The Impact of Stepfamily Structure on Older Parents’ Frequency of Contact With and Care Receipt From Adult Biological and Stepchildren in the Netherlands

Abstract

Objectives

Older people are increasingly entering their later years in stepfamilies. Because adult children play a central role in older parents’ support networks, there is concern that the generally weaker intergenerational ties found in stepfamilies may imply an impending deficit in the care available to stepparents. It is currently unclear whether there are differences across stepfamily types including stepfamilies with only biological children. The aim of the study is to examine whether there are differences in contact frequency with and care receipt from adult biological and stepchildren in biological and different types of stepfamilies.

Methods

Data are from the Longitudinal Aging Study Amsterdam (1992–2022; 10 observations); respondents’ ages varied between 54 and 101. An average of 3.7 observations are available from 2,761 parents in biological families and 647 parents in stepfamilies.

Results

Parents in biological families and in stepfamilies with joint children had more contact than in other stepfamily types. There was less contact in stepfamilies with biological and stepchildren formed in midlife and in families with only stepchildren. There were small differences in care receipt; the lowest likelihood was in composite families.

Discussion

Our study challenges the idea that the relationships of the adult child to older parents in all stepfamilies are weaker than in biological families and points to the importance of considering that only some stepfamilies are vulnerable in terms of contact frequency. We query whether stepfamilies are resilient, for example, to greater pressures from a sharp increase in care needs for one or both parents.

One of the consequences of population aging is that there will be an increase in older people who have complex health issues and need care in the future. Many Western countries are facing workforce shortages and capacity issues in professional care, which in turn increase the demand for informal care (Wieczorek et al., 2022). Most informal care received in later life is provided by family members, particularly partners and adult children (Broese van Groenou & de Boer, 2016).

Due to the rapid increase in divorce and remarriage, ties between parents and children have become more complex. Older parents today have fewer biological children and more stepchildren than in prior decades (Seltzer, 2019; Wiemers et al., 2019). Stepfamilies of older people are created in part by repartnering later in life (Brown & Lin, 2012). Because adult children play an important role of caregiver to older adults, there is concern that the generally weaker intergenerational ties found in stepfamilies may imply an impending deficit in the care available to stepparents, even those with biological children (Carr & Utz, 2020).

It is currently unclear whether there are differences between types of stepfamilies by the frequency with which older parents have contact with and receive care from biological and stepchildren (Carr & Utz, 2020; Lin & Seltzer, 2024; Patterson et al., 2022). Intergenerational contact indicates a potential to respond to parental care needs (Silverstein & Bengtson, 1997). As caregiving by adult children is an important source of care provision to the growing proportion of the population, it is important to answer three questions: (1) What is the presence and number of various types of later-life stepfamilies in the Netherlands; (2) How does stepfamily structure influence contact frequency with the group of adult biological and stepchildren; and (3) How does the stepfamily structure influence the receipt of care from these children.

Family structure and contact frequency and care receipt

We extend past research in two ways. First, an important and consistent finding in previous research is that adult children’s step relationships are significantly weaker and less stable than biological relationships (Kalmijn et al., 2019; Patterson et al., 2022; Pezzin et al., 2008; Wiemers et al., 2019). However, there is a need to consider all the children in the family together. Although dynamics in caregiver networks are complicated (Szinovacz & Davey, 2013), having frequent contact with at least one biological or stepchild as the (potential) primary caregiver can be considered to be important for a parent in need (Lin & Wolf, 2020). A focus on the average contact across multiple intergenerational dyads is less informative: a high average indicates adequate contact and (potential) care, but a lower average masks whether all children have some contact or caregiving or whether one child is the primary caregiver and the others have little to no contact or caregiving. In families with multiple children, children often share care of their older parent (Tolkacheva et al., 2014), but usually it is one child who has most caregiving responsibilities, oversees the care situation, and makes decisions about the cared-for parent. Recent figures are not readily available, but Cicirelli (1992) showed that in 79% of families there was one primary child caregiver (sometimes with some backup care from other siblings); in other families siblings’ caregiving was shared equally. We take the parental perspective and look at the detailed variation in stepfamily types by examining information on all adult biological and stepchildren in the family rather than parent–child dyads. More specifically, we study the contact frequency in the parent’s most contacted relationship (meaning other children have contact as often or less often) and whether the parent receives care from children, regardless of whether they are biological or stepchildren. Our approach fits with the family system theory, a useful framework for understanding the family structure of older parents and its association with contact with and care receipt from the group of children (Cox & Paley, 2003). The family is seen as an ecological setting within which the connections among all individuals in the family influence the outcomes rather than the individual dyads (Patterson et al., 2022).

