The Long Arm of Warm Parenting

Abstract

Does adulthood physical health continue to reflect parental support and warmth received during childhood? Although previous research supports this continuity, I examine this question according to a sex-matching perspective. Drawing on representative cross-sectional data featuring detailed measures of maternal and paternal warmth (1995 National Survey of Midlife Development in the United States), I focus on adults who came of age in traditional two-parent households containing a biological mother and father. Across three physical health outcomes, I find strong support for sex-matching, in that paternal warmth during childhood is usually more important to male adulthood health whereas maternal warmth tends to be more important to female adulthood health. Sex-matching effects are especially evident for physical health difficulties (number of chronic health conditions and functional limitations). These findings provide a new lens through which to view physical health disparities among today’s middle- to late-aged adults who came of age in traditional households.

Keywords

physical health parent/child relations mother–child relationship father–child relationship life course

Parental warmth or emotional support experienced during childhood provides “the underpinnings of adult health”: It does so by shaping a child’s lifelong “ability to develop psychosocial resources” (such as social support, problem-focused coping skills, and self-esteem) and by durably setting the flexibility and regulation of biological and physiological response systems (Shaw, Krause, Chatters, Connell, & Ingersoll-Dayton, 2004; Taylor, Baldwin, & Seeman, 2011, pp. 939-941). In essence, children who sustain low levels of parental support during the early years of life have been found to exhibit exaggerated stress response, diminished psychosocial resources, and a variety of metabolic, biological, and inflammatory symptoms that together culminate in an “integrated profile of risk” that systematically hampers physical health across the life course (Repetti, Taylor, & Seeman, 2002, p. 336; see also Braveman & Barclay, 2009; Case, Fertig, & Paxton, 2005; Caspi & Silva, 1995; Conger & Donnellan, 2007; Elder, 1998).

In this research, I focus on traditional two-parent households containing a biological mother and father. For this household type, I ask whether warmth from a same-sex parent experienced during childhood matters more to later adulthood physical health than warmth from an opposite-sex parent. Empirical studies have yielded mixed support for the special importance of father–son and mother–daughter emotional bonds during childhood to adult children’s mental health and to children’s mortality outcomes of parental absence or death in particular (e.g., Mallers, Charles, Neupert, & Almeida, 2010; Marks, Jun, & Song, 2007; Poon & Knight, 2012; Powell & Downey, 1997; Rostila & Saarela, 2011; Stewart-Brown, Fletcher, & Wadsworth, 2004). However, general physical health outcomes of early parent–child emotional bonds remain a neglected topic, especially the importance of children’s gender to shaping these effects.

Background

Do the health effects of supportive parenting remain as children age into later adulthood? Emergent adulthood is marked by the formation of new ties and intimate relationships, and the assumption of work and family roles, making it unclear whether parental relationships would continue to matter to adult children’s well-being (Amato, 1994; Elder, 1998). Indeed, newly forged relationships and work and family roles all have well-being implications. Thus, adulthood may, figuratively speaking, “rewrite” earlier experiences with parents (Amato, 1994; Hughes & Waite, 2009; Sampson & Laub, 1993; Wheaton & Reid, 2008).

In contrast, developmental and life-course theory posit a continuity model. According to a continuity model, emotional bonds with caregivers are profoundly formative, in that they provide deeply internalized models for how to form social bonds later on in life (Bowlby, 1979; Caspi, Bem, & Elder, 1989; Mikulincer & Shaver, 2003). Thus, poor relationships with parents during childhood may well continue to manifest themselves through stifled intimate relationships, social anxiety, or impaired social interaction later on in life. Rather than rewriting parental experiences, then, key transitions in socializing, love, work, or family may therefore play out in ways that accentuate or reproduce earlier emotional difficulties with parents. This should lead, in turn, to the perpetuation or enhancement of earlier health differences.

Consistent with a continuity perspective, studies have uncovered links between abuse, neglect, rejection, or parental absence during childhood and later physical health deficits during adulthood (Blackwell, Hayward, & Crimmins, 2001; Danese, Parlante, Caspi, Taylor, & Poulton, 2007; Haas, 2008; Hamil-Luker & O’Rand, 2007; Hayward & Gorman, 2004; Kestilä et al., 2006). Likewise, several studies have found that emotional quality of parent–child bonds, reported by the adult child retrospectively, indeed has health correlates that are observed at various stages of adulthood. Unfortunately, however, these studies draw on selective, small samples (Russek & Schwartz, 1997); focus on either adult men or women (Min, Minnes, Kim, & Singer, 2013; Russek & Schwartz, 1997; Walker et al., 1999); measure parent–child relationships using a single, abstract item rather than a targeted scale (Stewart-Brown et al., 2004); and/or conflate maternal and paternal warmth (Ryff, Singer, & Palmersheim, 2004; Shaw et al., 2004).

In part as a result of these shortcomings, prior work has yet to investigate whether sex-matching between parents and children structures health effects. The value of sex symmetry or matching to parent–child relationships is put forth by cultural narratives as well as social role-modeling theories. These paradigms see daughters as needing to benefit from the wisdom that supposedly only mothers can provide and sons needing the guidance of a father in order to adapt to social life (see Buchmann & DiPrete, 2006; Mokrue, Chen, & Elias, 2011; Powell & Downey, 1997; Videon, 2005, for specific arguments).

Given the fundamental importance of gender to social interaction (Ridgeway & Correll, 2004), children may struggle with many aspects of social life throughout adulthood without proper gender socialization from parents. For instance, it may be the case that problem-solving and coping skills are most relevant when provided by the same-sex parent. Indeed, most social ties and friendships across the life course are same-sex (Fischer, 2011; McPherson, Smith-Lovin, & Cook, 2001); in this regard, same-sex parent–child socialization may provide a particularly strong basis for developing psychosocial resources across the life course.

