Introduction

Consanguineous marriage and its effect remained an interest to the geneticists, medical and social scientists because of its effect on pregnancy outcomes and child mortality1,2,3,4.Several studies reported consanguineous marriage increases the risk of spontaneous abortion, stillbirths, neonatal, and child mortality1,5,6,7. For example, in a Palestinian based study, Assaf et al. (2009) found consanguineous marriage to be a significant risk factor for intrauterine life of foetus with age and parity as socioeconomic confounders5. In human populations, it is difficult to measure the direct effect of consanguinity on pregnancy outcome since several exogenous (i.e., socioeconomic, environmental, availability and utilization of health care services), and endogenous factors (i.e., inbreeding, birth injury, etc.) play a deterministic role8. Earlier studies emphasised on the improvement in both endogenous and exogenous factors to reduce pregnancy wastage1,8,9.

In a multicultural country like India, consanguinity is one of the preferential types of marriage practised by some communities. North–south regional dichotomies in culture largely determine the prevalence of consanguineous marriages10. Consanguineous marriage is customary in Southern India, while arranged marriage or exogamy is highly prevalent in North India11,12. Analysis of pan-Indian data reveals that consanguinity causes varying degrees of offspring mortality and congenital malformations across the different regions1,3,13. The incidence of congenital malformations is higher in consanguineous marriage, especially in case of uncle-niece marriage (f = 0.1250), where the degree of relatedness is higher compared to that of first cousin marriages (f = 0.0625)13. In relation to mortality, it was observed thatin south Indian state of Karnataka2, and the north Indian state of Jammu6, the likelihood of miscarriage and stillbirth due to consanguinity was higher compared to non-consanguinity.

In the post-globalization era, particularly since the beginning of the new millennium, India witnessed a significant change in the demographic and socio-economic profile14,15, along with the substantial increase in the utilisation of health care services, particularly maternal and child health13,14. In general, these developments led to an increased life expectancy and decline in infant and child mortality and morbidity rates15.

Earlier, Mosley and Chen (1984) used an analytical framework for the study of child survival in developing countries using several proximate determinants (i.e., maternal factors, environmental contamination, nutrient deficiency, injury and personal illness), and socio-economic variables (i.e., individual level variable, household level variables, community level variables)9 But, the complex web of cultural proscriptions and prescriptions, socio-cultural belief and practice attached to pregnancy and childbirth continues to prevail16,17. It is true that the proposed framework for child survival in developing countries by Mosley and Chen (1984) was not exclusively in the context of consanguineous marriage9). But we used this framework to control the endogenous, i.e., inbreeding or congenital, and exogenous, i.e., demographic, health care-seeking and socioeconomic variables as confounders to get a precise estimate of the association between consanguinity and adverse pregnancy outcome and offspring mortality in developing countries like India.

In India, there has been a significant improvement in accessibility to and availability of public health services during the last couple of decades; but social inequality still remain a constrain in getting uniform benefits of public health system18,19. India has also made significant progress in life expectancy, decline of infant and under-five mortality rates, and prevalence of communicable diseases18. Despite all these development in the public health, consanguinity still contributes to high rates of congenital malformations and child mortality in India. (1,3,13). India still requires a transformation in the public health system by involving people in remote areas, addressing social practices such as consanguinity, and utilizing modern technologies18,19.India’s population varies in terms of the benefits of public health accessibility, mainly due to social inequality18,19. India has made significant progress in life expectancy, infant and under-five mortality rates, and disease incidence. However, social practices of consanguinity still contribute to higher rates of congenital malformations and child mortality in developing nations like India1,3,13.

The present study hypothesized that consanguineous marriage will remain a risk factor for pregnancy outcome and offspring mortality, but the exogenous factors such as development in demographic, socioeconomic conditions and increased utilization of maternal and child health care services would work as a buffer in the reduction of child mortality rates. The analytical approach involves ‘stepwise’ incorporation of variables-consanguinity, demographic, maternal healthcare-seeking and socioeconomic factors to have a comprehensive understanding of the effect of consanguineous marriage on pregnancy outcome and offspring mortality in India.

