Health of Aging Parents and Childless Individuals

Abstract

This article reviews and presents research findings on the relationships between parenthood and health over the life span. Existing research shows lacunae. The links between reproductive behavior and longevity generally focus on family size rather than contrasting parents and nonparents. Studies of marital status differentials in survival generally confound the effects of parenthood and marital status. Studies of the effects of multiple roles (combining parenthood, marriage, and employment) have the drawback that parenthood is equated with currently having children in the home. The authors provide new evidence on the health of people who have reached old age, contrasting those with and without children, in an attempt to tease out the effects of parenthood, marital status, and gender. Data from Australia, Finland, and the Netherlands are used. Insofar as parenthood effects are found, they pertain to health behaviors (smoking, alcohol consumption, and physical exercise), providing evidence for the social control influences of parenthood.

Full text

http://jfi.sagepub.com
Journal of Family Issues
DOI: 10.1177/0192513X07303896
2007; 28; 1457 Journal of Family Issues
Hal Kendig, Pearl A. Dykstra, Ruben I. van Gaalen and Tuula Melkas
Health of Aging Parents and Childless Individuals
http://jfi.sagepub.com/cgi/content/abstract/28/11/1457
The online version of this article can be found at:
Published by:
http://www.sagepublications.com
can be found at:Journal of Family Issues Additional services and information for
http://jfi.sagepub.com/cgi/alerts Email Alerts:
http://jfi.sagepub.com/subscriptions Subscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://jfi.sagepub.com/cgi/content/refs/28/11/1457 Citations
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Health of Aging Parents
and Childless Individuals
Hal Kendig
The University of Sydney, Australia
Pearl A. Dykstra
Ruben I. van Gaalen
Netherlands Interdisciplinary Demographic Institute, The Hague
Utrecht University, Netherlands
Tuula Melkas
Statistics Finland, Helsinski
This article reviews and presents research findings on the relationships between
parenthood and health over the life span. Existing research shows lacunae. The
links between reproductive behavior and longevity generally focus on family
size rather than contrasting parents and nonparents. Studies of marital status
differentials in survival generally confound the effects of parenthood and marital
status. Studies of the effects of multiple roles (combining parenthood, marriage,
and employment) have the drawback that parenthood is equated with currently
having children in the home. The authors provide new evidence on the health
of people who have reached old age, contrasting those with and without children,
in an attempt to tease out the effects of parenthood, marital status, and gender.
Data from Australia, Finland, and the Netherlands are used. Insofar as parent-
hood effects are found, they pertain to health behaviors (smoking, alcohol con-
sumption, and physical exercise), providing evidence for the social control
influences of parenthood.
Keywords: childlessness; health; health behaviors; marital history; parent-
hood; social control
To gain insight into sources of inequality in society, we examine differ-
ences in health between aging parents and childless individuals. Insight
into the health-compromising and health-promoting aspects of parenthood as
it extends into later life will contribute to our knowledge about cumulative and
compounding influences over many decades of experience. Historical and
comparative perspectives can shed light on possible social structural influences
on health through the life span.
Journal of Family Issues
Volume 28 Number 11
November 2007 1457-1486
© 2007 Sage Publications
10.1177/0192513X07303896
http://jfi.sagepub.com
hosted at
http://online.sagepub.com
1457
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

In the first part of this article, we provide a critical review of relevant lit-
erature. Different bodies of research touch on issues linked with parenthood
differences in health; however, the implications of being childless for late-
life health are typically not addressed. We show that existing research has
had substantive foci that are of relevance to issues of childlessness; how-
ever, the question of health differences linked with nonparenthood has
rarely been addressed directly. We review studies from three research tradi-
tions: (a) reproductive history and longevity, (b) marital status and health,
and (c) multiple roles and health. In the second part of this article, we pre-
sent findings from three countries (Australia, Finland, and the Netherlands)
on the health of older parents and nonparents.
Review of Relevant Literature
Reproductive History
We start by reviewing studies of the links between reproductive history
and longevity. Three strands of theorizing can be identified in this body of
literature. The first involves the explanation of the U-shaped pattern that has
emerged in many studies of the association between parity and all-cause
mortality. Typically, these studies are based on people identified by their
demographic characteristics obtained from census data with follow-up infor-
mation from death registers. A consistent finding of these prospective stud-
ies is that ever-married women with no or only one child and those with five
or more children have relatively high mortality rates, whereas women with
two to four children have lower mortality rates (Hurt, Ronsmans, & Thomas,
2006). This pattern has been observed in the United States for ever-married
women age 45 years and older (Kitagawa & Hauser, 1973) and in two stud-
ies of women in Norway (Kvåle, Heuch, & Nilssen, 1994; Lund, Arnesen,
& Borgan, 1990). The relationship between parity and marital status was
1458 Journal of Family Issues
Authors’ Note: The findings reported in this article are from the 1992-1993 Australian Longitu-
dinal Study of Aging (ALSA), which was funded by the U.S. National Institute on Aging and the
South Australian Health Commission; the 1994 Health Status of Older People Survey, which was
funded by the Victorian Health Promotion Foundation; the 1994 Survey of Living Conditions in
Finland (SFLC) that was carried out by Statistics Finland; and the 1992-1993 Longitudinal Aging
Study Amsterdam (LASA), which is funded by the Dutch Ministry of Health, Welfare and Sport.
We would like to thank Marja Jylhä, Tanya Koropeckyj-Cox, and Renske Keizer for their thought-
ful suggestions. Please address correspondence to Hal Kendig, Research Professor of Ageing and
Health, Faculty of Health Sciences, University of Sydney, c/o CERA, Concord Hospital C25,
Concord, NSW 2139, Australia; e-mail: h.kendig@usyd.edu.au.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

also confirmed in England and Wales (Doblhammer, 2000; Green, Beral, &
Moser, 1988; Grundy & Tomassini, 2005) and Austria (Doblhammer, 2000).
Post hoc explanations advanced for the higher mortality rates of low-parity
women tend to center on selection effects as to who marries or has children
(Green et al., 1988; Kitagawa & Hauser, 1973). The most common argument
is that low parity is a sign of subfecundity or sterility, both of which in turn
are linked to poor health and higher mortality. Women might also restrict
their number of childbirths because of their health: The 1946 British cohort
study (Pless, Cripss, Davies, & Wadsworth, 1989) found that women who
had been chronically ill at ages 21 to 25 were less likely to have children
compared with their healthy counterparts. A posited causal effect is that the
higher mortality rates of high-parity women may arise from the adverse
physiological effects of multiple pregnancies (Kvåle et al., 1994). Green et
al. (1988) provided examples of health conditions in which risks are height-
ened by pregnancy (see also Beral, 1985; Friedlander, 1996). It is known
that pregnancy can trigger diabetes (starting with gestational diabetes) and
degenerative changes in arterial walls; repeated pregnancies might increase
the risk of ischemic heart disease. Toxemia of pregnancy might also make
women more susceptible to hypertension and its complications and its ongo-
ing effects.
The excess mortality of high-parity women is consistent with a second
strand of theorizing, which takes an evolutionary framework as its point of
departure. The disposable soma theory on the evolution of aging (Kirkwood,
1977; Kirkwood & Holliday, 1979) states that there is a trade-off between
longevity and reproduction. Longevity involves significant metabolic costs
in the form of investments in somatic maintenance. Reproduction makes
use of resources that could be used for the maintenance and repair of cells
and thus competes for resources, with faster aging as a result. Using histor-
ical, genealogical data on ever-married men and women of the British aris-
tocracy born between 700 and 1876, Westendorp and Kirkwood (1998)
demonstrated that there was an inverse relationship between the age at death
and the number of children ever born among women who lived to age 60
and older. It is surprising to note that a similar pattern was observed for
men. Another intriguing finding was that almost half of the married women
who survived to age 81 or older had no children at all. Although these find-
ings are consistent with the disposable soma theory, they do not go very far
in suggesting specific explanatory mechanisms as to disease pathways that
could affect differential survival.
The Westendorp and Kirkwood (1998) study has been subject to criticism
(Doblhammer & Oeppen, 2003; Gavrilova & Gavrilova, 1999; Le Bourg,
Kendig et al. / Health of Parents and Childless Individuals 1459
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