In terms of parental investment, there is some evidence that fathers favor sons and that mothers favor daughters. For instance, fathers spend more time with sons than with daughters and are more likely to invest in sibships that contain sons (Lundberg, 2005; Mammen, 2011; Raley & Bianchi, 2005). Reciprocally, children seem to favor spending time with same-sex parents when given a choice (Maccoby, 2003). In part, emotional associations between parents and children based on gender may be driven by cultural expectations that fathers should spend time with sons and mothers with daughters (Boyd, 1989; Lundberg, McLanahan, & Rose 2007), by intuitive beliefs about the necessity of same-sex socialization for favorable social development (Chodorow, 1978), and by similarity of interests between parents and their children due to common gender (Suitor & Pillemer, 2006).

Some research focusing on two-parent households indeed does find that father–son and mother–daughter relationships are especially important to adult children’s health. However, the importance of early parent–child relationships to adult children’s health depends on the outcome assessed and sometimes, though not always, on the child’s gender (Mallers et al., 2010; Marks et al., 2007; Poon & Knight, 2012; Umberson & Chen, 1994). For instance, within the realm of psychological health, reactivity to daily stressors during adulthood is linked to early father–son relationships but not to early father–daughter relationships; in contrast, nonspecific psychological distress is not patterned by the child’s gender (Mallers et al., 2010) nor is substance abuse (Rothrauff, Cooney, & An, 2009). Similarly, paternal death is more strongly linked to depressive symptoms and psychological distress for adult sons than daughters, whereas effects on self-esteem and personal mastery do not differ by the child’s gender (Barnett, Marshall, & Pleck, 1992; Marks et al., 2007).

In terms of physical health, however, men seem to be more adversely affected by a mother’s than a father’s death (Marks et al., 2007; Rostila & Saarela, 2011). However, this impact depends on whether the parent had a preexisting problem and also the nature of that problem (Umberson & Chen, 1994). Likewise, for certain measures of parental presence, daughters actually appear to benefit from father–child relationships more so than do sons. Stewart-Brown et al. (2004) found that daughters had greater susceptibility to chronic health conditions from having “uncertain” or ambivalent relationships with their fathers during childhood than did sons. However, the conceptual link between this item and detailed measures of parental warmth is unclear. Although they did not examine parent–child relationship quality, Hamil-Luker and O’Rand (2007) found that the absence of a father during childhood contributed more to later heart attack risk for daughters than for sons.

Parental warmth is distinguishable from parental death, absence, or ambivalence in important ways. Warmth represents a positive and sustained influence on children’s development, whereas parental death or absence represents an acute or chronic stressor, as does ambivalence. Warm or supportive parenting, as reported from a child’s perspective, represents a high-quality emotional bond with a parent, above and beyond that parent’s mere presence (as captured by measures of residence or custody; e.g., Powell & Downey, 1997) and the absence of maltreatment (Mallers et al., 2010; Russek & Schwartz, 1997; Ryff et al., 2004; Shaw et al., 2004). In this way, parental warmth may supersede or even dictate the meaning of a parental death or absence and thus have distinct effects on adult children’s health.

In this research, I focus on children who grew up in traditional two-parent households consisting of a mother and a father, and I examine how parental warmth experienced during a child’s early and adolescent years is linked to later life-course health differences. Of key interest, I examine differences by children’s sex. In the tracking of health outcomes, mortality and self-rated health show different patterns by sex than do functional limitations and disease-specific pathways, for example, both in general and in relation to childhood parental variables (Gorman & Read, 2006; Turrell, Lynch, Leite, Raghunathan, & Kaplan, 2007). Thus, sex-matching effects may be outcome-specific in some cases. Therefore, I examine three general physical health outcomes: self-rated health, number of chronic conditions, and functional limitations.

Method

Nationally representative data on adulthood health and parental warmth during one’s childhood are made available by the 1995 to 1996 National Survey of Midlife Development in the United States (MIDUS) administered by the John D. and Catherine T. MacArthur Foundation (available on the ICPSR website: http://www.icpsr.umich.edu). The main component of MIDUS is a probability sample consisting of English-speaking, noninstitutionalized adults residing in the contiguous United States (referred to herein as the random-digit-dial or RDD sample). Adults were aged 25 to 74 years at the time of the survey. About 70% of initially chosen RDD respondents responded to the first phase of the survey: a phone interview that gathered basic information. 87% of this phone-response RDD sample then also agreed to complete questionnaires that included items about parental background as well as detailed health measures. For the 1995 wave, complete data were available for n = 3,034 respondents. After excluding respondents not raised in households with both biological parents present (n = 724, about 24% of sample), those not reporting warmth for a female or male caregiver (n = 38), and after taking into account voluntary missingness for dependent and predictor variables (<5% of respondents for any given variable), estimation samples were approximately n = 1,930 (n = 942-995 for each respondent sex).

Inclusion in the analysis was positively linked to respondent education and age and was more common for males and less common for racial or ethnic minorities. Furthermore, included respondents reported higher levels of health and parental warmth than nonincluded respondents. All of these differences are consistent with previous research finding that two-parent households are less common among recent cohorts and among minorities and that children from intact, two-parent households tend to fare better in terms of educational attainment and mental and physical health than children from households marked by divorce, stepparenting, or single parenthood (Amato, 2005; Musick & Meier, 2010). Auxiliary analyses that used multiple imputations or a restricted set of respondent covariates led to identical findings, as did analyses that included nonresidential or nonbiological parents.

Measures

To offer a fairly comprehensive assessment of adulthood health (e.g., Jylhä, 2009), I employ three outcomes: self-rated physical health, number of chronic conditions, and functional limitations. These outcomes are desirable in that they cover many domains of physical well-being. In addition, chronic conditions and functional limitations are conceptually similar, in that an increasing number of chronic conditions is linked to increasing functional limitations and vice-versa. For this reason, these two outcomes both represent physical health difficulties (Haas, 2008; Hayward, Miles, Crimmins, & Yang, 2000; Yang, 2006). Having two distinct measures of health difficulties is useful, allowing me to check for consistent effects of parental warmth within the same general domain.