Data and methods

Data

Data for this study were pooled together from the fourth and fifth round of National Family Health Surveys (NFHS-4 and NFHS-5), which were carried out by the International Institute for Population Sciences (IIPS), Mumbai and International Classification of Functioning, Disability, and Health (ICF) during 2015–16 and 2019–21. NFHS is an Indian variant of Demographic and Health Surveys (DHS) which collects and disseminates information on socio-demographic, economic and cultural characteristics, and maternal and child health, and nutritional status based on a nationally representative household survey. Requisite permission in accessing and usage of the dataset was obtained from the DHS Program archive (www.dhsprogram.com). The survey collected information from 601,509 households in NFHS-4 and 636,699 in NFHS-5 and interviewed 699,686 women in NFHS-4 and 724,115 in NFHS-5 (in the age group 15–49) in all the states and union territories in India. Details on the sampling design, coverage, and findings of the survey are available in the national report16. A total of 493, 927 women in NFHS-4 and 514,927 in NFHS-5 were included in the analysis after excluding the 349,881 (169,814 in NFHS-4 and 180,067 in NFHS-5) women who were never married, 34,317 (18,718 in NFHS-4 and 15,599 NFHS-5) women who were not the usual residents of the households, 14,605 (10,183 in NFHS-4, and 4422 in NFHS-5)women who reported more than one marriage, and 16,774 (7674 in NFHS-4 , and 9100 in NFHS-5) women who belonged to the union territories of India.

The survey included two questions pertaining to consanguineous marriage. First, whether, the respondent was biologically related to her husband before marriage, and, second, if answered in the affirmative, the exact nature of the relationship. In the latter question, the respondents were supposed to report whether their spouse was an uncle, first cousin, a second cousin, or ‘other biological relative’. In order to understand the effect of consanguinity on offspring mortality precisely, the whole consanguineous group was further divided into two separate categories, close consanguinity (CC) and distant consanguinity (DC) based on the relative distance between husband and wife. In this regard, first cousin on paternal side and first cousin on maternal side have been included in the CC section, because the genetic impact of these marriages is reportedly more serious than that of DC marriage1. DC marriages included second cousin, uncle-niece and the marriages between ‘other blood relatives’ as mentioned in the dataset.

Additionally, all eligible women were asked to provide a complete birth history, comprising date of birth, survival status and age of child at death. The study was restricted to the births in the five years proceeding both the survey periods, i.e., NFHS-4 (2015–2016) and NFHS-5 (2019–2021). A total of 4,902 (2616 in NFHS-4, and 2286 in NFHS-5) stillbirths and 51,843 (26,494 in NFHS-4, and 25,349 in NFHS-5) spontaneous abortions were reported. Out of 259,627 live-births in NFHS-4, and 232, 920 live-births in NFHS-5 that occurred during the five years preceding the survey, 240,152 births in NFHS-4 and 217,520 births in NFHS-5 respectively were included in the analysis after excluding the 565 births (297 in NFHS-4 and 268 in NFHS-5) from women who were not married, 24,806 births (13,753 in NFHS-4 and 11,053 in NFHS-5) from women who are not the usual residents of the households, 3770 births (2592 in NFHS-4 and 1,178 in NFHS-5) from women reporting more than one marriage, and 5734 births (2833 in NFHS-4 and 2901 in NFHS-5) from women who belong to the union territories of India. Out of 457, 672 births, 12,388 (7,112 in NFHS-4 and 5276 in NFHS-5) died during the neonatal period (before completion of one month of birth), while 4888 (2752 in NFHS-4 and 2136 in NFHS-5) and 1839 (1,100 in NFHS-4 and 729 in NFHS-5) died during the post-neonatal (between 1 and 11 months) and childhood periods (between 12 and 59 months) respectively. The variables used in the study were described in Table 1.

Table 1 Description of variables included in the study.
Full size table

Methods

To assess the effect of consanguineous marriages on adverse pregnancy outcomes and offspring mortality, first the regional variations in the prevalence of consanguinity and types of consanguineous marriages (CC and DC) were estimated and compared with non-consanguineous (NC) marriages. Following Mosley and Chen (1984), the background characteristics included in the study were age at marriage (below 18, 18–24, and 25 and above), educational attainment of woman/mother (no education, primary, secondary, higher), types of family (nuclear, non-nuclear), religion (Hindu, Muslim, Others), caste affiliation (scheduled castes (SC), scheduled tribes (ST), other backward castes (OBC), and others), household wealth quintiles (poorest, poorer, middle, richer, richest), place of residence (urban, rural), and region of residence (north, east, central, north-east, west, south)9. Additionally, indicators of maternal care-seeking variables, namely, availed at least 4 ANC (no, yes), obtained institutional delivery (no, yes), and received PNC in 48 h of delivery (no, yes) were incorporated in the analyses of neonatal, post-neonatal and child mortality.