2001). In their analysis of a more complete genealogy of the British aris-
tocracy than the one used by Westendorp and Kirkwood, Doblhammer and
Oeppen (2003) confirmed that the number of children was positively
associated with mortality levels for women who lived to age 50 and older;
however, they found no statistically significant differences for men. Dribe
(2004), using family reconstitution data from West Sweden for the period
1766 to 1895, also reported a negative influence of parity on longevity after
age 50 among married women. Again, no associations between the number
of children ever born and longevity were found for men. These historical
studies pertained to time periods prior to the demographic and epidemio-
logical revolutions that led to sharp reductions in childbearing rates and
death rates.
The inverse relationship between parity and longevity has also been reported
for contemporary populations. Survivorship data of men and women born
between 1880 and 1929 who had participated in the Rancho Bernardo Heart
and Chronic Disease Survey showed that women with relatively more
children born had higher mortality than those who were childless or had few
children (Friedlander, 1996). Among men, no association between reproduc-
tion and survivorship was found. It is possible, however, that the findings
might be specific to this exclusive community in Southern California where
the people were relatively affluent and had high levels of childlessness
(more than 30%).
A third strand of theorizing focuses on the excess mortality of childless
and low-parity women in contemporary populations. As described previ-
ously, one possible mechanism may be the role of health as a selection factor
in marrying or having children. An alternative explanation draws attention
to the hormonal protection against genital and other sex-specific cancers that
is provided by pregnancy: In a prospective study of Norwegian women, the
risk of cancers of the breast, uterus, and ovary was shown to be inversely
related to the number of births (Kvåle et al., 1994). A British epidemiologi-
cal study also reported higher mortality from breast, ovarian, and endome-
trial cancer among ever-married childless women compared to ever-married
mothers (Beral, 1985). Although multiple pregnancies may have a negative
impact on health by triggering dormant pathological conditions, hormonal
changes associated with pregnancy may have a positive impact by reducing
the risk of specific types of cancer.
In summary, studies based on 20th-century populations show that all-
cause mortality is higher among childless women than among mothers, with
the exception of women with relatively large families (five or more children).
Health selection and poorer protection against hormonally related cancers
1460 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

are the dominant explanations advanced for the excess mortality among
childless women. However, research on the associations between reproduc-
tive history and longevity is generally restricted to ever-married women.
The interrelations of reproductive and marital histories and their impacts
on mortality are generally not considered. Relatively few studies have
included men in their samples. Another drawback is that a focus on number
of children can aggregate the childless in a low-parity group. A better under-
standing of the interlinkages between reproductive history and longevity
requires individual-level data on full fertility histories and full medical histo-
ries through time.
Marital Status
The second body of literature that we consider involves marital status
differentials in health (e.g., Ben-Shlomo, Smith, Shipley, & Marmot, 1993;
Joung, 1996; Verbrugge, 1979; Waite & Gallagher, 2000), often using mor-
tality as the outcome measure. However, as marriage and parenthood are
inherently linked, it is difficult to disentangle them. Usually, parenthood is
examined only indirectly.
One of the most influential contributions in this literature is Gove’s (1973)
article in the American Journal of Sociology. The basic premise is that varia-
tions in mortality rates arise from the psychological states and lifestyles asso-
ciated with the marital role. Gove argued that the married, compared to the
unmarried, (a) have higher psychological well-being and thus a lower like-
lihood of death by suicide, homicide, and accidents; (b) are less likely to engage
in self-destructive acts and thus are less likely to die from, for example, alcohol-
related causes; and (c) are more willing and able to undergo the treatment
required for diseases such as tuberculosis and diabetes. Using data from the
U.S. National Center for Health Statistics, Gove showed that for each of these
causes of death, the married had substantially lower mortality rates than the
unmarried.
The marital status differences in mortality were found by Gove (1973) to
be systematically larger for men than women (see also Berkman & Syme,
1979; Brockmann & Klein, 2004; Cheung, 2000; Hu & Goldman, 1990;
Joung, 1996; Lillard & Waite, 1995). Among men, the formerly married had
higher mortality rates than the never married. Among women, the mortality
rates of the formerly and never married were quite similar. The marital status
disparities were greatest in the age 25 to 44 range. Gove noted that these ages
correspond closely to the time when most families have young children in the
home and, following Durkheim (1896/1951), suggested that having children
Kendig et al. / Health of Parents and Childless Individuals 1461
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

might provide a form of protection through their effect on the concerns and
behavior of parents.
The explanations offered for marital status differentials in health broadly
divide into those based on health selection and those related to social causation
(e.g., Joung, 1996; Kobrin & Hendershot, 1977; Wyke & Ford, 1992). Health
selection implies that people with health problems (and fewer resources gen-
erally) have lower probabilities of marrying, staying married, and remarrying.
Health differences between the married and the unmarried may therefore
not be attributable to marriage per se but rather to the characteristics of the
people who are likely to marry and to remain married. Social causation
implies that marriage has a positive protective effect and that the health
advantages of the married are attributable to better material resources,
avoidance of risky health behavior, better coping resources, a lower suscep-
tibility to stress, and higher levels of available support. For the formerly mar-
ried, the stress accompanying the breakup or loss of the relationship also can
have adverse health consequences. Research findings indicate that selection
and causation explain part of the marital status differences in health and psy-
chological well-being (Brockmann & Klein, 2004; Joung, 1996; Mastekaasa,
1992; Simon, 2002; Weitoft, Burström, & Rosén, 2004).
The design of health studies seldom allows us to disentangle parenthood
and marital status effects. Notable exceptions do exist, however. For example,
Høyer and Lund (1993) linked data from the 1970 census in Norway to death
certificates for the years 1970 to 1985. Information on parity was available
only for the current marriage of women at the time of the census. The authors
restricted their analyses to suicide among women who were either never-
married or married at baseline, and age 25 and older. Among those age 25 to
64 at baseline, the risk of death from suicide was higher among never-mar-
ried women than among married women, regardless of whether they had
children. Among married women, childlessness was associated with a higher
rate of suicide in each age group. However, in the age group of 65 to 74 at
baseline, there was an apparent “crossover” effect with the suicide mortality
rate of never-married women being lower than that of childless married
women. Høyer and Lund’s findings suggest that childless women are more
prone to suicide and the more so if they are never-married and younger than
age 65.
Combining data from the 1985 and 1990 censuses, national health
registers, and a register of known biological relations between parents and
children, Weitoft et al. (2004) distinguished the effects of ever having
fathered a child, living with one or more biological children, and living with
a partner on the longevity of Swedish men age 29 to 54. The highest all-
cause mortality was observed among single noncustodial fathers, closely
1462 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

followed by single childless men. These differences were only partially
attributable to health selection (as indicated by the men’s previous inpatient
history) and to differences in socioeconomic status. A reduced risk of all-
cause mortality was also observed among childless men living with a partner,
again after controlling for health selection effects and socioeconomic status.
The lowest all-cause mortality was observed among fathers living with their
children and a partner. Singlehood, never having had children, and not living
with one’s biological children had negative impacts on longevity.
In summary, the effects of marital status on health are confounded by
other indicators of social position such as parental status, living arrangement
(the presence of an adult and/or children in the household), and socioeco-
nomic status. An analysis that does not unravel these characteristics can
lead to misleading conclusions about the extent and nature of differences in
health. The few studies with designs that allow the disaggregation of effects
show that permanent childlessness has a negative impact on longevity inde-
pendent of marital status and socioeconomic status. Given the nature of the
data sets with which they work, researchers in the marital status health dif-
ferentials tradition tend to focus on establishing effects, rather than investi-
gating potential mechanisms. Analyses of differences in morbidity and
mortality should consider not only sociostructural states but also determi-
nants in daily living such as stress, health practices, and access to support
and care.
Multiple Roles
A third body of research views parenthood as one of the multiple roles
people might occupy. The focus is on having the responsibility for children
as they grow up, typically measured as having children in the home. A draw-
back of this measure is that empty nesters and people who have not (yet)
become parents fall into the same category. The aim of this line of research
is to find out whether it is harmful or beneficial to combine the roles of parent,
spouse, and paid worker. Most of the research has focused on women; how-
ever, increasingly, men are being included in research designs (Arber, 1991;
Hibbard & Pope, 1991; Mastekaasa, 2000). As Macintyre (1992) pointed
out in her review article, there are many more studies of the health effects
of parenthood and employment among women than among men, presumably
at least in part because “being employed and a father are seen as normative
and unproblematic” (p. 461).
Competing hypotheses have been put forward regarding the effects of
multiple roles on health (Marks, 1977; Sieber, 1974). Are they a “double burden”
or a “double blessing” (Weatherall, Joshi, & Macran, 1994)? The role strain
Kendig et al. / Health of Parents and Childless Individuals 1463
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