Self-Rated Physical Health

The MIDUS self-rated health item asked: “In general, would you say your physical health is excellent, very good, good, fair or poor?” (1 = poor, 3 = good, 5 = excellent). A focus on physical health is desirable to keep the respondent from reporting mental health issues. Although it is a subjective, single-item measure, self-rated health is linked to the incidence and severity of diverse health problems, and it predicts mortality net of objective physician assessments (Idler & Benyamini, 1997; Jylhä, Volpato, & Guralnik, 2006).

Number of Chronic Conditions

I invoke a count measure of health constructed from items in the questionnaire. Across these items, the respondent was asked: “In the past 12 months, have you experienced or been treated for any of the following”: respiratory conditions (e.g., asthma, bronchitis, emphysema, tuberculosis), bone or joint diseases, thyroid disease, high blood pressure, alcohol or drug problems, neurological problems, diabetes, autoimmune disorders and stroke, as well as a variety of less serious health conditions. Altogether, this inventory makes for a rich measure of health as it combines the invaluable subjective nature of health assessment (Jylhä et al., 2006) with an objective checklist of pathology (Hayward et al., 2000; Herd, 2010). In all, 0.9% of respondents reported 12 chronic conditions or greater (12-27 conditions). I recoded the ceiling of this index to 12. Various recodes of this variable that move the ceiling value all produce substantively identical results but lead to slightly varying estimates.

Functional Limitations

Functional limitations were assessed by asking the respondent to what extent he or she had difficulty with various tasks such as lifting groceries, climbing flights of stairs, kneeling, walking, and moderate to vigorous activities (1 = a lot to 4 = not at all, for each type of activity; 7 items). Limitations were then computed as the mean of reverse-scored items for respondents giving at least one valid item response (α = .93).

Parental Warmth

Parental warmth was assessed using separate scales for maternal and paternal warmth. Respondents were asked to complete the maternal warmth scale in reference to their primary female caregiver and the paternal warmth scale in reference to the primary male caregiver, thus allowing for the measurement of warmth for nonbiological (e.g., adoptive or step) parents. Scales contained the same seven items, changing only the referent title and/or gender for the caregiver. Items included a general question about the respondent’s relationship with his/her caregiver growing up: “How would you rate your relationship with your mother/father during the years you were growing up?” (1 = excellent, 2 = very good, 3 = good, 4 = fair, 5 = poor) as well as six specific questions about the relationship: “How much did she/he understand your problems and worries?” “How much love and affection did she/he give you?” “How much time and attention did she/he give you when you needed it?” “How much could you confide in him/her about things that were bothering you?” “How much effort did he/she put into watching over you and making sure you had a good upbringing?” and “How much did he/she teach you about life?” The latter six items had a four-category response format (1 = a lot, 2 = some, 3 = a little, 4 = not at all). After rescaling the general question, all seven items were reverse-coded and then averaged to create a maternal or paternal warmth score (maternal warmth α = .91; paternal warmth α = .93). Maternal and paternal warmth correlated moderately (r = .5).

Parental Socioeconomic Status

Financial hardships affecting the household induce various types of role strain that may ultimately lead to deficits in parental warmth (Conger & Donnellan, 2007; Elder, 1998). To separate the effects of parental warmth from the socioeconomic conditions growing up, therefore, I made use of three separate socioeconomic indicators. First, I constructed an indicator for middle-class parental social background based on whether one or both of the respondent’s parents were employed in a managerial or professional occupation (1 = at least one managerial/professional parent, 0 = neither parent held managerial/professional occupation). For an additional covariate, I averaged parents’ levels of education when both were reported and used either value when one was missing (measured as 0-20 years). Finally, I also took into consideration the respondent’s subjective assessment of his/her financial situation growing up (“When you were growing up, was your family better off or worse off financially than the average family was at that time?” 1 = a lot better off, 4 = same as average family, 7 = a lot worse off; reverse-coded so that higher values reflect a better subjective financial situation). Parental SES variables are intercorrelated but only moderately (rs = 0.3-0.5). Their low VIFs across all models (<2) attest to their unique information in predicting health outcomes.

Parental Health

Parental illnesses, whether physical or mental, may greatly constrain the ability to provide warmth to children. Therefore, to ensure that warmth effects operated independently of parental health challenges, I made use of the respondent’s assessments of maternal and paternal health. Two items asked: “Looking back to when you were 16, how would you rate your biological mother’s (or father’s) health at that time?” (1 = excellent, 3 = good, 5 = poor). While these questions refer to biological parents in particular, no other questions queried the health statuses of nonbiological caregivers. Results do not differ when these items are excluded from models.

Number of Siblings

Number of siblings in the household shapes parental stresses in terms of time, energy, and financial resources and thus may well affect parental warmth received by any given child. In a series of six questions, the respondent indicated how many brothers or sisters she/he had growing up (older, same age, younger). I summed across these six questions to generate a sibling count. Counts greater than eight were recoded to eight.

Sociodemographic Factors

I control for a variety of respondent sociodemographic factors that are linked to health outcomes and also are likely to reflect childhood origins. Education was surveyed by MIDUS in terms of credential points (i.e., completion of junior high, high school, GED, associate’s degree, bachelor’s, master’s, or doctorate), with degree midpoints also present (e.g., 1 or 2 years of college, no degree yet). I recoded this variable so that education is measured as 0 to 20 years of formal instruction. Gender is measured as an indicator for male (0 = female, 1 = male). Age is measured in years. Race is measured as an indicator for whether the respondent reported being White, given other categories of Black and/or African American, Native American or Aleutian Islander/Eskimo, Asian or Pacific Islander, Other, or multiracial (0 = non-White, 1 = White). Marital status is assessed by two interrelated indicators capturing whether the respondent was currently married at the time of the baseline interview (0 = no, 1 = yes) or, if not, whether the respondent was currently living with a steady partner (0 = no, 1 = yes).