‘Scheduled castes’ and ‘scheduled tribes’ are castes and tribes identified by the Government of India as socially and economically backward and in need of special protection from social injustice and exploitation according to the article numbers 341 and 342 of the Constitution of India. ‘Backward caste’ is the term used by the Government of India to classify groups which are educationally or socially disadvantaged yet upper in the ladder compared to SC/ST in the caste hierarchy. The variable was classified as scheduled castes (SC), scheduled tribe (ST), other backward castes (OBC) and others. Household wealth index was used as a proxy for household economic status. The wealth index was computed from economic proxies such as household durable assets, availability of safe drinking water and sanitation, and landholding. For construction of wealth index, these variables were first divided into sets of dichotomous variables and indicator weights were assigned using principal component analyses (PCA). The index thus derived was divided into five quintiles: poorest, poorer, middle, richer and richest20.

As the response variable such as spontaneous abortion, stillbirth, and neonatal mortality are binary variables; binary logit regression models were carried out in all cases, though model characteristics vary according to the nature of outcome. It may be mentioned that in the case of spontaneous abortion and stillbirth, the unit of analysis was ever married women aged 15–49. To note, spontaneous abortion and stillbirth have been pooled together in the regression analyses for two reasons: first, occurrence of both of these events can be categorized as foetal wastage, and secondly, number of cases of these events are rare and potentially destabilised model estimates, if analysed separately. In the case of neonatal mortality, the unit of analysis was living children aged at least one month and children who died within one month of their birth.

In the binary logistic regression analysis, the dependent variables were recorded as 0 if no, and 1 if yes.

$$\log_{e} [P(Y_{i} \, = \,1| \, X_{i} )/1 - \, P(Y_{i} \, = \,1| \, X_{i} )]\, = \,\, = \,\alpha \, + \,\beta_{1} Xi_{1} , \ldots \ldots \ldots .. \, \beta_{k} X_{ik}$$
(1)

Where, Yi is the binary response variable and Xi is the set of explanatory variables like socio-demographic characteristics, and β1 β2…… βk are the coefficient of the Xi variables. The results are given in the conventional form of adjusted odds ratios (AOR).

In the case of post-neonatal mortality, the population comprised of all children aged between 1 and 11 months. If they have died before attaining one year, they were given a value of 0. If they are alive at the time of survey, they may survive to attain one year, or may die before the completion of one year. We are unable to observe the event of death because of the truncated nature of data. Similarly, child mortality comprises the population of all children with age 12 months to 59 months. The aforementioned problem continues to persist in this case. In such cases, survival analysis, with a logit regression model, is relevant21, indicating likelihood of dying between 1–11 months and 12–59 months as dichotomous outcome. Specifically, a proportional hazards model analysis22 has been used in this case. The general form of the Cox proportional hazard model is expressed as follows:

$$h\left( {t,X} \right) \, = \, h_{0} \left( t \right)e^{\Sigma \beta ixi}$$
(2)

where h0 denotes the baseline hazard function, which refers to the duration to mortality; the vector X (x1, x2, …xk) denotes the set of predictor variables and βi are the coefficients of these variables. An important assumption, implicit in the structure of the equation, is that the hazards at time t is modified by the characteristics of a unit (x1, x2,…xk) with eΣβixi as multiplier. It is assumed that this multiplier does not vary over time, hence the term ‘proportional’ hazards. The results of hazards model analysis are given in the conventional format of adjusted hazard ratios (AHR). The predictor variables may be continuous or in a categorical form. In the latter case, which is the present situation, one category is designated as ‘reference category’. The coefficients for the other categories then give the effects on hazards relative to that of the reference category (here, death of the child).