In stepfamilies, at least one of the parents has a child or children from previous unions; there are many unique stepfamily types (Ganong & Sanner, 2023). Based on the type of relationship the focal parent had (biological or not), and partner’s relationship with the children, we compared biological families and various types of stepfamily structure: reconstructed biological families with joint children, and families where the respondent had only biological children, biological and stepchildren, or had only stepchildren.

Previous studies focused on contact with the particular adult children in the context of biological versus stepfamilies and found that parents maintained less contact with their stepchildren (Kalmijn, 2013). Hämäläinen et al. (2024) found that children gave more support to their biological parent than to their stepparent. Pezzin et al. (2008) found that stepchildren are less likely than biological children to provide care to older parents; Wiemers et al. (2019) had similar results. Pezzin et al. (2008) did not find a difference between biological children in biological and in stepfamilies. Parents perceive lower levels of satisfaction and closeness with their adult stepchildren than biological children (Arránz Becker et al., 2013; Steinbach & Hank, 2016). Patterson et al. (2022) compared biological families and stepfamilies. Older parents in need of care were more than twice as likely to receive care from their adult biological children compared to parents who had stepchildren. There were few cases where different stepfamily types were compared. van der Pas and van Tilburg (2010) found that contact by individual biological children with their parent is perceived as more often regular and important in biological families and stepfamilies with both biological and stepchildren compared with other stepfamilies. Schoeni et al. (2022) found that parents spent the most time with children in stepfamilies with joint children, followed by biological families, and then other stepfamilies. Lin and Seltzer (2024) also found differences between these three types. In couples’ appraisals of positive parent–child interactions, stepfamilies with joint children and biological families were similar and exceeded other stepfamilies. For negative interactions, the contrast was between biological and stepfamilies. The “other” stepfamily types did not differ from each other.

Second, comparisons were made for the timing of the start of the stepfamily to provide a comprehensive assessment. The timing of the stepparent’s arrival in the child’s life may influence the quality of the intergenerational relationship. In taking the child’s perspective, Kalmijn (2013) found that the longer parents and children were together in childhood, the more contact there was in adulthood, more support was exchanged, and the child was more positive about the relationship. Hämäläinen et al. (2024) had similar findings. Detailed retrospective data allowed us to examine the ages of biological children when the stepparent entered the family, and the ages of stepchildren when they entered the stepfamily. We differentiated between stepfamilies formed when children were young (in parents’ “midlife”) and when they were grown-up (in “later life”).

The current study

The Netherlands is a good example of a modern Western European country. Data for our study covers three decades. Dutch divorce and marriage rates were average for Western Europe (Kalmijn, 2007) but are lower than U.S. divorce rates (Amato & James, 2010). The frequency of contact between adult children and older parents is also similar to that of other Western European countries (Hank, 2007).

The incidence of care need has been regarded as the starting point of caregiving behaviors, including frequent intergenerational contact (Broese van Groenou & de Boer, 2016). Previous research specified the care need in different ways, such as a decline in self-reported health (Chen et al., 2021), the incidence or increase of functional and cognitive impairments (Pezzin et al., 2015), increased frailty (Lin & Wolf, 2020), and a higher chronological age (Potter, 2019). The actual provision of care is facilitated or limited by the presence of other types of helpers in the family, the larger social network, and the community (Broese van Groenou & de Boer, 2016).

Method

Respondents

Data were obtained from the Longitudinal Aging Study Amsterdam (LASA; Hoogendijk et al., 2020). Samples of men and women born between 1908 and 1957 were taken from the population registers of three Dutch cities and six surrounding small municipalities in 1992, 2002, and 2012, with follow-up observations every 3 or 4 years. See Supplementary Figures 1 and 2 for sample sizes, attrition, and selection across observations.

The mode of data collection was a face-to-face interview (92%); telephone interviews (6%) and interviews with a proxy (2%) were used when a respondent was unable to participate in the face-to-face interview. We analyzed data from 1,866 men and 1,542 women with a mean of 3.7 observations and a maximum follow-up of 29.5 years. Ninety-five percent were born in the Netherlands. In the sample of pooled observations, the mean age was 69.8 years (range 54–101; Table 1).