Labor force status is assessed by whether the respondent was working full- or part-time at baseline (0 = no, 1 = yes; vs. all other labor force statuses). Furthermore, I account for whether the respondent has children, as child-raising has been shown to diminish well-being and to conflict with relationship and work roles, thus potentially altering health outcomes (Umberson, Pudrovska, & Reczek, 2010). In keeping with prior work, I distinguish between having any children less than 7 (0 = no, 1 = yes) and having any children aged 7 to 17 years (0 = no, 1 = yes). Finally, household income for the past year ($0-$300,000+; log-transformed) was included in all models to distinguish health effects of education from those of income.

Modeling Strategy

For ease of presentation, I fitted separate models of adulthood physical health by respondent sex. This is the most straightforward way of showing whether and to what extent maternal or paternal warmth matters for a given respondent sex. All models included all parental variables and covariates as listed above. To determine whether the importance of maternal or paternal warmth differed significantly by respondent sex, I estimated pooled regression models that included all respondents, as well as interactions of maternal and paternal warmth with respondent sex. The significance of the interaction terms from these models are reported in a separate (Diff) column.

Self-rated physical health was recoded into a binary variable (good, very good, or excellent health vs. poor or fair health) and then fitted using logistic regression. Relative to an ordinal logistic strategy, which retains all five outcome categories, a logistic strategy produced better model fit (determined via BIC) and more interpretable results; furthermore, substantive results were identical across both approaches. Number of chronic conditions was fitted using negative binomial regression, and functional limitations were fitted using ordinary least squares. Consistent with overdispersion, likelihood-ratio testing strongly supported the usage of negative binomial over Poisson regression for number of chronic conditions (ps < .0001). I log-transformed functional limitations to improve error-term normality as well as overall model fit.

Results

Descriptive statistics are presented in Table 1 by respondent sex. Respondents averaged good self-rated health (about 3.5 on a 5-point scale), carried 2 to 3 chronic conditions, and had only mild functional limitations (about 1.5 on a 4-point scale). Consistent with previous research, female respondents reported lower levels of physical health (higher rates of chronic conditions and higher functional limitations; Gorman & Read, 2006).

Table 1.

Descriptive Statistics (1995 MIDUS).

Variable	Male	Female	Variable	Male	Female
Self-Rated Health (1-5)	3.56	3.48	Education (0-20)	14.42	13.94
	(0.94)	(0.98)		(2.69)	(2.35)
Good, Very Good, or Excellent Self-Rated Health (≥Good)	0.88	0.85	Age (25-74)	46	46.72
Number of Conditions (0-12)	1.97	2.76		(12.47)	(12.55)
	(2.19)	(2.6)	White	0.9	0.91
Functional Limitations (1-4)	1.43	1.66		(0.3)	(0.29)
	(0.68)	(0.82)	Married	0.73	0.61
Maternal Warmth (1-4)	3.27	3.07		(0.45)	(0.49)
	(0.58)	(0.71)	With Partner	0.05	0.04
Paternal Warmth (1-4)	2.82	2.77		(0.21)	(0.19)
	(0.72)	(0.8)	Child < 7	0.18	0.15
Parental Education (0-20)	12.34	12.24		(0.38)	(0.36)
	(3.57)	(3.54)	Child 7-17	0.34	0.29
Finances Growing Up (1-7)	4.21	4.14		(0.47)	(0.45)
	(1.22)	(1.22)	Full-Time	0.76	0.55
Prof./Managerial Parent(s)	0.34	0.31		(0.43)	(0.5)
	(0.47)	(0.46)	Part-Time	0.04	0.12
Maternal Health (1-5)	3.62	3.6		(0.19)	(0.32)
	(1.1)	(1.07)	Household Income^a	79.19	64.05
Paternal Health (1-5)	3.65	3.55		(63.01)	(56.29)
	(1.06)	(1.02)	Num. of Siblings (0-8)	2.91	3.14
				(2.08)	(2.09)
			Observations	977	930

Note. Means appear above standard deviations. Ranges are in parentheses; all other variables are binary.

Household income is shown in thousands of dollars.

Self-Rated Physical Health

Maternal warmth during childhood was linked to improved probability of good or better self-rated health for female respondents (Table 2; b = .52, p < .01) but showed no relationship to health for male respondents (b = −.14, ns). Paternal warmth did not predict self-rated health for either sex (|b|s < .20, ps > .10). Maternal warmth was significantly more important to health among females according to a pooled interaction model (“Diff” column, p < .05).

Table 2.

Regressions of Adulthood Physical Health on Parental Warmth During Childhood (1995 MIDUS).

