Social cooperation is a hallmark of human evolution [1], closely associated with the expansion of the brain in the genus Homo. However, at an empirical level, evolutionary perspectives on cooperation have been dominated by research on men, with most studies in high-income western societies [2]. Addressing the experience of cooperation among women in non-western settings, and its health impacts, is therefore a priority [2]. Our study addresses this issue in rural lowland Nepal, focusing on the mothers of children enrolled in a birth cohort.

When it comes to the social component of being a mother, evolutionary biologists have offered two overarching frameworks for our species that have very different implications for social interactions. First, humans are widely presented as cooperative breeders, where costs and responsibilities of raising children are shared by mothers with various other members of the family and social community [3, 4]. Notably, these ‘others’ are not necessarily the mother’s husband/partner and are often other women such as relatives, neighbours or friends [3, 5]. Second, the value of cooperation is predicted to increase with genetic relatedness, promoting shared interests in childcare [6, 7]. On this basis, reduced cooperation, or even tension, might be expected between individuals that are not genetically related.

From a genetic perspective, the ideal source of social support for a mother is expected to be her own mother (maternal grandmother), with whom 50% of alleles are shared. No other category of relative is guaranteed to share such a high proportion of alleles with the mother. Among many non-human primates, as well as most mammals, most species are philopatric and females do indeed remain close to their natal kin [8]. Humans are a notable exception to this pattern, however, as adult women may live with either their maternal kin or their husband’s kin, or in nuclear households, depending on the society and circumstances. This brings into focus the extent to which individuals living in the same household are genetically unrelated. In patrilineal patrilocal societies, for example, mothers commonly reside with their parents-in-law [9], whereas in matrilineal matrilocal societies, they typically reside with their own parents [10].

To date, evolutionary studies of household composition have focused primarily on the implications of the presence/absence of different relatives for maternal fertility and child health. In low-income settings with high mortality risk, for example, the presence of maternal grandmothers has been associated with greater support for maternal reproduction [10, 11], providing advice on child care [12], and better survival and nutritional status of grandchildren [13–15], though some studies are not consistent [6]. In this way, maternal grandmothers may increase their own fitness through maintaining supportive and cooperative relationships with their daughters [16]. Conversely, the mother’s parents-in-law have often been presented in negative light, due to the expectation that they will act in the fitness interests of their son and grandchildren, potentially at a cost to the mother’s own interests [17]. In traditional South Asian societies, for example, where married women generally and live in extended family households, parents-in-law may constrain mothers’ agency and access to resources [18, 19], though again some studies show supportive associations [16, 20–22]. Overall, lack of support and psychosocial stress may adversely affect mothers’ ability to care for their children [23–25], and hence impact maternal fitness.

Recently, the evolutionary theoretical framing of parents-in-law was challenged by Dyble et al. [26], on the grounds that despite only being relatives through marriage, parents-in-law can still gain indirect fitness benefits through cooperating with and supporting their son’s wife. For example, mothers draw on different forms of embodied and social ‘maternal capital’ to invest in their offspring [27]. If genetically unrelated family members were to promote maternal health and autonomy as forms of maternal capital, they could gain fitness rewards, through promoting maternal investment in offspring with whom they share alleles. What, therefore, might constrain the support of parents-in-law for mothers?

A fundamental difference for maternal versus husband’s kin concerns paternity uncertainty. When this is zero in both generations, all four grandparents will share 25% of alleles with each grandchild. If paternity uncertainty is >0, however, each grandparent has a different pathway through which the likelihood of being genetically related to grandchildren is determined (Fig. 1). Only the maternal grandmother remains guaranteed to share 25% of her alleles with each grandchild. Consequently, the husband’s relatives could promote their own fitness not by supporting the mother but simply by controlling her behaviour (often termed ‘mate-guarding’ [28, 29]) in order to minimize paternity uncertainty. Moreover, with no direct ‘genetic stake’ in their son’s wife, parents-in-law could demand from the mother levels of investment in grandchildren and unpaid care work that impose substantial costs on the mother herself. For example, high levels of fertility (benefitting the fitness of parents-in-law) may lead to chronic maternal undernutrition, a scenario recognized in public health as ‘maternal depletion’ syndrome [30, 31]. If this strategy were to severely impair the mother’s economic or childcare productivity, parents-in-law could potentially minimize any fitness costs by replacing her with another wife for their son, while keeping the first wife’s offspring. There may even be an economic incentive for this, where wives bring dowry with marriage.