hypothesis (Marks, 1977) suggests decreased well-being, given the demands
of incompatible roles. For example, role overload may result from accumu-
lating too many responsibilities given available time and energy. Role con-
flict describes the difficulties involved in attempting to meet competing or
contradictory expectations and obligations, irrespective of time pressures.
The idea is that role overload and role conflict result in fatigue and exhaus-
tion and ultimately contribute to disease and death. The role accumulation
(Sieber, 1974) or role expansion hypothesis (Marks, 1977), on the other
hand, suggests increased well-being resulting from multiple roles. People ful-
filling multiple roles gain sources of self-esteem and identity. Employment
has the benefits of providing financial independence and access to social
contacts outside the home and marriage. Presumably, role accumulation
includes improved coping resources, a healthier lifestyle, and better disease
resistance.
The empirical evidence favors the role accumulation hypothesis: The
health of women who are married, who have children, and who are employed
is good compared to other groups of women (Arber, 1991; Fokkema, 2002;
Hibbard & Pope, 1991; Martikainen, 1995; McMunn, Bartley, Hardy, &
Kuh, 2006; Verbrugge, 1983; Waldron, Weiss, & Hughes, 1998; Weatherall
et al., 1994). Note that the effects of multiple roles are not of primary inter-
est to us. Instead, we focus on what we can learn from these studies about the
role of parenthood specifically and its links with health.
Verbrugge (1983) was among the first to examine the singular and com-
bined effects of marriage, parenthood, and employment on health. Using a
wide range of subjective and objective health measures, she analyzed data
from the 1978 Health in Detroit Study, a sample of White men and women
age 18 and older. Defining parent status as having one or more children in
the home, Verbrugge found that parenthood had the weakest health bene-
fits, whereas employment had the strongest. Each role in and of itself con-
tributed positively to health. The combination of roles conferred no special
additional health advantage or disadvantage. The health benefits of mar-
riage, employment, and parenthood were similar for men and women. Given
the cross-sectional nature of the data, Verbrugge was unable to unravel selec-
tion and causation effects.
Another study, which also included men, was carried out by Hibbard and
Pope (1991). Their data were from a 15-year follow-up of a sample of
members of a large Health Maintenance Organization (HMO) in the United
States who had been studied in 1970 to 1971. Hibbard and Pope controlled
for self-reported health at baseline to deal with the problem of the “healthy
worker” effect (i.e., the possibility that healthier people were more likely to
1464 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

enter the workforce and stay in it). Parental status, measured as having one
child or more younger than age 19 in the home, was unrelated to any of the
morbidity and mortality measures among men. Among employed women
(controlling for marital status), having no children at home increased the
risk of ischemic heart disease but had no effect on other serious morbidities
or on all-cause mortality. Among nonemployed women (controlling for mar-
ital status), parental status had no association with death or morbidity.
Although Hibbard and Pope (1991) reported few links between parental
status and health, Martikainen (1995) showed, in a 5-year follow-up of
Finnish women who were age 35 to 64 at the time of the 1980 census, that
mortality for most cancers, vascular diseases, “other” diseases, and acci-
dents and violence was higher among women who had no children at home
than mothers. This parental status difference remained after controlling for
marital status, labor force participation, occupational status, and educational
attainment.
A study by Weatherall et al. (1994) is unusual in the multiple roles liter-
ature because of its focus on ever having had children rather than having
children at home. Using a British database that matched 1971 census records
and death records, they found higher mortality risks for childless women and
for women whose youngest child was age 15 and older among married
women younger than age 60 in 1971. No interaction with employment status
was observed. In addition, they demonstrated health selection into mother-
hood in an analysis that matched data from the 1971 and 1981 census returns.
Age-adjusted mortality was highest for women who were childless at both
time points, and it was markedly lower for women who became mothers for
the first time in the interval between the two censuses. The authors conclude
that “health problems, which are later manifest as mortality, may bar entry
to motherhood” (p. 295).
Elstad (1996) examined changes over time in the effects of multiple roles
on women’s health, as indicated by having any long-standing diseases. He
used data from five national surveys conducted by Statistics Norway between
1968 and 1991, focusing on women age 31 to 60 at the time of data collec-
tion. Parental status was measured as having at least one child younger than
age 17 in the household. This measure excluded noncustodial children but
included stepchildren. Those who had more children had fewer health prob-
lems. Whereas this relationship had not changed over time, the effect on
health of having no children at home and being in full-time employment did.
The negative effect observed in the older data had disappeared in the sur-
veys conducted in the late 1980s. In Elstad’s view, this finding suggests an
increased importance of paid employment in women’s lives. He argued that
Kendig et al. / Health of Parents and Childless Individuals 1465
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

paid employment appears to be protective of health, even when women have
other characteristics that usually confer health disadvantage, such as being
childless. Elstad also drew attention to health selection into work, arguing
that good health has possibly become a more explicit criterion when women
enter paid employment.
The centrality of work orientation also emerged in a more recent study
using two Norwegian data sets (Mastekaasa, 2000). Both of these cross-
sectional samples were restricted to employed 18- to 59-year-olds, and
health was measured in terms of sick leave. Parental status was indicated by
the number and ages of children. Only children younger than age 16 were
considered. In the first data set (1995 national registry data of employment
relations), involving a sample of married men and women, he found few
parental differences. The few observed differences pertained to parents of
infants, who were more likely to be absent from work than the childless and
parents of older children. Their sickness absence was largely because of
respiratory conditions, probably involving infections contracted through
their children.
The second Norwegian data set was formed from the 1990 census supple-
mented with information from public registers. Married, cohabiting, never-
married, and previously married single women were distinguished. Sick leave
was low among single never-married childless women, a group that might
be particularly committed to work. Sick leave was relatively high among
young married women, a group that might be particularly family oriented and
correspondingly less work oriented. In this sample of Norwegian women,
work orientation appears to be the relevant determinant of health rather than
marital or parental status.
In summary, the picture that emerges from these studies is that women
who have no children at home tend to have poorer health. Parental status
appears to make less of a difference in terms of health than does employ-
ment status. Among the limitations of this research for gaining insight into
the effects of childlessness on late-life health are that (a) lifelong childless-
ness and no longer having children living at home are confounded, (b) there
is a focus on women in midlife and so one learns little about men and about
older women, and (c) apart from the issue of selection versus causation, little
insight is provided into the mechanisms that might underlie the salutary
effects of parenthood.
Social Networks
In our view, it is useful to turn to the social networks literature for insights
into how and why parenthood might affect health. Large, well-controlled
1466 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

prospective studies have repeatedly shown that social relationships have an
impact on health independent of potentially confounded factors such as socio-
economic status, prior health status, use of health services, and personality
(Uchino, 2004). There are several pathways by which social relationships
are posited to affect well-being (Berkman, Glass, Brissette, & Seeman, 2000).
The first is through social support,which involves behavioral exchanges
(giving and receiving) that are intended as helpful and are perceived as such.
Second, networks provide opportunities for social engagement and com-
panionship. Shared leisure activities serve as a source of pleasure and stimu-
lation, whereas the participation in meaningful community activities brings
social recognition and feelings of belonging. Social control is a third mech-
anism that operates directly on health when network members consciously
attempt to modify a person’s health behavior or indirectly when people inter-
nalize norms for healthful behavior. Fourth, relationships provide access to
resources that transcend an individual’s means. To have relationships is to
have access to other people’s connections, information, money, and time. A
fifth pathway by which networks may influence health is by person-to-person
contact resulting in disease transmission. The same network characteristics
than can be health promoting can be health damaging if they serve to expose
people to infectious agents.
We know that adult children serve several of the previously described
functions in their aging parents’ lives. In contrast to the “demands of par-
enthood” view that dominates the literature on earlier adulthood, the litera-
ture on late life emphasizes the “rewards of parenthood.” Study after study
has documented the centrality of adult children in the social networks of
aging parents as providers of support, bridges to social services, monitors of
health behavior, vehicles to new social circles, and so on (e.g., Choi, 1994;
Kendig, 1996; Phillipson, Bernard, Phillips, & Ogg, 2001; Wenger, 1984;
see also the article by Wenger, Dykstra, Melkas, & Knipscheer, 2007 [this
issue]). The importance of adult children in older parents’ lives has given
rise to the question of what happens to older people who have no children
on whom they can rely (Kreager, 2004; Wenger, Scott, & Patterson, 2000).
Generalizability of Findings to Older Adults
As the review of the literature has revealed, childless men and women
generally have poorer health than parents, though the differences are small.
Most of the findings pertain to midlife adults; little research has compared
the health of older parents and nonparents. As yet, it is not clear whether
parenthood differences in health also exist at advanced ages. Two issues are
relevant here.
Kendig et al. / Health of Parents and Childless Individuals 1467
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