							Functional limitations
	Self-rated health (≥ good; logistic)			Chronic conditions (neg. binomial)			(OLS regression)
Independent variables	Males	Females	(Diff)	Males	Females	(Diff)	Males	Females	(Diff)
Maternal Warmth	−0.142	0.518**	*	−0.049	−0.208**	*	0.030	−0.065**	**
	(0.215)	(0.146)		(0.069)	(0.046)		(0.023)	(0.020)
Paternal Warmth	−0.019	−0.192	ns	−0.156**	0.031	*	−0.052**	0.027	**
	(0.169)	(0.141)		(0.055)	(0.043)		(0.018)	(0.019)
Education	0.170**	0.141**	—	−0.051**	−0.026^	—	−0.016**	−0.021**	—
	(0.043)	(0.051)		(0.013)	(0.015)		(0.004)	(0.006)
Age	−0.011	−0.015	—	0.013**	0.012**	—	0.004**	0.010**	—
	(0.012)	(0.011)		(0.004)	(0.003)		(0.001)	(0.001)
White	−0.921*	0.167	—	−0.043	−0.126	—	−0.023	−0.066	—
	(0.467)	(0.353)		(0.117)	(0.112)		(0.038)	(0.047)
Married	0.095	−0.051	—	−0.075	−0.092	—	−0.028	−0.027	—
	(0.263)	(0.226)		(0.087)	(0.068)		(0.028)	(0.029)
With Partner	0.780	−0.759	—	0.209	0.143	—	–0.106^	0.073	—
	(0.659)	(0.472)		(0.166)	(0.159)		(0.056)	(0.069)
Child <7	0.400	0.241	—	0.031	−0.087	—	−0.013	−0.005	—
	(0.386)	(0.359)		(0.102)	(0.104)		(0.032)	(0.042)
Child 7-17	0.091	−0.188	—	−0.005	−0.092	—	0.016	−0.033	—
	(0.263)	(0.259)		(0.077)	(0.076)		(0.025)	(0.031)
Full-Time	0.983**	0.639**	—	−0.180^	−0.113	—	−0.170**	−0.107**	—
	(0.273)	(0.230)		(0.095)	(0.074)		(0.032)	(0.031)
Part-Time	−0.000	0.988*	—	0.128	−0.075	—	−0.107^	−0.136**	—
	(0.491)	(0.389)		(0.177)	(0.105)		(0.061)	(0.044)
Household Income	0.119*	0.043	—	−0.000	0.014	—	−0.014^	−0.012	—
	(0.058)	(0.046)		(0.023)	(0.018)		(0.007)	(0.008)
Parental Education	0.043	0.057^	—	−0.009	−0.005	—	−0.001	0.002	—
	(0.035)	(0.035)		(0.012)	(0.011)		(0.004)	(0.005)
Finances Growing Up	−0.062	0.131	—	−0.006	−0.045^	—	−0.015	−0.010	—
	(0.095)	(0.088)		(0.029)	(0.027)		(0.010)	(0.011)
Prof./Mgr. Parent(s)	0.226	−0.189	—	0.031	0.009	—	−0.023	−0.009	—
	(0.285)	(0.254)		(0.086)	(0.076)		(0.027)	(0.032)
Constant	2.139^	0.076	—	0.048	0.036	—	0.769**	0.837**	—
	(1.118)	(0.915)		(0.034)	(0.030)		(0.121)	(0.132)
Observations	995	946		992	944		993	945
(Pseudo) R²	0.125	0.114					0.134	0.187
Alpha				0.566	0.453

Note. Unstandardized estimates are shown and they appear above standard errors given in parentheses. All models also include maternal and paternal health and number of siblings. Differences of warmth coefficients by respondent sex were tested using pooled interaction models. Significance of interaction coefficients from these models are reported in the (Diff) column. ns = not significant.

p < .01. *p < .05. ^p < .10 (two-tailed).

Number of Chronic Conditions

Maternal warmth buffered chronic health conditions in adulthood among female respondents (Table 2; b = −.208, p < .01) but not males (b = −.049, ns). In contrast, paternal warmth protected against conditions for males (b = −.156, p < .01) but not for females (b = .031, ns). Within a pooled interaction model, sex-matching emerged for both types of warmth (ps < .05). Figure 1 depicts matching for paternal warmth based on pooled model results. For males, paternal warmth provides protection against about one condition (across its domain) whereas for females no protection is evident.

Figure 1.

Effect of paternal warmth during childhood on rate of chronic conditions in adulthood.

Functional Limitations

Maternal warmth protected against functional limitations for female respondents only (Table 2; b = −.065, p < .01, male b = .030, ns), whereas paternal warmth was protective only for males (b = −0.052, p < .01, female b = .027, ns). A pooled model demonstrated that both types of warmth shaped adulthood health through sex-matching (ps < .01). Figure 2 depicts both interactions based on pooled model results. Here, one can see that opposite-sex parental warmth produces no health gains and in fact is predicted to produce slight increases in functional limitations that do not differ from zero. In contrast, same-sex warmth produces gains in health or fewer limitations with increasing warmth.

Figure 2.

Effects of maternal and paternal warmth on logged functional limitations in adulthood.

Discussion

Very few studies have tracked the adulthood health effects of parental warmth or support experienced during childhood, focusing instead on the life-course sequelae of originating socioeconomic status or more distal or extreme measures of parenting such as parental absence, maltreatment, or neglect. In this study, I added to research by employing rich measures of early parental warmth and by examining whether sex-matching shapes the later adulthood physical health outcomes of such warmth. Across three general physical health measures collected during middle to late adulthood, I found strong support for sex-matching on parental warmth. None of these measures revealed results contrary to the sex-matching hypothesis. In other words, no opposite sex-matching effects were apparent (e.g., daughters benefitting more from paternal warmth than from maternal warmth), contrary to some previous research looking at specific psychological and physical health outcomes of parental bonds (Hamil-Luker & O’Rand, 2007; Powell & Downey, 1997; Stewart-Brown et al., 2004). Across all general physical health outcomes, paternal warmth was important only to male adulthood health, whereas maternal warmth was important only to female adulthood health. Effects of maternal and paternal warmth were stronger for same-sex children in five of six cases.

Although some previous research on sex-matching and children’s well-being has focused on single-parent custody arrangements (e.g., Powell & Downey, 1997), this research took as its focus traditional two-parent households containing a biological mother and a father. Importantly, it examined parent–child emotional bonds rather than simply examining custody, as custody does not necessarily imply a supportive parent–child relationship. Within two-parent households, mutual selection processes may be especially likely to operate, such that same-sex parent–child bonds may be especially formative and important to children’s development. Relationships with nonresidential parents, such as in the wake of divorce, may have mixed or sometimes negative effects on children’s well-being (Amato & Gilbreth, 1999; Musick & Meier, 2010).

Continuing on this point, it is important to keep in mind that individuals who came of age in traditional two-parent households are measurably different from the general United States population in a number of ways. First, they are likely to originate in slightly older cohorts (due to lesser odds of traditional family formations among recent cohorts), more likely to be male, and less likely to be a racial or ethnic minority. Furthermore, the range of health disparities is perhaps more limited among these individuals relative to the general United States population, as previous research has documented the higher levels of mental and physical health that typically result from growing up in a two-parent household relative to nontraditional types such as single-parent or stepparent households (Amato, 2005; but see Musick & Meier, 2010).