Table 1.

Descriptive Statistics

Variable	N observations	N respondents	Mean	SD	Percentage
Highest contact frequency among adult biological and stepchildren in family
Every day	11,328				35
Few times a week	11,328				39
Once a week	11,328				18
Not frequent or not important—once a fortnight	11,328				7
Care receipt from the group of adult biological and stepchildren	12,702				4
Number of adult biological and stepchildren (1–8)		3,408	3.0	1.7
Care by partner	12,702				18
Other informal care	12,702				2
Paid care	12,702				19
Female (vs male)		3,408			45
Time since first observation (0–29 years)	12,730		6.5	6.3
Age (54–101 years)	12,730		69.8	8.0
Physical functioning (6–30)	12,719		27.8	4.1
Cognitive functioning (0–16)	12,406		14.8	1.6

Note: SD = standard deviation.

The LASA study has been conducted in line with the Declaration of Helsinki. LASA has received approval from the medical ethics committee of VU University Medical Center (IRB numbers: 92/138, 2002/141, and 2012/361).

Measures

Identification of children

We applied three methods to identify children. First, in the baseline interviews, we asked about the number of children. It was explained that we were asking about all biological children, as well as step, adopted, and foster children. Second, in all observations during the face-to-face interview, the personal network was delineated by asking: “Name the people you have frequent contact with and who are also important to you” (van Tilburg, 1998). On average, more than 17 people were identified. Criteria for importance were left to the respondent’s interpretation, and only those 18 years and older could be considered. Contact frequency with younger children is unique and less comparable because they may not play substantial roles in providing practical and emotional support to older adults, particularly in ways that significantly affect the older adult’s social support or health outcomes. Third, follow-up observations at T5, T6, and T9 updated information on all children. In all methods, follow-up questions included name, gender, age, the type of relationship, and other characteristics. At each observation, respondents were asked if any child had died, and if so, their names.

We defined stepchildren as the biological children of the partner and not biological children of the respondent. Stepchildren became former stepchildren when the partnership ended. Adopted children were treated as biological children. Foster children were not counted.

Care receipt

Respondents in all types of data collection were asked whether they received any personal care or help with household and nursing tasks. Later questions were about the types of their caregiver, for example, the partner, children, other kin, nonkin, publicly or privately paid professional carers. Care receipt was identified as parents receiving care from the group of children, regardless of whether they were biological or stepchildren.

Use of other care options

We categorized care providers into the following categories: partner, other informal caregivers, and paid care.

Contact frequency

After the network delineation, for all relationships the respondents were asked about the frequency of contact they had. The eight answering possibilities varied between “never” and “daily.” Coresiding biological and stepchildren were assumed to have daily contact. This variable provided information about individual children.

Partner status and duration of partnership

At baseline, partner status was established. Unions could be either same-sex or opposite-sex, and in marriage or not, and cohabiting or not. Respondents were asked the year and month the union had begun. Follow-up observations asked whether the partner status had changed, and if so, when. Additional information on the date of (dissolution of) the union was taken from the population register.

Care need

Physical functioning was self-reported for six activities of daily living, for example, walking up and down 15 steps. Cronbach’s alpha = 0.83. The sum score ranged from 6 “cannot do it at all” to 30 “no difficulties” on all items. Cognitive functioning was measured by a 9-item version of the Mini-Mental State Examination (Folstein et al., 1975). The sum score ranged from 0 = poor to 16 = excellent. Age was measured as chronological age. Absolute correlations between these three indicators ranged between 0.20 and 0.31.

Procedure

For family structure, we took into account whether there were biological and stepchildren, whether the partner was a stepparent, and when the stepfamily started (van der Pas & van Tilburg, 2010). For the latter, we distinguished between whether the youngest child was 15 years old or younger (in the Netherlands there is compulsory education up to age 15), or older.