Paternity uncertainty takes on particular importance in patriarchal patrilocal societies, common in South Asia, where the transfer of material property down the male line (representing long-term paternal investment) will only benefit the alleles of a woman’s marital kin if her offspring are indeed her husband’s [32]. This perspective may help understand why norms for controlling the behaviour of women of reproductive age are common in such societies. In the Maithili-speaking Madhesi population of lowland Nepal where our study was carried out, for example, mothers-in-law and their husbands act as ‘household guardians’ and maintain a strong hierarchy over younger household members including their sons. It is not culturally appropriate for daughters-in-law, particularly when newly married, to go outside the house except for specific reasons such as accessing healthcare, when they would be accompanied by the mother-in-law or husband [19]. O’Hara and Clement [33] described a set of entrenched social norms in Nepal that collectively maintain women’s subordinate position: that women should obey their husbands in all things; that women’s work should be limited to household chores; that education is not valuable for women; that women should leave politics to men; that there are times when women deserve to be beaten; and that a woman should tolerate violence in order to keep her family together. While individuals vary in the extent to which they hold these norms [33], the prevalence of violence against women is common in Nepal [34], especially in the lowland Terai region [35], and the fear of violence is likely to be one factor that acts to maintain women’s subservience. The curtailment of women’s independence can be seen as a form of exploitation of the daughter-in-law by members of her household that are genetically unrelated to her, while also minimizing the likelihood of paternity uncertainty.

From a different theoretical perspective, the role of the parents-in-law in family dynamics has been approached through the lens of the ‘patriarchal bargain’ [36]. According to this perspective, a young bride is obliged to accept a position subordinate to both men and senior women in the marital household. The daughter-in-law may serve the mother-in-law’s own interests, including her wellbeing, health and food security in old age. In the short term, the daughter-in-law’s labour and reproduction are subordinate to, and dependent on, the economic interests of male household members, with whom the mother-in-law is aligned, but in the long-term, the daughter-in-law may gain her own authority, eventually at the expense of her own daughter-in-law when her son marries [36]. Like evolutionary approaches, this framework predicts conflict between household members, whereby parents-in-law favour a subservient daughter-in-law who will not challenge their influence over their son.

Based on the evolutionary framework set out above, we therefore predict that the mental health of a woman living with her husband’s kin may be strongly shaped by whether household relations are more supportive, or more controlling and exploitative. We further predict that her mental health may depend on which relatives are co-resident. For example, the behaviour of parents-in-law might change if the husband has temporarily migrated away for work, and can neither contribute directly to childcare nor guarantee his wife’s fidelity.

To date, studies addressing the mental health of mothers in their own right do not tend to use an evolutionary perspective [37]. Nevertheless, research from South Asian countries indicates that mothers with difficult relationships with their parents-in-law may experience psychological distress [38–40], especially during the antenatal and postnatal periods [41–43]. In Nepal, studies have shown that women who were pressured into childbearing by their mothers-in-law, particularly to have a son, and those with a greater number of children, tend to be more distressed [44, 45]. For Maithili-speaking Madhesi women in Nepal with limited autonomy in household decision-making, psychological distress is also related to the pressure to be a dutiful wife and mother [38]. In Nepal and Pakistan, women living in extended families may also be exposed to intimate partner violence, which negatively affects their mental health [46–48].

Our study was conducted in a Maithili-speaking Madhesi society, where women are socially isolated, largely restricted to the husband’s home, and tend to be accompanied by a relative when venturing outside the home. From an evolutionary perspective, this isolation may be considered beneficial for the fitness of the husband and his genetic kin, both through maximizing the wife’s childcare and unpaid care work (reproductive investment), while also minimizing her opportunities for extra-pair mating. However, as recognized by Dyble et al. [26], supporting the mother could also bring fitness benefits to marital kin, through promoting maternal capital and increasing the survival and health of the offspring, and subsequently the quantity and phenotypic quality of grandchildren. These contrasting strategies are expected to impact both of the outcomes we investigated, maternal mental health and economic autonomy. Our study therefore explored how these outcomes varied by household composition.

We tested competing hypotheses, to identify whether specific types of relative were more likely to provide support or constraint to mothers. By using the co-residence of both husband and parents-in-law as the reference group, we could evaluate maternal outcomes when specific categories of relative were not co-resident, and thereby learn about the influence of these relatives on maternal outcomes. We found evidence for both of our overarching hypotheses: individuals who are genetically unrelated to mothers may provide both support and constraint, depending on the outcome and the category of relative.

When parents-in-law but not husbands were co-resident, mothers had similar mental health outcomes as the reference group. Conversely, when the husband was present but not the parents-in-law, maternal mental health was worse, with mothers reporting losing sleep, feeling unhappy and worthless, and greater overall distress. When neither category of relative was present, maternal outcomes did not differ from the reference group. Overall, this suggests that parents-in-law provided some support to mothers, thereby improving their mental health. Importantly, the fact that the contrast in the distress score was greatest between both parents-in-law and husband being co-resident (lowest score), versus husband only (highest), suggests that parents-in-law may actively have reduced negative impacts of the husband. These differences are of public health significance, as mothers were two times more likely to be distressed if they co-resided with only their husbands, compared to both husband and parents-in-law.

When parents-in-law but not husbands were co-resident, mothers had similarly low economic autonomy as those in the reference group. Figure 4 shows that when they were co-resident, parents-in-law were the primary economic decision-makers in the household, regardless of whether the husband was co-resident or not. Conversely, when the husband was present but not the parents-in-law, maternal autonomy increased, and paradoxically, this increment was further enhanced if the husband himself was not co-resident, having migrated for work. In the absence of parents-in-law, the mother was the primary economic decision-maker in the household. These patterns suggest that overall, parents-in-law inhibited maternal economic autonomy, whereas husbands promoted it. This can be attributed to women living without their parents-in-law being required to take more responsibility for the household, though husbands often continued to make economic decisions if they were overseas.