Indeed, it is probably not the case that the current findings carry over to other household arrangements. For instance, Powell and Downey (1997) powerfully showed that a child’s well-being is not compromised when they are placed in the custody of an opposite-sex parent. Also, Biblarz and Stacey (2010) show that children in households with two fathers or two mothers do not fare any worse on well-being and attainment outcomes. Thus, coming of age within a traditional two-parent household may well serve as a precondition for observing sex-matching effects on adult children’s physical health. Reciprocally, a parent may not deem it desirable to focus on a same-sex child instead of an opposite-sex child unless this choice between children is inherent to the family structure. Overall, it may be the case that a lack of warmth from a same-sex parent within a traditional two-parent household is likely to be construed by a child as rejection or disapproval from a same-sex role model. That is, rejection from a same-sex parent within a domestic environment where one is readily available and present on a frequent basis (as in a traditional household) carries ramifications for children’s later physical health. Note that this argument does not readily pertain to single-parent households or households with two mothers or two fathers, as in such cases a same-sex parent may not be reasonably or readily available—and thus may not be relevant to a children’s favorable emotional and social development.

Auxiliary analyses using continuous and dichotomous treatments of respondent age both showed that the present findings did not differ by respondent cohort (i.e., were the same across younger and older respondents). In total, then, the specificity of the finding to traditional two-parent households and the absence of cohort effects suggest that this pattern of health disparities continues to be relevant to today’s adults but specifically for adults raised in traditional two-parent households. The relevance of sex-matching effects to households containing stepparents awaits further study, as MIDUS does not provide an adequately large sample of stepmother and stepfather families to provide a satisfactory test.

Given previous investigations of two-parent households, differing life outcomes can produce theoretically diametric results. Certain adulthood health outcomes seem to depend on parent–child gender matching whereas others do not (Hamil-Luker & O’Rand, 2007; Mallers et al., 2010; Rothrauff et al., 2009). Whether father–son or mother–daughter bonds are especially important depends on the analytic domain (i.e., mental or physical health), the specific outcome within that domain (e.g., depression or self-esteem; mortality or self-rated health), and also on whether parental absence or mortality is being examined rather than parental warmth or support. Future research will need to develop domain-, outcome-, and operationalization-specific theories for sex-matching effects. At present, it is unclear why sex-matching effects are not consistent across domains and outcomes and across diverse operationalizations of parent–child bonds.

In digesting the current study’s findings, it is important to bear in mind a number of limitations. Perhaps most obviously, parental warmth was measured retrospectively. As such, data may be biased by the respondent’s current life circumstances. For example, poor health could be linked to a tendency to remember parenting as unsupportive. However, it can be difficult to distinguish between bias and random error in retrospective reports of parenting, and such reports seem valid and reliable in any case (Kendler, Myers, & Prescott, 2000; McCrae & Costa, 1988) and are quite common in the literature. Supposed golden standards of parenting measurement, such as laboratory-based observations of parent–child interactions during the first years of life, can also be error-prone and may not map directly onto children’s recollections of interactions with parents. Due to their personal salience, subjective recollections may be more important for later health outcomes than what “in fact” transpired from a third-person perspective. The most robust approach seems to be using multiple informants whenever possible (e.g., Kendler et al., 2000).

Warmth may not be optimally measured as an independent construct. Ample theoretical and empirical work points to the fact that parenting styles, rather than being discrete properties of parents, actually shape and reflect children’s personalities and any problem behaviors exhibited by children (e.g., Kochanska, Friesenborg, Lange, & Martel, 2004). Thus, children’s early expressions of warm or defiant behavior could be driving much of the observed variability in parental support and warmth and potentially also may explain some of the variation in adulthood health outcomes as well.

For self-rated health, I found sex-matching for maternal health only, whereas for physical outcomes (chronic conditions and functional limitations), full matching effects were present. In one sense, this is unexpected, given self-rated health’s well-established links to various measures of morbidity, including morbidity related to physical impairment (Idler & Benyamini, 1997; Jylhä et al., 2006). However, Mirowsky and Ross (2003) make the important point that self-rated health is a uniquely encompassing measure, in that it reflects the presence of physical well-being as well as the absence of disease. Thus it is in keeping with the World Health Organization’s espoused focus on health as a general physical state indicated by thriving above and beyond low malaise (Jylhä, 2009; Ross & Wu, 1995).

Researchers should strive to reveal the “why” behind parental effects whenever possible. For instance, Shaw et al. (2004) showed how impaired social relationships and lower psychological well-being explain much of the association between parental warmth during childhood and later adulthood health. Kestilä et al. (2006) similarly focused on the behavioral mediator of smoking to shed light on disease pathways. Although these mediators are generally relevant, the “why” behind sex-matching effects of parental warmth constitutes fruitful grounds for research. Relevant mediators for sex-matching effects perhaps include marginalized gender identities resulting from suboptimal same-sex socialization experiences or impaired social relationships resulting from the same.

A variety of other mediators should be examined. For instance, health behaviors such as substance abuse, smoking, drinking, overeating, or lack of moderate or vigorous physical activity may play into sex-matching pathways. Specifically, if low levels of warmth from a same-sex parent are indeed experienced as an especially powerful form of social rejection, then mediation may be similar to other documented pathways in which social isolation or rejection is related to physical health through compromised health behaviors (Repetti et al., 2002). Psychobiological pathways are potentially relevant as well, as elevated levels of stress hormones and chronic activation of threat-related neurobiological systems may well figure into sex-matching effects. While psychobiological pathways have established relevance to the general effects of parental warmth irrespective of sex-matching between parents and children, sex-matching may still operate along similar pathways, as same-sex parental rejection may represent a unique domestic stressor that is specific to traditional two-parent households.

In total, what the current research offers is a new lens through which to view general health differences among middle- to late-aged adults who came of age in traditional households. These pathways are decidedly gendered and are products of mutually reinforcing effects between parents and their children. As they pertain to physical health, these pathways provide a new look at the important issue of how deficits in functioning may be traceable to emotional processes set into motion early in life.

Footnotes

Acknowledgements

Thanks to Jennifer Glanville, Steve Hitlin, Robin Simon, and Mark Walker for their comments on an earlier draft.