In the analysis of contact frequency, we performed multinomial logistic regression. Because of the small numbers we merged several categories. Biological and stepchildren not identified in the personal network were given a category of “not frequent or not important” contact, along with children with whom contact was “once every two weeks” or less frequent. The threshold in our measure was higher than that used by Arránz Becker and Hank (2022) because the cell frequency of low contact frequency was often small (Supplementary Table 1). Other categories were “once a week,” “a few times a week,” and “every day” contact. We identified the child with the highest frequency of contact as representative of the group of biological and stepchildren. Observations were nested in respondents. The longitudinal design was incorporated by calculating the time since the first observation for each observation. Predictors were family structure and number of biological and stepchildren. Control variables were gender, age, and physical and cognitive functioning. To avoid multicollinearity with time, respondent’s age was calculated as the mean across their observations. In the analysis of care receipt, we performed multilevel binomial logistic regression. The independent variables were similar to the analysis of contact frequency; we added the use of other care options. In a cross-tabulation of family structure and received care, the expected cell frequency was found to be low (Supplementary Table 5), so we also present results for a condensed typology of family structure. For both dependent variables, we tested whether differences between categories of family structure changed over time. The number of children and control variables were centered around the grand mean to facilitate interpretation of the coefficients. The small number of missing values was imputed with the grand mean. To interpret the results, we calculated marginal probabilities for selected values of independent variables.

Results

Description of Family Structure

We distinguished 10 types of family structures (Table 2; the seven main types are illustrated in Figure 1). The vast majority of respondents (81%) had a biological family. Both respondent and partner were the joint parents of an average of 2.9 children. Ninety-seven percent were in the first union; others had a former union without children. The average length of the union was 39 years, and the union began on average at the respondent’s age of 25.7 years. Three types of stepfamilies were reconstructed into biological families by joint children. Very few respondents had joint children and children who were biological children from the previous union of both the respondent and the partner, and these families are not considered below. There were also reconstructed biological families with respondents who had joint and biological children (the partner being a biological and stepparent, respectively), and with respondents who had joint and stepchildren. In this family structure, the family started in midlife. Furthermore, in additional partner families, only the respondent had biological children and the partner was stepparent. In composite families, both the respondent and the partner had biological children from a former union; thus, both had stepchildren. Finally, we had respondents with only stepchildren (additional respondent families). Additional partner and additional respondent families have the same structure but differ because we focus on respondent’s contact with and care from children. Respondents had only biological children in 50% of the stepfamilies. The distinction stepfamilies formed in midlife and later in life corresponded to the average age of respondents (n = 571) at the start of the union, which was 42.9 and 62.1 years, respectively. A relatively large number of stepfamilies (69%) started later in life. Few respondents (<1%) had frequent and important contact with a former stepchild.

Table 2.

Different Family Structures According to Relationship Type of the Respondent and Partner and When the Stepfamily Started (N = 3,408)

Family structure	Type of respondent’s children	Parental position		Starta	N b	N of children (M)		In first union (%)	Length unionc (years; M)	Respondent’s age at start (years; M)
		Respondent	Partner			Biological	Step
Biological	Biological	Biological (CU)	Biological (CU)		2,761 (10,711)	2.9	0	97	39.2	25.7
Reconstructed biological	Joint, biological, step	Biological (FU; CU); step (FUp)	Biological (FU; CU); step (FUr)		3 (6)	2.3	3.0	0	30.7	39.6
	Joint, biological	Biological (FU; CU)	Biological CU; step (FUr)		43 (153)	3.9	0	0	33.1	35.0
	Joint, step	Biological (CU); step (FUp)	Biological (FU; CU)		30 (98)	2.2	1.3	80	32.8	30.8
Additional partner	Biological	Biological (FU)	Step (FUr)	≤15	62 (216)	2.5	0	0	21.9	42.3
				≥16	221 (559)	2.6	0	2	5.4	63.9
Composite	Biological, step	Biological (FU); step (FUp)	Biological (FU); step (FUr)	≤15	40 (139)	2.5	2.0	3	17.1	46.0
				≥16	186 (657)	2.5	2.5	3	5.9	60.8
Addition of respondent	Step	Step (FUp)	Biological (FU)	≤15	23 (73)	0	2.3	57	25.2	39.3
				≥16	39 (118)	0	2.2	36	8.2	57.7

Notes: CU = current union; FU = former union; p = partner; r = respondent.

^aWhen biological children, age of respondent’s youngest biological child; otherwise, age of youngest stepchild. Moment of start is not meaningful if there are joint children.

^b N of respondents (N of observations between parentheses).

^cLength of partnership at the first observation.

Figure 1.

Illustration of the seven main types of families, with three children. C = child; P = partner; R = respondent. For children, white circles are respondent’s biological children (corresponding with respondent’s color), diagonal stripes are joint and gray are stepchildren (corresponding with partner’s color).