A stepwise systematic model-building procedure was adopted. For predicting spontaneous abortion and stillbirth, consanguinity type was included in Model I, while demographic variables were added in Model II and variables related to socioeconomic status, and place and region of residence were incorporated in Model III successively. For predicting neonatal, post-neonatal and child mortality, consanguinity type was included in Model I, whereas demographic variables, maternal health care-seeking variables, and variables related to socioeconomic status, and place and region of residence were incorporated in Model II, III and IV respectively.

Ethical approval

The research was carried out by using publicly available database. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees in human experimentation and with the Helsinki Declaration of 1978, as revised in 2008.

Results

The overall prevalence of consanguinity is 13.6%; the prevalence of CC (8.2%) was nearly three percentage point higher than that of DC (5.3%). CC was found to be higher in southern India (19.7%), while that of DC in northeast India (6.8%) (Table 2).

Table 2 Distribution of ever married women age 15–49 years, by region, according to their marriage pattern, India, 2015–21.
Full size table

The prevalence of CC was higher among the respondents who were married before 18 years of age (8.8%), educated up to secondary level (8.9%) and exposed to media (9.0%). The prevalence of CC was higher among the Muslims (12.6%) and OBC social group (9.3%), and those living in middle wealth quintile (10.3%), nuclear families (8.4%) and urban areas (8.9%) and the southern region of this country (19.7%) (Table 3).

Table 3 Sample characteristics of Mothers aged 15–49, 2015–21.
Full size table

The male neonatal mortality was higher in DC, while female post-neonatal and child mortality were higher in DC and CC respectively. Mortality rate of children (from neonatal to child) was higher among both CC and DC for those with birth order above four, and born below 24 months birth interval. Spontaneous abortion (8.9%) and stillbirth (5.9%) were high in DC who got married between 18 and 24 years; while neonatal mortality (35.7%) was high in CC and those got married beyond 25 years of age; post-neonatal (14.1%) and child mortality (7.7%) were high in CC and who got married below the age of 18 years. Spontaneous abortion increased with the increase in the level of education for CC and DC; but DC showed higher values than CC. Thetrends were reverse for both CC and DC with respect to the relationships between education and prevalence and still birth, neonatal, post-neonatal and child mortality. Non-working CC and DC respondents had higher number of spontaneous abortion and stillbirth than the working respondents of this category; but DC showed higher values in both the cases than the CC. For neo-natal mortality and beyond, the trends were reverse in terms of prevalence and working status (Table 4).

Table 4 Stillbirths, neonatal, post-neonatal and child mortality by marriage type and selected background characteristics, India 2015–21.
Full size table

The prevalence of all the types of mortality was higher for those not exposed to mass media, irrespective of the consanguinity types, with lower prevalence in DC compared to CC. The prevalence of all post-birth mortality was higher irrespective of the consanguineous types for those who have not visited ANC (≥ 4); but DC showed higher prevalence than CC for neonatal and child mortality. The trend in prevalence was similar for PNC visit (within 2 days); but reverse was the trend when compared for the prevalence in CC and DC. The post-neonatal and child mortality rates were higher in CC compared to DC for those who went for institutional delivery (Table 4).

Among the Hindus, the prevalence of all types of mortality (barring child mortality) was higher in DC than CC. Among the Muslims, the trend was reverse (barring spontaneous abortion). For variables related to ‘caste’, the prevalence of most of the mortality types was higher in DC compared to CC (Table 4).

Stillbirth, neonatal, post-neonatal and child mortality rates were higher among the respondents belonging to the poorest wealth quintile, irrespective of CC and DC. Marrying to distant cousins (DC) resulted in higher rates of spontaneous abortion and stillbirth among those living in non-nuclear family, whereas neonatal and post-neonatal mortality rates were higher in DC and those who live in nuclear families. Central region of the country showed slightly higher rates of mortality, irrespective of consanguineous types (Table 4).

Results of multivariate logistic analysis

Consanguinity and spontaneous abortion and stillbirth

Marrying to close cousins (CC) and distant cousins (DC) enhanced the likelihood of spontaneous abortion and stillbirth compared to NC (CC, AOR = 1.11, p < 0.001 and DC, AOR = 1.55, p < 0.001 in Model I). The value of AOR for CC and DC increased significantly in Model III after the incorporation of demographic and socioeconomic variables compared to NC (CC, AOR = 1.32, p < 0.001 and DC, AOR = 1.62,p < 0.001 in Model III). The value of AOR remains higher for DC than CC for each Model (Table 5).