Author’s Note

Neither the Inter-University Consortium for Political and Social Research (ICPSR) nor the John D. and Catherine T. MacArthur Foundation and the Research Network on Successful Midlife Development bear any responsibility for any interpretations or inferences herein.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: A Presidential Graduate Fellowship at The University of Iowa facilitated the preparation of this article.

References

Amato

P. R.

(1994). Father-child relations, mother-child relations, and offspring psychological well-being in early adulthood. Journal of Marriage and the Family, 4, 1031-1042.

Amato

P. R.

(2005). The impact of family formation change on the cognitive, social, and emotional well-being of the next generation. Future of Children, 15, 75-96.

Amato

P. R.

Gilbreth

J. G.

(1999). Nonresident fathers and children’s well-being: A meta-analysis. Journal of Marriage and the Family, 61, 557-573.

Barnett

R. C.

Marshall

N. L.

Pleck

J. H.

(1992). Adult son-parent relationships and their associations with sons’ psychological distress. Journal of Family Issues, 13, 505-525.

Biblarz

T. J.

Stacey

(2010). How does the gender of parents matter? Journal of Marriage and Family, 72, 3-22.

Blackwell

D. L.

Hayward

M. D.

Crimmins

E. M.

(2001). Does childhood health affect chronic morbidity in later life? Social Science & Medicine, 52, 1269-1284.

Bowlby

(1979). The making and breaking of affectional bonds. London, England: Tavistock.

Boyd

C. J.

(1989). Mothers and daughters: A discussion of theory and research. Journal of Marriage and the Family, 51, 291-301.

Braveman

Barclay

(2009). Health disparities beginning in childhood: A life-course perspective. Pediatrics, 124, S163-175.

10.

Buchmann

DiPrete

T. A.

(2006). The growing female advantage in college completion: The role of family background and academic achievement. American Sociological Review, 71, 515-541.

11.

Case

Fertig

Paxton

(2005). The lasting impact of childhood health and circumstance. Journal of Health Economics, 24, 365-389.

12.

Caspi

Bem

D. J.

Elder

G. H.

Jr. (1989). Continuities and consequences of interactional styles across the life course. Journal of Personality, 57, 375-406.

13.

Caspi

Silva

P. A.

(1995). Temperamental qualities at age three predict personality traits in young adulthood: Longitudinal evidence from a birth cohort. Child Development, 66, 486-498.

14.

Chodorow

(1978). The reproduction of mothering: Psychoanalysis and the sociology of gender. Berkeley: University of California Press.

15.

Conger

R. D.

Donnellan

M. B.

(2007). An interactionist perspective on the socioeconomic context of human development. Annual Review of Psychology, 58, 175-199.

16.

Danese

Parlante

C. M.

Caspi

Taylor

Poulton

(2007). Childhood maltreatment predicts adult inflammation in a life-course study. Proceedings of the National Academy of Sciences, 104, 1319-1324.

17.

Elder

G. H.

Jr. (1998). The life course as developmental theory. Child Development, 69, 1-12.

18.

Fischer

C. S.

(2011). Still connected. New York, NY: Russell Sage Foundation.

19.

Gorman

B. K.

Read

J. N. G.

(2006). Gender disparities in adult health: An examination of three measures of morbidity. Journal of Health and Social Behavior, 47, 95-110.

20.

Haas

(2008). Trajectories of functional health: The “long arm” of childhood health and socioeconomic factors. Social Science & Medicine, 66, 849-861.

21.

Hamil-Luker

O’Rand

A. M.

(2007). Gender differences in the link between childhood socioeconomic conditions and heart attack risk in adulthood. Demography, 44, 137-158.

22.

Hayward

M. D.

Miles

T. P.

Crimmins

E. M.

Yang

(2000). The significance of socioeconomic status in explaining the racial gap in chronic conditions. American Sociological Review, 65, 910-930.

23.

Hayward

M. D.

Gorman

B. K.

(2004). The long arm of childhood: The influence of early-life social conditions on men’s mortality. Demography, 41, 87-107.

24.

Herd

(2010). Education and health in late-life among high school graduates: Cognitive versus psychological aspects of human capital. Journal of Health and Social Behavior, 51, 478-496.

25.

Hughes

M. E.

Waite

L. J.

(2009). Marital biography and health at mid-life. Journal of Health and Social Behavior, 50, 344-358.

26.

Idler

E. L.

Benyamini

(1997). Self-rated health and mortality: A review of twenty-seven community studies. Journal of Health and Social Behavior, 38, 21-37.

27.

Jylhä

(2009). What is self-rated health and why does it predict mortality? Social Science & Medicine, 69, 307-316.

28.

Jylhä

Volpato

Guralnik

J. M.

(2006). Self-rated health showed a graded association with frequently used biomarkers in a large population sample. Journal of Clinical Epidemiology, 59, 465-471.

29.

Kendler

K. S.

Myers

Prescott

C. A.

(2000). Parenting and adult mood, anxiety and substance use disorders in female twins: An epidemiological, multi-informant, retrospective study. Psychological Medicine, 30, 281-294.

30.

Kestilä

Koskinen

Martelin

Rahkonen

Pensola

Pirkola

. . . Aromaa

(2006). Influence of parental education, childhood adversities, and current living conditions on daily smoking in early adulthood. European Journal of Public Health, 16, 617-626.

31.

Kochanska

Friesenborg

A. E.

Lange

L. A.

Martel

M. M.

(2004). Parents’ personality and infants’ temperament as contributors to their emerging relationship. Journal of Personality and Social Psychology, 86, 744-759.

32.

Lundberg

(2005). Sons, daughters and parental behavior. Oxford Review of Economic Policy, 21, 340-356.

33.

Lundberg

McLanahan

Rose

(2007). Child gender and father involvement in fragile families. Demography, 44, 79-92.

34.

Maccoby

E. E.

(2003). How do children affect parents’ marriage and other family relationships? The gender of child and parent as factors in family dynamics. In Crouter

A. C.

Booth

(Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 191-206). Mahwah, NJ: Lawrence Erlbaum.

35.