The first is whether parenthood differences that are observed in midlife
will continue to exist in late life. There is the possibility that survival through
midlife may lead to selection effects in which only the healthier individ-
uals move into old age and progressively through later life. However, if mar-
ital status health differentials serve as an indication, there is reason to
suspect that family variables will have continuing effects to the end of life.
Even at very advanced ages, marital status has been found to predict longevity
(Martelin, Koskinen, & Valkonen, 1998; Shye, Mullooly, Freeborn, & Pope,
1995). In their review article, Alwin and Wray (2005) pointed out that the
literature is quite consistent in the finding that socioeconomic and sociode-
mographic inequalities in health persist across the life course. The cumulative
advantages and disadvantages literature shows increasing disparities across
social categories with increasing age (Crystal & Shea, 2003; Dannefer, 2003;
O’Rand, 1996).
The second issue is whether findings based on younger cohorts can be
generalized to older cohorts. More particularly, the question is whether the
opportunities to enter marriage or parenthood have been similar for younger
and older cohorts. Women born in the first decades of the 20th century faced
relatively greater impediments to combining professional aspirations with
family responsibilities. Those who remained unmarried at that time were
relatively resourceful women (Bernard, 1982; see also the article by Hagestad
& Call, 2007, part one of this special issue). The rise in women’s employ-
ment has made men’s role as principal providers for the family less impor-
tant (Kalmijn, 1998; Sweeney, 2002). Successive cohorts have been more
likely to marry and become parents during the course of the past century. This
suggests that health differentials between the married and the unmarried, and
between the parents and the childless, may have diminished.
Clearly, there are opposing views regarding the applicability to older
people (and earlier cohorts) of findings from people in early and midlife
showing marital and parenthood differentials in health. It remains an open
question as to how differences in health between parents and childless
individuals may accumulate through the life course and vary for successive
cohorts.
Empirical Analysis of Health Differences
In the remainder of this article, we examine differences in health between
older people with and without children. In addition to analyzing differences
in physical and mental health outcomes, we look at the health behaviors that
1468 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

may contribute to these outcomes. The focus on health behaviors is based on
the notion that parents have been (and remain) subject to informal social
control for much of their adult lives. Presumably, the influence of social con-
trol is evident in exhibited behavior, in this case, refraining from damaging
health behaviors and engaging in a health-promoting lifestyle.
Several authors provided reasons why parents would have more health-
ful behaviors than childless individuals. One idea is that parenthood is a
transforming event (Akerlof, 1998; Eggebeen & Knoester, 2001). When people
become parents, their behavior changes. They act more responsibly, become
more caring, and wish to be good role models (cf. Backett & Davison, 1995).
Another idea is that living with others (a partner, children) requires regulation
of activities and a division of responsibilities and obligations (Umberson,
1987). Sharing a household implies being subject to informal pressure
toward regularity of habits (Anson, 1989): sleeping times, meal times, and
avoidance of adverse exposures imposed on others (e.g., smoking). Having
day-to-day obligations toward others means people can less easily afford to
enter the sick role (Gove & Hughes, 1979). Older people report that spouses
and adult children are positive influences that facilitate physical activity,
social activity, and healthy eating (Kendig, 1996).
The links between health behaviors (e.g., smoking, alcohol consumption,
regularity of meals, sleeping habits, exercise) and health outcomes have
been examined extensively elsewhere (Belloc & Breslow, 1972; Berkman,
Breslow, & Wingard, 1983; Camacho & Wiley, 1983; Huijbregts et al.,
1997; Joung, Stronks, van de Mheen, & Mackenbach, 1995; Wyke & Ford,
1992). Here, we restrict ourselves to the first part of the causal chain, from
sociodemographic predictors to health behaviors. The second part of
the causal chain, running from health behaviors to health outcomes, is not
considered here.
Our working hypothesis is that compared to parents, childless individ-
uals will have poorer physical and mental health and exhibit fewer positive
and more negative health behaviors. Implicit in our thinking is that late-life
health is a continuation of patterns established earlier in life. It is expected
that the effects of family circumstances will be ongoing even after children
have left the parental home.
Data Sources
We use survey data from three countries: Australia, Finland, and the
Netherlands. Although the surveys were not designed to examine the impact
of childlessness on late-life health, the data sets have comparable (though
not identical) questions on relevant variables.
Kendig et al. / Health of Parents and Childless Individuals 1469
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

The Australian data are from two surveys, which were combined to
obtain a sufficiently large number of childless older adults. We analyzed
data from the first wave of the Australian (Adelaide) Longitudinal Study of
Aging (ALSA), which was designed to assess how psychosocial, biomedical,
lifestyle, and environmental factors are associated with health and well-
being in late life. Respondents (N=1,947) were between the ages of 70 and
85 at the time of the first interview. The sample was drawn from the State
Electoral Data Base (with a response of 60%), stratified by gender and 5-year
age bands, and included community-dwelling and institutionalized individ-
uals. Wave 1 (1992-1993) included a comprehensive face-to-face interview
using computer-assisted interviewing, a physiological assessment, and self-
completion questionnaires. Details on the study can be found in Andrews,
Cheok, and Carr (1989) and in Andrews and Myers (2000).
The second Australian data set is from the Health Status of Older People
(HSOP) survey, which was conducted in the Melbourne metropolitan area in
1994. Questions in the survey concentrated on the following topics: health-
related actions, health history, functional health, perceived quality of life,
service use, transport, and neighborhood. The survey instruments included
a face-to-face interview, a brief physical examination, and a self-completion
questionnaire. Respondents (N=1,000) were age 65 and older at the time of
the interview and were living in private dwellings. The response rate was
70% for full interviews. Details on the study can be found in Kendig et al.
(1996).
The Finnish data are from the 1994 Survey of Living Conditions in
Finland (SFLC), which was part of a series of periodic surveys carried out
by Statistics Finland on the living conditions of noninstitutionalized adults.
Personal interviews were conducted with people age 15 and older (N=
8,650). The systematic random sample was stratified by home municipal-
ity, occupational class, and income. The response rate among persons age
55 to 74 was 73%, and 66% among those age 75 and older. The analyses
reported in this article are based on the 1,551 respondents who were age 65
or older at the time of data collection. Background information on the sur-
vey is provided in Ahola, Djerf, Heiskanen, and Vikki (1995) and on the
Web site of Statistics Finland (http://www.stat.fi/).
The Dutch data come from the first wave (1992-1993) of the Longitu-
dinal Aging Study Amsterdam (LASA). The purpose of LASA was to gain
insight into the predictors and consequences of changes in physical, cogni-
tive, emotional, and social functioning in older persons. The sample, which
was stratified by gender and 5-year cohort bands (1908-1937 birth cohorts),
was drawn from the population registers of the city of Amsterdam and two
1470 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

rural communities in the western part of the Netherlands, one city and four
rural communities in the south, and one city and four rural communities in
the northeast. Older adults in private households and in institutions were
included in the sample. Respondents (N=3,107) were between the ages of
55 and 85 at the time of the interview. The analyses reported in this article
are based on the 2,141 respondents age 65 and older. Face-to-face inter-
views were conducted using laptop computers. Respondents also filled in
a self-completion questionnaire, and a separate visit was made by a
nurse–interviewer for clinical measurements. The response rate was 51% for
the interviews. More detailed information can be found on the LASA home page
(http://www.lasa-vu.nl/) and in Deeg and Westendorp-De Serière (1994).
Analyses
Three sets of health outcomes were considered to examine the influ-
ence of marital and parenting experiences. We selected measures that were
highly similar across the surveys. The first set pertained to physical health
and included self-rated general health a rating of one’s health compared to
peers, functional capacity, pain, and the use of prescription medicine. The
second set pertained to mental health and included measures of depression
and difficulty of falling asleep. The third set pertained to health behaviors
(fruit consumption, smoking, alcohol consumption, and physical exercise).
To avoid what Connidis and McMullin (1993) referred to as the “all-or-
nothing” approach to research on the impact of parent status in older age,
where differentiations among childless individuals and among parents are
overlooked, we separate the effects of parental and marital status.Respondents
were categorized as childless if they had no living biological or adoptive
children. Three marital status categories were distinguished: never married,
formerly married, and currently married. The “currently married” category
included a small number of individuals in consensual unions. Given the small
numbers of divorcées, they were combined with widows into a “formerly
married” category.
Findings are presented separately for men and women, in each of the mari-
tal and parenting groups, given the potential influence of gender on health
status (see Table 1 for the numbers of Australian, Finnish, and Dutch respon-
dents in each of the groups). The article by Koropeckyj-Cox and Call (2007
[this issue]) shows that in these data sets relatively more men are currently mar-
ried and relatively more women are either previously married or never married.
The analyses applied controls for age and educational attainment given
their potential to independently affect health and their importance for selection
into the marital and/or parental status categories. The level of educational
Kendig et al. / Health of Parents and Childless Individuals 1471
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