Table 5 Adjusted odds ratio (AOR) estimated from binary logit regression of spontaneous abortion and stillbirth by background characteristics, India 2015–21.
Full size table

Respondents, who married between 18 and 24 years and 25 years and above, showed significantly more risk of spontaneous abortion and stillbirth compared to those who got married before 18 years. The level of educational attainment of the respondents enhanced the likelihood of spontaneous abortion and stillbirth. For working women, the likelihood of spontaneous abortion and stillbirth was less compared to those who were non-working. The likelihood of spontaneous abortion and stillbirth was significantly higher among the Muslims and lower in other religious groups compared to the Hindus. Caste groups, ST, OBC and ‘other’ social groups’ showed lower risk of spontaneous abortion and stillbirth compared to the SC group. The risk of spontaneous abortion and stillbirth was significantly lower for all the wealth quintile categories starting from poorer to the richest and those living in nuclear families compared to those belonging to poorest quintile and living in non-nuclear families respectively. Results showed that living in urban areas significantly increased the likelihood of spontaneous abortion and stillbirth compared to the respondents belonging to rural areas. Barring the central region of India, all the other regions showed lower likelihood of spontaneous abortion and stillbirth compared to the northern region (Table 5).

Consanguinity and neonatal mortality

The neonatal mortality was significantly more likely in both CC (AOR = 1.10, p < 0.001) and DC (AOR = 1.16, p < 0.001) compared to NC independently (Model I). The odds of consanguinity types in predicting neonatal mortality increased gradually from Model I through Model III and remained significant. In the final Model (Model IV), after incorporating demographic, health care seeking variables and socioeconomic variables, the risk of neonatal mortality became insignificant for CC, unlike DC (1.30, p < 0.01) (Table 6).

Table 6 Adjusted odds ratio (AOR) estimated from binary logit regression for neonatal mortality by background characteristics, India 2015–21.
Full size table

The table also revealed that the likelihood of neonatal mortality among females was consistently lower in all the models compared to males. The risk of dying during neonatal period was significantly higher for the children of higher birth order (4 or more) with less than 24 months birth interval compared to the first birth order in Models II and III; however, such risk declined after inclusion of demographic and socioeconomic variables (Model IV). The risk of neonatal mortality was significantly higher for those respondents who got married at the age of 25 years or later compared to those who got married before 18 years. More than four ANC visits, higher level of education and belonging to the richest wealth quintile category significantly reduce the likelihood of neonatal mortality. Working status of the respondents and exposure to media significantly increased the risk of neonatal deaths. The risk of neonatal mortality was significantly lower among the children of ST’s and ‘other’ social groups (comprised of non-SC/ST/OBC) compared to the children of SC group. Neonates belonging to the southern region of the country showed significantly lower risk of death compared to the children belonging to the northern region, whereas neonates of the central region showed higher risk of death.

Consanguinity and post-neonatal mortality

The adjusted hazard ratios (AHR) were estimated from Cox proportional hazard regression models of post-neonatal mortality by consanguinity status. Post-neonatal mortality was significantly higher irrespective of the consanguineous types in all the Models. The likelihood of post-neonatal mortality was higher in both CC (AHR = 1.17, p < 0.001) and DC (AHR = 1.18, p < 0.001) compared to NC (Model I). The risk of post-neonatal mortality was significantly higher for both CC and DC in Models II and III compared to NC after controlling the socio-demographic and maternal health care-seeking variables as confounders in the respective models (AHR = 1.16, p < 0.001 for CC and AHR = 1.20, p < 0.001 for DC in Model II; AHR = 1.25, p < 0.001 for CC and AHR = 1.33, p < 0.001 for DC in Model III). The likelihood of post-neonatal death significantly increased for CC (AHR = 1.49, p < 0.01) and also for DC (AHR = 1.87, p < 0.001) compared to NC after controlling the socio-demographic, health care seeking and socio-demographic variables as confounders in Model IV (Table 7).