We interpret these two sets of results in combination as suggesting that mothers are exposed to trade-offs: different categories of marital kin balance the control and constraint of women with different forms of support. We suggest that wives in this population may be overall subjected to two forms of mate-guarding, that differ in how they relate to the household’s need for women to express economic autonomy. When co-resident, parents-in-law appear to constrain women’s autonomy, but having negated their agency, a form of mate-guarding, they also seem to provide some social support that benefitted their mental health. In contrast, husbands appear to promote the economic autonomy of their wife, and reduce her unpaid care work, but by undermining her sense of worth and mental health, husbands may also operationalize an alternative tactic for mate-guarding. We found some support for this second hypothesis, as when parents-in-law were not co-resident, mental health outcomes were better for women whose husband was also not co-resident, compared with those where he was. As the majority of mothers in our sample were either completely uneducated or had only primary levels of education, the co-residence status of PILs and H, rather than her education, are likely to be the key factors associated with maternal mental health and economic autonomy.

In this study, we focused on continuous markers of mental health and autonomy that can be measured across the whole population. However, other work has focused on more extreme markers of the costs inflicted to women, such as violence by one or both of husband and parents-in-law [34, 35, 64, 65]. Women in this society are therefore likely to be familiar with the threat of violence, reducing their incentive to resist familial constraints on their autonomy.

Our findings are consistent with other studies noting mental health benefits of social support [39, 44]. Regarding autonomy, other studies have also found parents-in-law to be the most frequent household economic decision-makers [19] but also that non-resident husbands may continue to make decisions [39]. Other studies have also found that parents-in-law reduce mothers’ income earning opportunities [44], which is of interest given our prediction that parents-in-law would act in the reproductive interests of their son.

Our measure of unpaid care work included support for elder members of the household. This direction of support is also given emphasis through the lens of the patriarchal bargain framework [36], which considers that the interests of maintaining senior household member’s health, autonomy and food security shape the experience of younger members. However, Gram et al. also found evidence of husbands pushing back against parents-in-law influence, by amassing enough resources to eventually form a nuclear household with their wives [53]. This may relate both to the husband’s own reproductive interests, but also to his interest in maintaining a good quality relationship (pair-bonding) with his wife [66].

In a society where decisions are typically made collectively, and where the opinion of older family members reflects their greater experience and authority, women lacking co-resident family members may find the pressure to run the household and taking on decision-making responsibility stressful, as reported in other studies [67]. Consequently, a lack of economic autonomy may not be perceived by some mothers as a sign of disempowerment, as they may not want the responsibility of making decisions, which in itself may cause distress, especially in the absence of their husbands [68]. Efforts to avoid stressful situations may be a covert form of agency or empowerment in this population [50]. While acknowledging the diversity of definitions of empowerment [50], we use it as a broader concept reflecting women’s ability to overcome barriers and achieve their goals and desired outcomes.

In the context of our study, households are primarily patrilocal and patriarchal. Future studies could test our competing evolutionary hypotheses more broadly, by comparing patrilocally and matrilocally married women, with and without husbands and either parents-in-law or parents. For example, in matrilineal societies, a mother’s own mother may help her to meet high costs of reproductive investment, if demanded by her husband. Whether families are characterized by cooperation or conflict may also be elucidated by considering factors such as maternal age, income earning opportunities outside the home, parity and family members beyond parents-in-law, parents and husbands, all of which may shape maternal mental health and economic autonomy. For example, our recent fieldwork in this cohort will explore the implications for mental health of women becoming increasingly related to other group members as they age, by transitioning to the role of mother-in-law [69].

Overall, our results indicate that mothers face trade-offs regarding their household activities, and may experience both mental health benefits and costs through co-residence with other family members, alongside variable levels of economic autonomy [26]. As recently highlighted [26], a person related by marriage (affine) may gain fitness benefits by providing support to an unrelated individual, who will carry their genes via reproduction with a relative of the affine. Both husbands and parents-in-law could gain such fitness pay-offs, by supporting mothers and reducing conflict and distress, with positive effects on the quality of her offspring [24]. However, marital kin could also increase their fitness by reducing paternity uncertainty, and we suggest that this strategy is evident in our study population, as indicated by very restricted physical mobility and social contact outside the family.

Our interest here is in maternal mental health in its own right, irrespective of its implications for Darwinian fitness. Our findings suggest that parents-in-law offer women greater support than do their husbands, which may be possible because they also restrict their economic autonomy. When parent-sin-law are not co-resident, husbands may use more controlling tactics (psychosocial pressure, potentially leading to distress), to counterbalance their requirement for greater economic autonomy by their wives. Future work could consider the mental health of the parents-in-law and husbands from the same perspective. Overall, we favour a nuanced framework that acknowledges negotiation and nurturing among genetically unrelated family members, as well as hierarchical power dynamics.