Analysis of Contact Frequency

In 35% of the observations, the highest frequency of contact with a child in the family, regardless of whether they were biological or stepchildren, was “every day,” in 39% “a few times a week,” in 18% “once a week,” in 4% “once every two weeks,” in 2% “monthly” or less frequently, and in 2% of families none of the children were identified as having “frequent or important” contact. When the respondent had both biological and stepchildren, the highest frequency of contact was often with a biological child (59% of observations), in 22% it was with a stepchild, and in 20% it was with both a biological child and a stepchild. Looking at the least contacted child in a family, the percentages were 9% “every day,” 25%, 27%, 12%, 13% “monthly or less” and 14% no “frequent or important” contact, respectively. Thus, in one of the four observations, contact with at least one child was scarce or absent. At the parent level, 114 (3.5%) of the 3,196 respondents with biological children identified none of their biological children as someone with whom they had regular contact and who was also important to them; another three respondents reported never having contact, and 19 respondents said they had contact a few times a year or less. Of the 316 respondents with stepchildren, 138 (44%) identified none of them as a network member, and another 12 respondents had contact a few times a year or less.

Regression of the highest contact frequency shows differences across family structure types (F_{(241, 11277)} = 21.1; p < .001). Parameter estimates are shown in Supplementary Table 2, and marginal probabilities are illustrated in Figure 2. Across observations, 34% of respondents in biological families (reference category) had daily contact. In reconstructed biological families, there was daily contact in 43% (p > .05) of observations among respondents with joint and biological children, and 47% (p > .05) of observations among respondents with joint and stepchildren.

Figure 2.

Marginal probabilities of highest frequency of contact among adult biological and stepchildren in family by family structure. *p < .05. **p < .01. ***p < .001.

Contact frequency in other types of stepfamilies differed significantly from biological and reconstructed biological families, and almost all distributions of contact frequency differed among the other types of stepfamilies (Supplementary Table 3). We discuss the differences between midlife and later-life families (shown in Figure 2). In additional partner families, there was daily contact in 18% of observations; in later-life families, there was once-a-week contact in 24% of observations compared to 17% in midlife families. However, looking at the distribution as a whole, we do not see that there is more contact in later-life families. In composite families, about 11% of respondents had daily contact. In later-life families there is more frequent contact: in 41% and 32% of observations there was contact a few times and once a week, respectively, compared to 36% and 26% in midlife families. In additional respondent families, respondents had little contact with their stepchildren. In midlife families, contact in 6% of observations was daily (11% in later-life families); a few times a week in 20% (vs 14%), once a week in 32% (vs 16%), and scarce or absent contact in 60% (vs 41%).

The more children the higher the frequency of contact (F_{(3, 11277)} = 13.7; p < .001). For example, in midlife stepfamilies where the respondent had joint and biological children, 40% of respondents had daily contact when there were two biological children (one jointly with the current partner and one from another partner), compared to 61% of respondents when there were eight or more children. The number of children varied by family structure. When we corrected for these differences, the above estimates for daily contact by family structure changed slightly for most types, but most for later-life stepfamilies with biological and stepchildren (from 12% to 16%), as they have the highest number of children.

All covariates were significant. We present marginal probabilities for selected levels of contact frequency (Supplementary Table 4) among respondents in a stepfamily where the partner was added to the family later in life. Gender differences were small. Female and male respondents had similar levels of daily contact (17%–18%), but female respondents were more likely to have contact a few times a week (38% vs 33%). Over time, contact frequency decreased. At the first observation, 26% of respondents had daily contact. Ten years later (randomly selected, close to the mean), only 14% had daily contact. Adding an interaction between family structure and time did not improve the model (F_{(24, 11253)} = 1.2; p > .05). Older respondents had less often contact than younger respondents. For example, 36% of average 55-year-old respondents had daily contact, compared with 9% of 80-year-old respondents. Physical functioning was positively related to contact frequency but differences were small. Respondents with better cognitive functioning had more often contact, in particular in the category “few times a week.”

Analysis of Care Receipt

Respondents did not often receive care. In 36% of the observations, care was received from some source. When care was received, 48% had care from their partner, 11% from biological or stepchildren, 5% from other informal caregivers, and 52% had paid care.