Mallers

M. H.

Charles

S. T.

Neupert

S. D.

Almeida

D. M.

(2010). Perceptions of childhood relationships with mother and father: Daily emotional and stressor experiences in adulthood. Developmental Psychology, 46, 1651-1661.

36.

Mammen

(2011). Fathers’ time investments in children: Do sons get more? Journal of Population Economics, 24, 839-871.

37.

Marks

N. F.

Jun

Song

(2007). Death of parents and adult psychological and physical well-being: A prospective U.S. national study. Journal of Family Issues, 28, 1611-1638.

38.

McCrae

R. R.

Costa

P. T.

Jr. (1988). Do parental influences matter? A reply to Halverson. Journal of Personality, 56, 445-449.

39.

McPherson

Smith-Lovin

Cook

J. M.

(2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415-444.

40.

Min

M. O.

Minnes

Kim

Singer

L. T.

(2013). Pathways linking childhood maltreatment and adult physical health. Child Abuse & Neglect, 37, 361-373.

41.

Mirowsky

Ross

(2003). Education, social status, and health. New York, NY: Aldine de Gruyter.

42.

Mikulincer

Shaver

P. R.

(2003). The attachment behavioral system in adulthood: Activation, psychodynamics, and interpersonal processes. In Zanna

M. P.

(Ed.), Advances in experimental social psychology (Vol. 35, pp. 53-152). San Diego, CA: Academic Press.

43.

Mokrue

Chen

Y. Y.

Elias

(2011). The interaction between family structure and child gender on behavior problems in urban ethnic minority children. International Journal of Behavioral Development, 36, 130-136.

44.

Musick

Meier

(2010). Are both parents always better than one? Parental conflict and young adult well-being. Social Science Research, 39, 814-830.

45.

Poon

C. Y. M.

Knight

B. G.

(2012). Emotional reactivity to network stress in middle and late adulthood: The role of childhood parental emotional abuse and support. The Gerontologist, 52, 782-791.

46.

Powell

Downey

D. B.

(1997). Living in single-parent households: An investigation of the same-sex hypothesis. American Sociological Review, 62, 521-539.

47.

Raley

Bianchi

(2006). Sons, daughters, and family processes: Does gender of children matter? Annual Review of Sociology, 32, 401-421.

48.

Repetti

R. L.

Taylor

S. E.

Seeman

T. E.

(2002). Risky families: Family social environments and the mental and physical health of offspring. Psychological Bulletin, 128, 330-366.

49.

Ridgeway

C. L.

Correll

S. J.

(2004). Unpacking the gender system: A theoretical perspective on gender beliefs and social relations. Gender & Society, 4, 501-531.

50.

Ross

C. E.

(1995). The links between education and health. American Sociological Review, 60, 719-745.

51.

Rostila

Saarela

J. M.

(2011). Time does not heal all wounds: Mortality following the death of a parent. Journal of Marriage and Family, 73, 236-249.

52.

Rothrauff

T. C.

Cooney

T. M.

J. S.

(2009). Remembered parenting styles and adjustment in middle and late adulthood. Journal of Gerontology: Social Sciences, 64, 137-146.

53.

Russek

L. G.

Schwartz

G. E.

(1997). Feelings of parental caring predict health status in midlife: A 35-year follow-up of the Harvard mastery of stress study. Journal of Behavioral Medicine, 20, 1-13.

54.

Ryff

C. D.

Singer

B. H.

Palmersheim

K. A.

(2004). Social inequalities in health and well-being: The role of relational and religious protective factors. In Brim

O. G.

Ryff

C. D.

Kessler

R. C.

(Eds.), How healthy are we? A national study of well-being at midlife (pp. 90-123). Chicago, IL: University of Chicago Press.

55.

Sampson

R. J.

Laub

J. H.

(1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press.

56.

Shaw

B. A.

Krause

Chatters

L. M.

Connell

C. M.

Ingersoll-Dayton

(2004). Emotional support from parents early in life, aging, and health. Psychology and Aging, 19, 4-12.

57.

Stewart-Brown

S. L.

Fletcher

Wadsworth

M. E. J.

(2004). Parent-child relationships and health problems in adulthood in three UK national birth cohort studies. European Journal of Public Health, 15, 640-646.

58.

Suitor

Pillemer

(2006). Choosing daughters: Exploring why mothers favor adult daughters over sons. Sociological Perspectives, 49, 139-161.

59.

Taylor

S. E.

Baldwin

M. W.

Seeman

T. E.

(2011). Early adversity and adult health outcomes. Development and Psychopathology, 23, 939-954.

60.

Turrell

Lynch

J. W.

Leite

Raghunathan

Kaplan

G. A.

(2007). Socioeconomic disadvantage in childhood and across the life course and all-cause mortality and physical function in adulthood: Evidence from the Alameda County Study. Journal of Epidemiology and Community Health, 61, 723-730.

61.

Umberson

Chen

M. D.

(1994). Effects of a parent’s death on adult children: Relationship salience and reaction to loss. American Sociological Review, 59, 152-168.

62.

Umberson

Pudrovska

Reczek

(2010). Parenthood, childlessness, and well-being: A life course perspective. Journal of Marriage and Family, 72, 612-629.

63.

Videon

T. M.

(2005). Parent-child relations and children’s psychological well-being: Do dads matter? Journal of Family Issues, 26, 55-78.

64.

Walker

E. A.

Gelfand

Katon

W. J.

Koss

M. P.

Von Korff

Bernstein

Russo

(1999). Adult health status of women with histories of childhood abuse and neglect. American Journal of Medicine, 107, 332-339.

65.

Wheaton

Reid

(2008). The role of timing vs. duration in the cumulative work history effects of job exits and nonemployment on women’s mental health. In Turner

H. A.

Schieman

(Eds.), Advances in life course research (Vol. 13, pp. 195-232). New York, NY: Kluwer Academic/Plenum.

66.

Yang

(2006). How does functional disability affect depressive symptoms in late life? The role of perceived social support and psychological resources. Journal of Health and Social Behavior, 47, 355-372.