1472
Table 1
Mean Age and Years of Education of Older Adults in Three Countries, Categorized by
Gender and Parental and/or Marital Status
Childless Men Fathers Childless Women Mothers
Never Formerly Currently Formerly Currently Never Formerly Currently Formerly Currently
Married Married Married Married Married Married Married Married Married Married
N
Australia 56 31 66 322 1,048 66 70 57 596 774
Finland 41 19 41 104 413 92 54 28 484 274
Netherlands 56 34 76 208 674 79 70 55 497 392
Age
Australia 74.3 80.4 78.3 80.7 75.8 80.5 79.7 76.2 78.5 73.8
Finland 69.4 73.6 69.5 71.1 68.4 73.1 77.1 69.5 73.4 67.6
Netherlands 80.0 78.0 70.7 74.8 69.7 72.2 75.7 68.5 74.3 66.8
Years of education
Australia 8.9 8.5 8.8 8.0 9.0 9.7 9.1 9.1 8.9 8.7
Finland 6.6 6.1 7.2 6.1 7.7 8.0 5.1 7.2 5.8 7.4
Netherlands 8.6 8.0 10.2 8.7 9.7 10.4 8.7 8.9 7.7 8.0
Source: 1992-1993 Australian Longitudinal Study of Aging (ALSA; see Andrews, Cheok, & Carr, 1989; Andrews & Myers, 2000) combined with
the 1994 Health Status of Older People (HSOP) survey (see Kendig et al., 1996); 1994 Survey of Living Conditions in Finland (SFLC; see Ahola,
Djerf, Heiskanen, & Vikki, 1995); and 1992-1993 Longitudinal Aging Study Amsterdam (LASA; see Deeg & Westendorp-De Serière, 1994).
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

attainment was measured in terms of the number of years that are generally
required to complete a certain level of schooling. As Table 1 shows, the for-
merly married were generally older than the currently married. Never-married
Australian women and never-married Dutch men also had reached rela-
tively advanced ages. Among men, currently married fathers had the high-
est level of education, followed by those who were currently married and
childless. Among women across the three countries, those who had never
married and remained childless consistently had the highest level of educa-
tion, followed by those who were currently married and childless. (Differences
in educational attainment are described in greater detail in the article by
Koropeckyj-Cox & Call, 2007.)
ANOVAs were carried out with parental status (has living children or not),
marital status (never married, formerly married, and currently married), and
gender as determinants. Age and level of educational attainment served as
covariates. Health status was examined by means of one-way ANOVA
using 10 gender/marital/parental status categories. Differences described in
the text below are significant at least at the .05 level. In the tables, we pre-
sent means adjusted for the effects of age and level of educational attainment.
The findings are presented for five categories of adults: never married, for-
merly married and currently married childless older adults, and formerly
married and currently married parents. In addition, they are presented sepa-
rately for men and women and for the three countries.
Results
Physical Health
Table 2 shows differences in self-reports of physical health, adjusted for
age and years of education. Across the three countries, few consistent dif-
ferences were observed regarding the reports of poor (including “fair”) self-
rated health. In Australia and Finland, formerly married childless men were
most likely to report poor health, whereas in the Netherlands they were
least likely to report poor health. In the Netherlands, never-married childless
men were most likely to report poor health, whereas this group reported
favorably about their health in Finland. Among Finnish women, the never-
married childless and currently married mothers were least likely to report
poor health.
The next indicator of general health was whether respondents perceived
they were fitter than age peers. The pattern of findings is quite similar for
Kendig et al. / Health of Parents and Childless Individuals 1473
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

1474
Table 2
Physical Health of Older Adults in Three Countries, Categorized by
Gender and Parental and/or Marital Status (in percentages)
Childless Men Fathers Childless Women Mothers
Never Formerly Currently Formerly Currently Never Formerly Currently Formerly Currently
Married Married Married Married Married Married Married Married Married Married
General health
Australia (fair or poor) 32 39 26 28 28 22 21 31 23 29
Finland (fairly bad or bad) 17 50 25 27 24 15 37 38 31 13
Netherlands (fair or poor) 42 11 18 28 31 34 31 35 42 36
Health compared to peers
Australia (more fit) 46 69 54 62 62 66 69 45 56 54
Finland (no data)
Netherlands (more fit) 21 67 24 38 33 28 27 18 31 29
Functional capacity
Australia (no ADL 79 83 92 83 68 62 75 67 47 68
problems)
Finland (no data)
Netherlands (no ADL 45 57 64 62 68 50 52 55 54 62
problems)
Pain
Australia (no pain) 57 58 58 64 62 44 59 62 56 68
Finland (no pain) 42 36 45 35 34 30 30 61 37 29
Netherlands (no data)
Prescription medicine
Australia (yes) 89 84 87 82 82 86 81 92 83 87
Finland (yes) 60 86 95 71 67 80 82 85 81 76
Netherlands (yes) 73 59 56 63 63 70 70 66 70 65
Source: Please see Table 1.
Note: ADL =activities of daily living. The proportions are adjusted for differences in age and years of education.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Australia and the Netherlands. The favorable ratings of formerly married
childless men stand out. Of all categories of men, they were most likely to
perceive they were fitter than age peers. Never-married childless men were
least likely to perceive they were fitter than their peers. Among women, cur-
rently married childless women were least likely to report they were fitter
than their peers. This question was not in the Finnish survey.
Functional capacity, measured in terms of having no limitations with the
activities of daily living (ADL), was the third indicator of general health.
Finnish data on ADL limitations were not available. Findings were not con-
sistent across Australia and the Netherlands. The Australian data showed a
main effect for gender: Women were more likely to have ADL limitations
than men. Of all the groups distinguished in the Australian analyses, cur-
rently married childless men were least likely to have ADL limitations,
whereas formerly married mothers were most likely to have ADL limita-
tions. In the Dutch data, using a less stringent measure, never-married
childless men were least likely to have ADL limitations. Other differences
between the groups in the Dutch analyses were not significant.
The absence of pain was the fourth measure of general health. Dutch data
on pain were not available. In Australia and Finland, few differences across
the distinguished gender/marital/partner status categories were observed. In
the Australian data, never-married childless women stood out, with a high
likelihood of reporting pain. In the Finnish data, currently married childless
women were distinctly less likely to report pain.
The last measure of general health was the use of prescription medicine.
Neither the Australian nor the Dutch data showed any significant gender,
marital status, or parental status differences in the likelihood of using pre-
scription medicine. Few significant differences were found in the Finnish
data as well, with one exception: Never-married childless men were least
likely to report using prescription medicine.
Mental Health
We turn next to our measures of mental health. As Table 3 shows, diffi-
culty falling asleep was clearly more frequent among women than men.
This pattern was observed in each of the three countries. In Australia and in
the Netherlands, never-married childless women were less likely to report
sleeping problems than women in other marital and/or parental status cate-
gories. In Finland, formerly married childless men were most likely to report
difficulty falling asleep, whereas formerly married childless women in the
Netherlands were the most likely to report sleep problems.
Kendig et al. / Health of Parents and Childless Individuals 1475
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

1476
Table 3
Mental Health of Older Adults in Three Countries, Categorized by
Gender and Parental and/or Marital Status (in percentages)
Childless Men Fathers Childless Women Mothers
Never Formerly Currently Formerly Currently Never Formerly Currently Formerly Currently
Married Married Married Married Married Married Married Married Married Married
Falling asleep
Australia (trouble) 17 9 15 18 13 21 27 25 31 26
Finland (difficulties) 25 50 30 27 31 49 37 46 47 44
Netherlands (often 15 15 22 15 10 19 35 23 25 23
difficulties)
Depression
Australia (above mean) 48 39 43 42 32 42 50 30 47 38
Finland (yes) 8 57 30 34 28 37 41 46 42 33
Netherlands (above cutoff) 25 23 8 21 7 18 19 14 23 12
Source: Please see Table 1.
Note: The proportions are adjusted for differences in age and years of education. See text for differences in the depression measures.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Measures of depression differed between the surveys. In the Australian
and Dutch surveys, depressive symptoms were measured using the Center
for Epidemiological Studies Depression Scale (Radloff, 1977). In the Finnish
survey, a one-item question assessed whether respondents felt low spirited
or depressed. Being formerly married was associated with a greater likeli-
hood of depression in all three countries. Consistent parental status differ-
ences and gender differences were not observed. Although never-married
men were least likely to report depressive feelings in Finland, they were
highly likely to do so in Australia. Formerly married childless men in Finland
and formerly married childless women in Australia were also highly likely
to report feelings of depression. Finally, currently married Dutch men, regard-
less of parental status, were unlikely to report depression.
Health Behaviors
The first indicator of positive health behavior was the daily consumption
of fruit in winter (see Table 4), included in the Australian and Dutch but not
the Finnish survey. In the Australian and Dutch data, a clear gender differ-
ence emerged: Men were less likely to eat fruit daily than women. In both
countries, formerly married fathers were least likely of all to consume fruit
on a daily basis in winter. In the Netherlands, never-married men were also
unlikely to eat fruit daily.
Each survey included information on smoking behavior. Gender and
marital status differences were found in Finland and the Netherlands but
not in Australia. Overall, Australians were least likely to report being cur-
rent smokers. In Finland and the Netherlands, men were more likely to be
current smokers than women, and the unmarried were more likely to be cur-
rent smokers than the married. In both countries, parenthood differences in
the likelihood of being a smoker were observed; however, the pattern dif-
fered by gender. In Finland, childless men were more likely to be smokers
than fathers, whereas in the Netherlands, childless women were more likely
to be smokers than mothers. In all three countries, the likelihood of being a
current smoker was highest among never-married childless men.
The fourth health behavior considered in the analyses was the daily con-
sumption of alcohol. Daily alcohol consumption levels were quite similar in
Australia and the Netherlands but much lower in Finland. Policy restrictions
(the state has a monopoly on the sale of alcohol) might account for why
Finnish older adults are less likely to drink alcohol daily than Australian and
Dutch. In each country, men were more likely to consume alcohol daily than
women. In Australia, high rates of daily alcohol consumption were also observed
Kendig et al. / Health of Parents and Childless Individuals 1477
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

