childbearing complications may initiate or exacerbate each other. Chronic disease, physical impairment, frequent aches and pains, poor subjective health, and low energy and fitness may be depressing (Turner & Noh, 1988). They may discourage the afflicted and also may indicate problems in the physiological systems that sustain emotional well-being.

Age at first birth can be an internal modifier of the association between parenthood and depression. Among parents the age at first birth may influence current depression. If so then it also alters the expected difference in depression between parents and nonparents. There are reasons to expect psychological benefits of delay over much of the range of age at first birth, as detailed above. Thus the association of parenthood with depression may switch from positive to negative with delay in first birth beyond a pivotal age.

The pivotal age at first birth: We define the pivotal age at first birth as the one that divides parenthood's likely emotional liability from its likely emotional benefit. Parents who had children before that age feel depressed more often than nonparents; those who had children after that age feel depressed less often than nonparents.

The optimal age at first birth: The physical strains of parenthood may limit the benefits of delay, particularly for women. If so then there may exist an optimal age at first birth, defined by the lowest ratio of depression's frequency among parents compared to that among nonparents. This would not rule out the existence of a pivotal age at first birth, as defined above. However it could create a range of ages at first birth over which parents are less depressed than nonparents. The pivotal age at first birth marks the lower boundary of that range. The peripheral age at first birth, as defined here, marks the upper boundary. The upper limit is peripheral in two senses: it marks the distal limit of a positive association between parenthood and emotional well-being, and it is likely to be well above the average age at first birth.

The paradox of delayed parenthood: If a pivotal age at first birth exists, then it creates a seemingly paradoxical effect. Over some range of ages it is better to have become a parent than not, but better yet to have delayed beyond that range. Indeed, for a man there may be no upper limit to the beneficial length of delay, so long as he did eventually become a father. This would be implied by a monotonic decrease in depression with older age at first birth. For women a parabolic relationship might limit the paradoxical improvement in parenthood's emotional benefits with longer delay. However it might create a paradox of its own. The emotional distress associated with having become a parent too early might be large and relatively obvious, but the benefits of having waited for the optimal age might be small and relatively obscure. Speaking in graphical terms, the slope to the left of the pivotal age will be steeper than the slope to the right between it and the optimal age. Such a pattern might undermine the motivation to wait for the optimal age. That in turn might reinforce a tendency to begin parenthood before or soon after the pivotal age, and well before the optimal one.

Surprisingly few previous studies examine the association between depression throughout adulthood and age at first birth. None directly compare the parents with nonparents to see if age at first birth modifies the association between depression and parenthood. It follows that none attempt to map the pivotal age dividing positive from negative association. However all of them find later first birth associated with lower depression or malaise. Our search of the abstract databases for population, social, psychological, and health sciences found only three directly relevant studies. One interviewed the mothers of children born in April of 1972 through March of 1973 at a large maternity hospital in Dunedin, New Zealand (Williams et al., 1997). The study measured the emotional well-being of the mothers at 2-year intervals from 3 to 14 years after the index child's birth and once again at 22 years after. The youngest mothers were just getting out of adolescence when first assessed and the oldest were in their 50s when last assessed. The mental health scales changed over time, measuring neuroticism in the first follow-up, malaise in the second through fourth, and depression in the rest. However the results remain essentially the same for all measures at all follow-ups. The mothers who had a first birth when under age 21 (not necessarily the index child) reported significantly more symptoms than the mothers who were older at first birth, even within categories of education (more than high school degree or not).

Two studies found similar results for US mothers and fathers, although spanning a narrower range of adulthood ages and with an indirect outcome measure for the mothers. The mothers were ages 14–24 in 1968, 29–39 in 1983, and 34–44 in 1988 (Waldron et al., 1998). The analysis related age at first birth to a health index with 11 questions about physical difficulty and six about psychosomatic distress including weakness and fatigue. The results showed a linear relationship in the first follow-up, and a quadratic one in the second with an optimum age at first birth around 27. In both periods the mothers who were under age 20 at first birth had the highest predicted level of health problems, adjusting for education, race, age, hours employed, marital status, and number of children. The study of fathers looked at men age 18–40 and found more major depression among those who fathered during adolescence, adjusting for SES, race, fertility, and age (Heath et al., 1995).

Although the studies described above do not compare parents with nonparents, other information suggests that the pivotal age at first birth may be somewhat older than the typical age at first birth. Studies usually find either no association between children in the household and depression, or an association that somewhat favors persons not raising children (McLanahan & Adams, 1987; Umberson & Williams, 1999). That pattern of findings would be consistent with a pivotal age at first birth that is near but somewhat older than the typical age at first birth.

This study tests the following hypotheses: (a) the association between parenthood and depression depends on age at first birth, (b) the association switches from detrimental (a positive association of parenthood with depression) to beneficial (a negative association) with delay in first birth beyond a pivotal age, (c) age at first birth has a parabolic effect on the relationship between parenthood and depression for women, implying an optimal age at first birth associated with parenthood's largest beneficial (negative) association, (d) the impact of age at first birth on the association between parenthood and depression depends on sex, and (e) age at first birth modifies the association between parenthood and depression because of its relationship with problems in family formation, socioeconomic attainment, and physical health.