Results of the regression of care receipt from children are presented in Supplementary Tables 8–10. Due to low expected cell frequencies (Supplementary Tables 5–7), we present results for two variations in condensed categories of family structure (Supplementary Tables 11 and 12). Distinguishing between biological, reconstructed biological, partner addition, composite, and respondent addition, there were differences in the likelihood of receiving care between family structure types (F_{(4, 12682)} = 2.8; p < .05). The lowest likelihood was observed in composite families (3%), but differences were small (other categories were between 4% and 5%; Figure 3). Care receipt was more likely in larger families than in small families (F_{(1, 12682)} = 94.2; p < .001), and 3% of respondents with one child received care compared to 13% of respondents with eight or more children. Distinguishing between biological, reconstructed biological, and formed in midlife and later life, there were no significant differences in the likelihood of receiving care between family structure types (F_{(3, 12683)} = 2.6; p > .05).

Figure 3.

Marginal probabilities of care receipt from the group of adult biological and stepchildren by family structure. (A) Categories of stepfamily formed in midlife or later in life were merged. (B) Categories of stepfamilies with partner addition, composite families, and families with respondent addition were merged.

The probability of receiving care was related to most of the covariates (Supplementary Table 13). Care by the partner and especially by other informal caregivers was associated with a greater likelihood of receiving care from children; probability increased by two and seven percentage points, respectively, among respondents in additional partner families. Receiving paid care reduced the probability by two percentage points. Gender was significantly related; the probability for women was one percentage point higher. In follow-up observations, the probability did not increase, nor was there an interaction with family structure (F_{(4, 12678)} = 0.6; p > .05). Older respondents received more often care from children than younger respondents. For example, the difference between an average 55-year-old respondent and an 80-year-old respondent was four percentage points. Physical functioning was associated with care receipt; the difference between someone with a problem in one domain and without problems was three percentage points. Cognitive functioning was not related to care receipt.

Discussion

In this study, we examined whether the structure of stepfamilies is associated with contact frequency with and care receipt from adult biological and stepchildren. Using LASA data, we show that in the Netherlands 19% of later-life families were stepfamilies. Seven out of 10 stepfamilies started in later life.

The frequency of contact was measured as the highest among all children. Having children does not automatically mean you can count on them for contact or care (Arránz Becker & Hank, 2022; Reczek et al., 2023); a small number of biological parents and almost half of stepparents are “estranged” from all their biological or stepchildren. In families with biological and stepchildren, it is important not to look only at biological children and if that were the case, a stepchild with much (and sometimes more) contact is overlooked in 41% of those families.

We found evidence that there is less contact in stepfamilies than in biological families. In our study, we not only examined whether stepchildren existed but also focused on the differences in family structure. If the focal parent had only biological children in the stepfamily, the partner may be a stepparent to all of the children, but can also be a stepparent to some of the children and have children with the focal parent. In the latter case of families with joint children, contact frequency was relatively high and similar to biological families. This finding is consistent with the findings of Schoeni et al. (2022) and, for positive relationship aspects, with Lin and Seltzer (2024). In contrast, we found differences among other stepfamily types: contact frequency was lowest in families with only stepchildren.

Stepfamilies are formed at different stages of the life course, and it is important to take this into account. When a new parent enters the family shortly after the birth of the focal parent’s biological children, the stepparent can participate in the upbringing and take on a role that is similar to that of a biological parent. At the other end of the spectrum, there are older people who have lost their partner at an advanced age and enter into a new partnership, and the biological and stepchildren themselves can already be middle-aged or older. It is found that a longer duration of the stepfamily coincided with better intergenerational relationships (Arránz Becker et al., 2013), but this was not confirmed in some other studies (Lin & Seltzer, 2024; Schmeeckle et al., 2006). We found not less but more frequent contact in composite stepfamilies formed in later life than those formed in midlife. Apparently, contact with one’s own biological children is not harmed by the presence of the “new” stepchildren. For both parents, the composition of the composite family is the same. When such a family is formed in midlife, each parent has parenting responsibilities for both biological and stepchildren at certain times. In adulthood, children can go to their own biological parent if necessary, and there is no need to fight for attention from a parent by different types of children. Alternatively, adult stepchildren may see their stepparent as a friend or casual acquaintance of their biological parent, or as “support” for their biological parent with whom they feel no obligations or close tie (Ganong et al., 2011). Contact with the stepparent is mainly via their biological parent. This can lead to a low frequency of contact with the stepparent, especially if the stepparent and partner do not live together because of the children (de Jong Gierveld & Merz, 2013). Because of the importance of care from children, what such a relationship means and what it means for filial responsibility can be examined in future research. We did not find such an advantage for contact with parents in stepfamilies with only biological children in later life. Within these stepfamilies there might be less competing interests and the biological parent may be preferred, both in midlife and in later life (Stewart, 2005).