1478
Table 4
Health Behaviors of Older Adults in Three Countries, Categorized by
Gender and Parental and/or Marital Status (in percentages)
Childless Men Fathers Childless Women Mothers
Never Formerly Currently Formerly Currently Never Formerly Currently Formerly Currently
Married Married Married Married Married Married Married Married Married Married
Fruit consumption in winter
Australia (daily) 62 65 60 49 64 80 75 57 69 76
Finland (no data)
Netherlands (daily) 49 64 52 48 63 87 79 87 81 75
Current smoker
Australia (yes) 15 8 8 12 8 8 13 9 10 7
Finland (yes) 53 31 29 26 20 4 0096
Netherlands (yes) 50 47 27 47 30 19 35 13 20 12
Alcohol consumption
Australia (daily) 33 34 37 30 37 17 18 36 17 20
Finland (daily) 17 7 10 2 5 0 0031
Netherlands (daily) 29 56 41 41 31 13 18 17 15 17
Walks outdoors
Australia (no difficulty) 83 89 79 82 82 70 77 68 68 75
Finland (no difficulty) 73 57 86 74 86 67 61 61 62 89
Netherlands (yes) 55 65 90 78 98 82 70 78 75 88
Sports/physical exercise
Australia (sports regularly) 12 43 32 27 25 11 12 11 13 12
Finland (yes physical 83 57 85 74 73 79 44 54 64 75
exercise)
Netherlands (sports 28 29 51 41 59 49 41 54 68 68
last 2 weeks)
Source: Please see Table 1.
Note: The proportions are adjusted for differences in age and years of education. See text for differences in the depression measures.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Kendig et al. / Health of Parents and Childless Individuals 1479
for currently married childless women. Finnish childless men were most
likely to report the daily use of alcohol, particularly if they were never married.
In the Netherlands, formerly married men, regardless of parental status, were
more likely to report daily use of alcohol than currently married men. Among
Dutch men, a parental status difference in daily alcohol consumption was
also observed: Ever-married childless men were more likely to report the
daily use of alcohol than ever-married fathers.
Two physical activity indicators were used. The first was whether the
older adult walked outdoors (which can also be viewed as a measure of
functional capacity). The second was a direct question about engaging in
sports or physical exercise. In each country, women generally were less likely
than men to walk outdoors; however, there were variations by parental and/or
marital status. Finnish and Dutch formerly married men were less likely than
currently married men to walk outdoors, particularly if they were childless.
In Finland and the Netherlands, currently married mothers were most likely
of all groups of women to walk outdoors. In the Netherlands, a high likeli-
hood of walking outdoors was also observed for never-married women.
Questions regarding sports or physical exercise differed between the coun-
tries, which might account for the differences in overall patterns. Australian
respondents were least likely to report engaging in sports or physical exer-
cise (a restrictive measure was used). Australian men were more likely to
report being physically active than Australian women, with the exception of
never-married childless men who were particularly unlikely to engage in
sports. A gender difference was not observed in Finland or the Netherlands;
instead, parenthood rather than gender or marital status accounted for differ-
ences in physical activity. Finnish and Dutch formerly married childless men
were unlikely to engage in sports or physical exercise. Among the Dutch, this was
also the case for never-married men. Childless women in the Netherlands
were less likely to engage in physical exercise than mothers. In Finland, this
was so for ever-married childless women, not for the never married.
Summary and Conclusion
We used survey data from three countries to examine parental status dif-
ferences in late-life health in conjunction with gender and marital status.
The findings in old age show outcomes that reflect accumulating life course
experiences since childhoods dating back to the early 20th century. Thirteen
measures, which were sufficiently similar across the surveys to warrant
comparisons, were incorporated in the analyses. They pertained to general
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

health, mental health, and health behaviors. All analyses of health differ-
ences between the groups applied controls for age and education differences.
Results showed a number of consistent patterns across measures and across
countries.
First, marital status differences in health provide a context in which to
consider the importance of parenting status. The Australian and Dutch data
showed that never-married men (and to some extent formerly married men)
had relatively poor health status compared to their currently married coun-
terparts in terms of physical health (general health and own health compared
to that of peers), depression, and several of the health behaviors (daily fruit
consumption, smoking, physical exercise). Previous authors have described
never-married men as having few resources (e.g., Bernard, 1982; Kiernan,
1988; Wenger, 2001); however, the findings here emerged after controlling
for education, suggesting an independent marital status effect on men’s
health. The different pattern in the Finnish data, with never-married men
having favorable scores on several health measures, may have arisen
because residents of institutions were not included in the sample. Previous
research has demonstrated that never-married older adults are overrepre-
sented in homes for the elderly and care institutions in all three countries
(Finne-Soveri & Noro, 2000; Harmsen, Keij, & Schapendonk-Maas, 2001;
Kendig, 2000).
Second, few effects for parenthood were found independent of marital
status. Where they were observed (and this was the case in Finland and the
Netherlands), they pertained to health behaviors, and more particularly smok-
ing, alcohol consumption, and engaging in physical exercise. Our results thus
provide evidence for the social control function of parenthood: Becoming
a parent or having adult children means that people refrain from health-
compromising behavior either because they have internalized norms to do so
or because they are explicitly admonished by others to do so. It is unclear
why parenthood effects were not observed in the Australian data. Smoking
rates were low in that country, allowing for little differentiation, and gender
rather than parental and/or marital status accounted for differences in alcohol
consumption and engaging in physical exercise.
Third, the results showed that the health of formerly married men was
particularly poor if they were childless. This pattern was observed for several
of the health measures (but not always in all three countries): general health,
difficulty falling asleep, depression, walking outdoors, and engaging in phys-
ical exercise. Apparently, the loss of the spouse has a more detrimental impact
on health and health behaviors if men are childless than if they are fathers.
1480 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Kendig et al. / Health of Parents and Childless Individuals 1481
Fourth, parental and marital status differences in health, insofar as they
existed, were greater among men than among women. Earlier we noted the
relatively poor health of never-married men and that of formerly married
men. Among women, no parental and/or marital status group emerged as
having consistently poor or consistently good health across the various mea-
sures. This may suggest that the social control effects of having a spouse or
a child have a stronger effect on men than on women. Overall, gender dif-
ferences were reasonably consistent across the three countries. Women
were more likely than men to have difficulty falling asleep, to eat fruit on a
daily basis in winter, and to refrain from smoking and drinking.
Research Directions
This article has aimed to provide an integrated review and some analysis
of the effects of parenting—independent of marital status and gender—on
health throughout the life course. Our comprehensive literature review in
the first part of the article showed that previous studies shed some light on
the topic; however, the research base is limited. Available studies in the inter-
national literature seldom disentangle the effects of parenting from marital
history. Most are limited by cross-sectional research designs that cannot
separate selection and causal effects and do little to examine the mechanisms
by which parenting may influence health. Many studies are conducted only
for women, and the few that relate social and biological factors focus almost
entirely on pregnancy and childbirth. The few available longitudinal studies
also are limited to specific groups, such as women after childbearing, and
they generally are conducted over relatively short periods of time.
Our comparative analyses of health in later life have presented new evi-
dence that attempts to separate the impact of parenthood, marital status, and
gender as they have accumulated throughout the life span and influence late-
life health. Our analyses have their limitations, however. One problem is that
data collected among older adults are inevitably riddled with misleading selec-
tion effects: People with the most health difficulties are less likely to reach old
age or, if they do survive, to be able to participate in a lengthy interview. We
have no information on those who did not make it to old age. Another prob-
lem is that our sample sizes are too small to fully investigate the likely inter-
action effects among our key variables. For example, it appears that the effects
of parenting and being married are particularly influential among men; social
class may also be important (Arber, 1991; Kendig, Browning, & Teshuva,
1998). We do not know if the differences between the three countries arise
from sampling artifice or genuine social structural differences.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