Section snippets

Model

The model treats age at first birth as an internal modifier of the difference in depression expected between parents and nonparents (Cohen, 1968; Mirowsky, 1999). The internal modifier model has two advantages over traditional approaches to analyzing the effects of internal variables. Researchers studying the effects of internal variables often restrict the analysis to the cases for whom the variable is defined. In the present case that would mean restricting the sample to parents and analyzing

Analyses

We report two sets of regressions. The first set analyzes the form of the relationship between depression and parenthood, testing whether it depends on age at first birth. The second set analyzes possible mediators of the association.

In order to eliminate the possible bias introduced by the oversample of seniors, all regressions adjust for age (linear and squared terms). This has the same effect on estimated regression coefficients as age-weighting the sample, but avoids the bias to estimated

Internal modifier analyses

The regressions in Table 1 show that the association between depression and parenthood depends on the age at first birth. Parenthood appears to have no overall association with depression because the association is positive for earlier first births and negative for later ones. Model A shows that, overall, parenthood has no significant association with depression. Model B tests the linear trend for age at first birth. The significant negative coefficient indicates two things: (a) among parents,

Discussion

The results support the hypothesis that the association between parenthood and depression depends on the age at first birth. The pivotal age dividing positive from negative association in the US is near the reference age of 23. We chose that age at first birth as the reference point in our analyses because it allows just enough time to complete college, then get a job and get married, and then start a family, and because it is about midway between the median age at first birth for the mothers

Conclusion

The association between depression and parenthood depends on the age at first birth in the US. Parents who had a first child before the age of 23 feel depressed more often than nonparents of similar age, sex, race and status of origin. Those who had a first child after 23 feel depressed less often than nonparents. Fathers are less depressed the longer they delayed parenthood. Mothers are less depressed with delay up to about age 30. The lower depression associated with delayed parenthood

Acknowledgements

The survey of Aging, Status and the Sense of Control (ASOC) was funded by a grant from the National Institute on Aging to John Mirowsky and Catherine Ross (RO1-AG12393). Sampling, pretesting, and interviewing for the surveys were conducted by the Survey Research Laboratory of the University of Illinois. The analysis reported here was supported by a grant from the National Institute of Mental Health to John Mirowsky and Catherine Ross to study “Children, Child Care and Psychological Well-Being”

References (71)

  • S. Desai et al.

    Women's employment during pregnancy and after the first birthOccupational characteristics and work commitment

    American Sociological Review

    (1991)
  • M.L. Dewit et al.

    Education and timing of parenthood among Canadian women–a cohort analysis

    Social Biology

    (1992)
  • DiGiulia, P., Lesthaeghe, R., Moors, G., & Pinnelli, A. (1999). Fertility tempo and quantum: an empirical test of major...
  • dos Santos Silva, I., & Beral, V. (1997). Socioeconomic differences in reproductive behavior. In: M. Kogevinas, N....
  • A.B. Elster et al.

    Adolescent fathersThe under studied side of adolescent pregnancy

  • J. Glass et al.

    Housework, paid work, and depression among husbands and wives

    Journal of Health and Social Behavior

    (1994)
  • R. Gosden et al.

    Delayed childbearing

    British Medical Journal

    (1995)
  • W.R. Gove et al.

    The effect of children and employment on the mental health of married men and women

    Social Forces

    (1977)
  • D.T. Heath et al.

    The consequences of adolescent parenthood on men's depression, parental satisfaction, and fertility in adulthood

    Journal of Social Service Research

    (1995)
  • S. Hofferth et al.

    Early childbearing and later economic well-being

    American Sociological Review

    (1979)
  • S.D. Hoffman et al.

    Reevaluating the costs of teenage childbearing

    Demography

    (1993)
  • D.B. Kandel et al.

    The stressfulness of daily social roles for womenMarital, occupational, and household roles

    Journal of Health and Social Behavior

    (1985)
  • K.E. Kiernan et al.

    The age at which childbearing starts–a longitudinal study

    Population Studies

    (1983)
  • O. Kravdal

    The importance of economic activity, economic potential and economic resources for the timing of 1st births in Norway

    Population Studies–A Journal of Demography

    (1994)
  • P.M. Lantz et al.

    The social construction of the breast cancer epidemic

    Social Science and Medicine

    (1998)
  • M.C. Lennon et al.

    Relative fairness and the division of houseworkThe importance of options

    American Journal of Sociology

    (1994)
  • M.C. Lennon et al.

    Infant care and wives depressive symptoms

    Women and Health

    (1991)
  • Lund, L., & Wright, W. E. (1994). Mitofsky-Waksberg vs. screened random digit dial: Report on a comparison of the...
  • S.S. McLanahan et al.

    Parenthood and psychological well-being

    Annual Review of Sociology

    (1987)
  • S.S. McLanahan et al.

    Growing up with a single parent

    (1994)
  • S.D. McLaughlin et al.

    The timing of first birth and changes in personal efficacy

    Journal of Marriage and the Family

    (1983)
  • S.D. McLaughlin et al.

    The effects of the sequencing of marriage and first birth during adolescence

    Family Planning Perspectives

    (1986)
  • Milan, A. M. (1991). Social determinants of fertility for women aged twenty to thirty-nine in Canada and the United...
  • J. Mirowsky

    Analyzing associations between social circumstances and mental health

  • J. Mirowsky et al.

    Social patterns of distress

  • Cited by (151)

    View all citing articles on Scopus
    View full text