Our study takes a family systems approach to address how different stepfamily types shape contact frequency and care receipt. This approach emphasizes the need to consider the complexity of different stepfamilies. Our study challenges the idea that parents in later-life stepfamilies are less likely to have contact with and receive care from their children than parents in biological families and point to the importance of considering that only some later-life stepfamilies are vulnerable in terms of contact. The uncertainty of expectations and roles within these stepfamilies may affect the family as a whole—including biological relationships and not just step relationships.

We found evidence that over time contact frequency decreases in biological and stepfamilies. This decrease might be related to adult children’s increasing competing interests in caring for their own children and paid work over the life course. We also found that older respondents have less contact than young–old respondents. Children of an older respondent were generally also later in their life course, so that competing tasks can also play a role in the effect of age. The variable for the age of the respondent also reflects generational differences—the data in this study cover 50 birth years. Later generations—the young–old respondents in our study—have fewer children on average than earlier generations, an effect that has already been taken into account in the positive effect of the variable for the number of children on contact frequency. Taking this difference into account, it points us to much more frequent contacts in recent cohorts of older people. Previous analysis of part of the data of the current study (van der Pas et al., 2007) additionally showed that in individual parent–child relationships, more support was exchanged in recent cohorts than in earlier cohorts.

We subsequently examined the influence of stepfamily types on care receipt from adult biological and stepchildren. The prevalence of care receipt was low and differences between family structure types were small. The presence of a partner, who is often the primary caregiver when the respondent received care (Polenick et al., 2017; Swinkels et al., 2022), and the good accessibility of formal care in the Netherlands—although this is under pressure (da Roit, 2013)—make care receipt from children occasional and additional. Also, non-familistic Netherlands has lower expectations for family members to be primary caregivers (Baranowska-Rataj & Abramowska-Kmon, 2019). Still, even if there is little parent–child contact in stepfamilies, it is still possible for care potential to be used if the care needs of a parent in a stepfamily necessitate it.

We end with some limitations and suggestions for future research. First, the small sample size of all stepfamily types not only meant that some parameters had large confidence intervals but also that further differentiation was not possible. Intergenerational relationships in stepfamilies are gendered (Kalmijn, 2013; Lin & Seltzer, 2024), but we preferred a detailed description of family structures to a gender-specific description of families. Moreover, we cannot rule out potential false negatives for care receipt. The differences were small, but we found that composite families had the lowest probability of care receipt. Future research may examine the extent to which children in these stepfamilies are available for care to their stepparent.

Second, despite the lower frequency of contact in stepfamilies, older adults in most stepfamily types seem to receive adequate care relative to biological families. What happens in contact with stepchildren after the partner relationship ends by separation or widowhood needs further investigation. Given the low frequency of contact in families with only stepchildren, we predict that the likelihood of care receipt by former stepchildren is small.

Third, we lack comparable data on contact frequency and care receipt and therefore do not know whether the care received was actually provided by the most contacted child. However, Grundy and Read (2012) demonstrated that frequent contact with children predicted care receipt.

We conclude that only some later-life stepfamilies are vulnerable in terms of contact. This is strongest in stepfamilies with biological and stepchildren formed in midlife and in families where the respondent has only stepchildren. For now, we see limited vulnerability to receiving care. The question is whether stepfamilies can withstand more pressure due to a sharp increase in care needs in one or both parents.

Funding

The Longitudinal Aging Study Amsterdam is supported by a grant from the Netherlands Ministry of Health, Welfare and Sport, Directorate of Long-Term Care. Data collection in 2012-213 was conducted with a grant from the Netherlands Organization for Scientific Research [grant number 480-10-014].

Conflict of Interest

None.

Data Availability

The data underlying this article are available upon request (https://lasa-vu.nl/en/request-data/). A replication packet with the code is available online (https://osf.io/bjksz).

Author Contributions

S. van der Pas planned the study. T. G. van Tilburg performed all statistical analyses. Both authors contributed equally to the writing of the paper.