1482 Journal of Family Issues
We commend the value of bio-psycho-social and life course perspec-
tives (Ryff & Singer, 2005; Sauvain-Dugerdil, Leridon, & Mascie-Taylor,
2006) for further research on the influence of marital status, childlessness,
and health as people move through adulthood into late life. A life history
framework is especially valuable for identifying the critical influences and
possibly enduring effects of family events such as marriage, divorce, and wid-
owhood, as well as childbirth, coresident parenting, the departure of children
from home, and ongoing interaction with them afterward. With the emer-
gence of chronic disease and mental health as major health concerns for the
future (World Health Organization, 2005), we need to know more about the
“trigger points” and durations of health-risking and health-promoting expo-
sures throughout the life course. Understanding family influences on health
has potential to identify opportunities and guide actions that can improve
health in later life (Kendig, Browning, & Wells, 1998).
References
Ahola, A., Djerf, K., Heiskanen, M., & Vikki, K. (1995). Elinolotutkimus 1994: Aineiston keruu
[Survey of living conditions 1994: Data collection]. Helsinki: Statistics Finland.
Akerlof, G. A. (1998). Men without children. Economic Journal, 108, 287-309.
Alwin, D. F., & Wray, L. (2005). A life-span developmental perspective on social status and
health [Special issue II]. Journal of Gerontology: Social Sciences, 60B, 7-14.
Andrews, G., Cheok, F., & Carr, S. (1989). The Australian Longitudinal Study of Aging.
Australian Journal on Aging, 8, 31-35.
Andrews, G. R., & Myers, G. C. (2000). Australian (Adelaide) Longitudinal Study of Aging,
waves 1-5 (1992-1997) [Computer file, 3rd ICPSR version].Adelaide: Flinders University
of South Australia, Centre for Ageing Studies.
Anson, O. (1989). Marital status and women’s health revisited: The importance of a proximate
adult. Journal of Marriage and Family, 51, 185-194.
Arber, S. (1991). Class, paid employment and family roles: Making sense of structural disadvan-
tage, gender and health status. Social Science & Medicine, 32, 425-436.
Backett, K. C., & Davison, C. (1995). Lifecourse and lifestyle: The social and cultural location
of health behaviours. Social Science & Medicine, 40(5), 629-638.
Belloc, N. B., & Breslow, L. (1972). Relationship of physical health status and health practices.
Preventive Medicine, 1, 409-421.
Ben-Shlomo, Y., Smith, G. D., Shipley, M., & Marmot, M. G. (1993). Magnitude and causes of
mortality differences between married and unmarried men. Journal of Epidemiology and
Community Health, 47, 200-205.
Beral, V. (1985). Long term effects of childbearing on health. Journal of Epidemiology and
Community Health, 39, 343-346.
Berkman, L., Glass, T., Brissette, I., & Seeman,T. E. (2000). From social integration to health:
Durkheim in the new millenium. Social Science & Medicine, 51, 843-857.
Berkman, L., & Syme, S. (1979). Social networks, host resistance, and mortality: A nine-year
follow-up of Alameda County residents. American Journal of Epidemiology, 109, 186-204.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Kendig et al. / Health of Parents and Childless Individuals 1483
Berkman, L. F., Breslow, L., & Wingard, D. (1983). Health practices and mortality risk. In
L. F. Berkman & L. Breslow (Eds.), Health and ways of living: The Alameda County Study
(pp. 61-112). Oxford, UK: Oxford University Press.
Bernard, J. (1982). The future of marriage (2nd ed.). New Haven, CT: Yale University Press.
Brockmann, H., & Klein, T. (2004). Love and death in Germany: The marital biography and its
effect on mortality. Journal of Marriage and Family, 66, 567-581.
Camacho, T. C., & Wiley, J. (1983). Health practices, social networks, and change in physical
health. In L. F. Berkman & L. Breslow (Eds.), Health and ways of living: The Alameda
County Study (pp. 176-209). Oxford, UK: Oxford University Press.
Cheung, Y. B. (2000). Marital status and mortality in British women: A longitudinal study.
International Journal of Epidemiology, 29, 93-99.
Choi, N. G. (1994). Patterns and determinants of service utilization: Comparisons of the child-
less elderly and elderly parents living with or apart from their children. The Gerontologist,
34, 353-362.
Connidis, I. A., & McMullin, J. A. (1993). To have or have not: Parent status and the subjective
well-being of older men and women. The Gerontologist, 33, 630-636.
Crystal, S., & Shea, D. (Eds.). (2003). Economic outcomes in later life: Public policy, health,
and cumulative advantage. New York: Springer.
Dannefer, D. (2003). Cumulative advantage/disadvantage and the life course: Cross-fertilizing
age and social science theory. Journal of Gerontology: Social Sciences, 58, S327-S337.
Deeg, D. J. H., & Westendorp-De Serière, M. (Eds). (1994). Autonomy and well-being in the
aging population I: Report from the Longitudinal Aging Study Amsterdam 1992-1993.
Amsterdam: VU Uitgeverij.
Doblhammer, G. (2000). Reproductive history and mortality later in life:A comparative study
of England and Wales and Austria. Population Studies, 54, 169-176.
Doblhammer, G., & Oeppen, J. (2003). Reproduction and longevity among the British peerage:
The effect of frailty and health selection. Proceedings of the Royal Society of London, Series
B, Biological Sciences, 270, 1541-1547.
Dribe, M. (2004). Long-term effects of childbearing on mortality: Evidence from pre-industrial
Sweden. Population Studies, 58, 297-310.
Durkheim, E. (1951). Suicide: A study in sociology. Glencoe, IL: Free Press. (Original work
published 1896)
Eggebeen, D. J., & Knoester, C. (2001). Does fatherhood matter for men? Journal of Marriage
and Family, 63, 381-393.
Elstad, J. I. (1996). Inequalities in health related to women’s marital, parental, and employment
status: A comparison between the early 1970s and the late 80s, Norway. Social Science &
Medicine, 42, 75-89.
Finne-Soveri, H., & Noro, A. (2000). RAI uusi apuvaline vanhuspalveluiden suunnitteluun
[A new tool for planning of the elderly care]. Sairaala, 4(62), 12-14.
Fokkema, T. (2002). Combining a job and children: Contrasting the health of married and
divorced women in the Netherlands? Social Science & Medicine, 54, 741-752.
Friedlander, N. J. (1996). The relation of lifetime reproduction to survivorship in women and
men: A prospective study. American Journal of Human Biology, 8, 771-783.
Gavrilova, L. A., & Gavrilova, N. S. (1999). Is there a reproductive cost of human longevity?
Journal of Anti-Aging Medicine, 2, 121-123.
Gove, W. R. (1973). Sex, marital status, and mortality. American Journal of Sociology, 79, 45-67.
Gove, W. R., & Hughes, M. (1979). Possible causes for the apparent sex differences in physical
health: An empirical investigation. American Sociological Review, 44, 126-146.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

1484 Journal of Family Issues
Green, A., Beral, V., & Moser, K. (1988). Mortality in women in relation to their childbearing
history. British Medical Journal, 297, 391-394.
Grundy, E., & Tomassini, C. (2005). Fertility history and health in later life: A record linkage
study in England and Wales. Social Science & Medicine, 61, 217-228.
Hagestad, G. O., & Call, V. R. A. (2007). Pathways to childlessness: A life course perspective.
Journal of Family Issues, 28, 1338-1361.
Harmsen, C., Keij, I., & Schapendonk-Maas, H. (2001). Zelfstandig oud [Independently old].
Maandstatistiek van de Bevolking, 49, 4-7.
Hibbard, J. H., & Pope, C. R. (1991). Effect of domestic and occupational roles on morbidity
and mortality. Social Science & Medicine, 32, 805-811.
Høyer, G., & Lund, E. (1993). Suicide among women related to number of children in marriage.
Archives of General Psychiatry, 50, 134-137.
Hu, Y., & Goldman, N. (1990). Mortality differentials by marital status: An international com-
parison. Demography, 27, 233-250.
Huijbregts, P., Feskens, E., Räsänen, L., Fidanza, F., Nissinen, A., Menotti, A., et al. (1997).
Dietary patterns and 20 year mortality in elderly men in Finland, Italy, and the Netherlands:
Longitudinal cohort study. British Medical Journal, 315, 13-17.
Hurt, L. S., Ronsmans, C., & Thomas, S. L. (2006). The effect of number of births on women’s
mortality: Systematic review of the evidence for women who have completed their child-
bearing. Population Studies, 60, 55-71.
Joung, I. (1996). Marital status and health: Descriptive and explanatory studies. Unpublished
doctoral dissertation, Erasmus University, Rotterdam, Netherlands.
Joung, I. M. A., Stronks, K., van de Mheen, H., & Mackenbach, J. P. (1995). Health behaviours
explain part of the differences in self-reported health associated with partner/marital status
in The Netherlands. Journal of Epidemiology and Community Health, 49, 482-488.
Kalmijn, M. (1998). Intermarriage and homogamy: Causes, patterns, trends. Annual Review of
Sociology, 24, 395-421.
Kendig, H. (1996). Understanding health promotion for older people: Sociological contribu-
tions. In V. Minichiello, N. Chappell, H. Kendig, & A. Walker (Eds.), Sociology of aging:
International perspectives (pp. 360-375). Melbourne, Australia: International Sociological
Association, Research Committee on Aging.
Kendig, H. (2000). Family change and family bonding in Australia. In W. Liu & H. Kendig
(Eds.), Who should care for the elderly? An East-West divide (pp. 107-125). Singapore:
Singapore University Press.
Kendig, H., Browning, C., & Teshuva, K. (1998). Health actions and social class among older
Australians. Australian and New Zealand Journal of Public Health, 22, 808-813.
Kendig, H., Browning, C., & Wells, Y. (1998). Australian research on ageing, families, and health
promotion. Social Policy Journal of New Zealand, 11, 169-181.
Kendig, H., Helme, R., Teshuva, K., Osborne, D., Flicker, L., & Browning, C. (1996). Health
Status of Older People project: Preliminary findings from a survey of the health and lifestyles
of older Australians. Melbourne, Australia: Victorian Health Promotion Foundation / La
Trobe University, Lincoln Gerontology Centre / University of Melbourne, National Ageing
Research Institute.
Kiernan, K. E. (1988). Who remains celibate? Journal of Biosocial Science, 20, 253-263.
Kirkwood, T. B. L. (1977). Evolution of aging. Nature, 270, 301-304.
Kirkwood, T. B. L., & Holliday, R. (1979). The evolution of ageing and longevity. Proceedings
of the Royal Society of London, Series B, Biological Sciences, 205, 531-546.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Kendig et al. / Health of Parents and Childless Individuals 1485
Kitagawa, E. M., & Hauser, P. M. (1973). Differential mortality in the United States: A study in
socioeconomic epidemiology. Cambridge, MA: Harvard University Press.
Kobrin, F. E., & Hendershot, G. E. (1977). Do family ties reduce mortality? Evidence from the
United States, 1966-1968. Journal of Marriage and Family, 39, 737-745.
Koropeckyj-Cox, T., & Call, V. R. A. (2007). Characteristics of older childless persons and
parents: Cross-national comparisons. Journal of Family Issues, 28, 1362-1414.
Kreager, P. (2004). Where are the children? In P. Kreager & E. Schröder-Butterfill (Eds.), Ageing
without children: European and Asian perspectives (pp. 1-45). Oxford, UK, and New York:
Berghahn Books.
Kvåle, G., Heuch, I., & Nilssen, S. (1994). Parity in relation to mortality and cancer inci-
dence: A prospective study of Norwegian women. International Journal of Epidemiology,
23, 691-699.
Le Bourg, E. (2001). A mini-review of the evolutionary theories of aging. Is it the time to accept
them? Demographic Research, 4, 1-28.
Lillard, L. A., & Waite, L. J. (1995). Till death do us part: Marital disruption and mortality.
American Journal of Sociology, 100, 1131-1156.
Lund, E., Arnesen, E., & Borgan, J.-K. (1990). Pattern of childbearing and mortality in married
women—A national prospective study from Norway. Journal of Epidemiology and Community
Health, 44, 237-240.
Macintyre, S. (1992). The effects of family position and status on health. Social Science &
Medicine, 35, 453-464.
Marks, S. R. (1977). Multiple roles and role strain: Some notes on human energy, time and
commitment. American Sociological Review, 42, 921-936.
Martelin, T., Koskinen, S., & Valkonen, T. (1998). Sociodemographic mortality differences
among the oldest old in Finland. Journal of Gerontology: Social Sciences, 53B, S83-S90.
Martikainen, P. (1995). Women’s employment, marriage, motherhood and mortality: A test of the
multiple role and role accumulation hypotheses. Social Science & Medicine, 40, 199-212.
Mastekaasa, A. (1992). Marriage and psychological well-being: Some evidence on selection
into marriage. Journal of Marriage and Family, 54, 901-911.
Mastekaasa, A. (2000). Parenthood, gender and sickness absence. Social Science & Medicine,
50, 1827-1842.
McMunn, A., Bartley, M., Hardy, R., & Kuh, D. (2006). Life course social roles and women’s
health in mid-life: Causation or selection? Journal of Epidemiology and Community Health,
60, 484-489.
O’Rand, A. (1996). The precious and the precocious: Understanding cumulative advantage
and cumulative disadvantage over the life course. The Gerontologist, 36, 230-238.
Phillipson, C., Bernard, M., Phillips, J., & Ogg, J. (2001). The family and community life of
older people: Social networks and social support in three urban areas. London: Routledge.
Pless, I. B., Cripss, H. A., Davies, J. M. C., & Wadsworth, M. E. J. (1989). Chronic physical illness
in childhood: Psychological and social effects in adolescence and adult life. Developmental
Medicine and Child Neurology, 31, 746-755.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement, 1, 385-401.
Ryff, C. D., & Singer, B. H. (2005). Social environments and the genetics of aging: Advancing
knowledge of protective health mechanisms. The Journals of Gerontology Series B: Psycho-
logical Sciences and Social Sciences, 60, 12-23.
Sauvain-Dugerdil, C., Leridon, H., & Mascie-Taylor, N. (Eds.). (2006). Human clocks: The bio-
cultural meanings of age (Population, Family, and Society Series, Vol. 5). Berlin, Germany:
Peter Lang.
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from

Shye, D., Mullooly, J. P., Freeborn, D. K., & Pope, C. R. (1995). Gender differences in the rela-
tionship between social network support and mortality: A longitudinal study of an elderly
cohort. Social Science & Medicine, 41, 935-947.
Sieber, S. D. (1974). Toward a theory of role accumulation. American Sociological Review, 39,
567-578.
Simon, R. W. (2002). Revisiting the relationships among gender, marital status, and mental
health. American Journal of Sociology, 107, 1065-1096.
Sweeney, M. M. (2002). Two decades of family change: The shifting economic foundations of
marriage. American Sociological Review, 67, 132-147.
Uchino, B. N. (2004). Social support and physical health: Understanding the health conse-
quences of relationships. New Haven, CT: Yale University Press.
Umberson, D. (1987). Family status and health behaviors: Social control as a dimension of
social integration. Journal of Health and Social Behavior, 28, 306-319.
Verbrugge, L. M. (1979). Marital status and health. Journal of Marriage and Family, 41, 267-285.
Verbrugge, L. M. (1983). Multiple roles and physical health of women and men. Journal of
Health and Social Behavior, 24, 16-30.
Waite, L. J., & Gallagher, M. (2000). The case for marriage: Why married people are happier,
healthier, and better off financially. New York: Doubleday.
Waldron, I., Weiss, C. C., & Hughes, M. E. (1998). Interacting effects of multiple roles on
women’s health. Journal of Health and Social Behavior, 39, 216-236.
Weatherall, R., Joshi, H., & Macran, S. (1994). Double burden or double blessing? Employment,
motherhood and mortality in the Longitudinal Study of England and Wales. Social Science
& Medicine, 38, 285-297.
Weitoft, G. R., Burström, B., & Rosén, M. (2004). Premature mortality among lone fathers and
childless men. Social Science & Medicine, 59, 1449-1459.
Wenger, G. C. (1984). The supportive network: Coping with old age. London: Allen & Unwin.
Wenger, G. C. (2001). Ageing without children: Rural Wales. Journal of Cross-Cultural
Gerontology, 16, 79-109.
Wenger, G. C., Dykstra, P. A., Melkas, T., & Knipscheer, K. C. P. M. (2007). Social embedded-
ness and late life parenthood: Community activity, close ties, and support networks. Journal
of Family Issue, 28, 1419-1456.
Wenger, G. C., Scott, A., & Patterson, N. (2000). How important is parenthood? Childlessness
and support in old age in England. Ageing & Society, 20, 161-182.
Westendorp, R. G. J., & Kirkwood, T. B. L. (1998). Human longevity at the cost of reproductive
success. Nature, 396, 743-746.
World Health Organization. (2005). Preventing chronic diseases:A vital investment (WHO Global
Report).Geneva, Switzerland: WHO Press.
Wyke, S., & Ford, G. (1992). Competing explanations for associations between marital status and
health. Social Science & Medicine, 34, 523-532.
1486 Journal of Family Issues
at University of Groningen on November 6, 2009 http://jfi.sagepub.comDownloaded from