Table 7 Adjusted hazard ratios (AHR) estimated from Cox proportional hazard regressions for post- neonatal mortality by background characteristics, India, 2015–21.
Full size table

The likelihood of post-neonatal mortality for female child was significantly higher compared to males after controlling confounders in Model IV. Children with birth order 2–3 and those born more than 24 months birth interval were significantly at a lower risk of post-neonatal mortality compared to the children of first birth order; while such risk was significantly higher among children belonging to rest of the birth orders and birth intervals. The likelihood of post-neonatal mortality was significantly higher among the respondents who got married between 18 to 24 years compared to those got married below 18 years in Model IV. The same table revealed that four or more ANC check-ups, higher level of education of mothers and those belonging to richest wealth quintile significantly reduced the risk of post-neonatal mortality. Exposure to mass media and belonging to ‘other’ religious group (other than Hindu and Muslim) significantly increased the risk of post-neonatal mortality. Post-neonatal mortality was significantly higher in the northern region, compared to the central regions.

Consanguinity and child mortality

Close consanguinity (CC) significantly increased the risk of child mortality (Model I) independently (AHR = 1.35; p < 0.001) and remained as risk factor even after the inclusion of demographic variables in Model II (AHR = 1.37; p < 0.001); however such effect becomes non-significant after controlling the demographic, maternal health care-seeking and socioeconomic confounders in Model IV (Table 8).

Table 8 Adjusted hazard ratios (AHR) estimated from Cox proportional hazard regressions for child mortality by background characteristics, India, 2015–21.
Full size table

The likelihood of child mortality was consistently higher and significant for children born in four or more birth order, at less than 24 months birth interval compared to those of the first birth order in all the models. Higher age at marriage, institutional delivery and higher educational attainment significantly reduced the risk of child mortality. Respondents of the central region showed lower risk of child mortality compared to the North region.

Discussion

We hypothesized in this study that the development in demographic, socioeconomic conditions and increased utilization of maternal and child health care services (exogenous factors) during post globalization era would reduce the risk factor for pregnancy outcome and offspring mortality owing to consanguinity. We used Indian national level data of the last two decades in this study. The trend in the result suggests no significant association between consanguinity and child mortality, suggesting exogenous factors reduces the risk of child death.

As stated earlier, the demographic, socioeconomic and health infrastructure facilities in India have significantly improved during the last few decades, following globalization14,18,19. The Government of India has also taken initiatives like Indian Newborn Action Plan to reduce newborn deaths23. Thus, it is hypothesized in this study that consanguineous marriage will remain a risk factor for pregnancy outcome and offspring mortality, but the development in demographic, socioeconomic conditions and increased utilization of maternal and child health care services would work as a buffer in the reduction of child mortality rates.

Using pooled data from NFHS-4 and NFHS-5, this study corroborates with earlier studies that identified consanguinity posing a significant challenge on pregnancy outcomes and offspring mortality, except child mortality1,3. Although the framework for child survival in developing countries proposed by Mosley and Chen (1984) was not exclusively for measuring the pregnancy outcomeof arising from consanguineous marriage, yet we used this framework to control the demographic, health care-seeking and socioeconomic variables as confounders to precisely examining the association of consanguinity with adverse pregnancy outcome and offspring mortality9. The present study conforms to the findings of previous studies carried out using pan-Indian datasets1,3 to a certain extent. The analytical framework used in these earlier studies was a ‘two step’ model1. In this study, we included the confounder variables in three and four successive steps (Models) for spontaneous abortion and stillbirth and post-birth mortality respectively. For example, the odds of spontaneous abortion/stillbirth and post-neonatal mortality increased due to consanguinity after the inclusion of socio-economic characteristics of the participants in the final model. But reverse trends were observed for neonatal and child mortality; here inclusion of socio-economic and health care variables reduces the odds of neonatal and child mortality owing to consanguinity. In both the cases, socioeconomic variables and health care facilities remain pronounced in delineating mortality due to consanguinity, but in contrasting ways. Unlike, Grant and Bittles (1997), Fareed et al. (2017), the finding of this study supports that development in healthcare seeking and socioeconomic variables significantly lowered the risk of consanguinity on child mortality6,24.

We argue that perhaps the family members show more concern in availing public health care facilities for the survival of a child, rather than for the ones who are yet to born. For example, most of the women of this country are reluctant to take up intra uterine check-ups periodically, mostly due to the discouragement from the in-laws17. This results in adverse intrauterine health of the baby including infections, placental abnormalities, fetal and maternal haemorrhage25. The risk of adverse intra uterine health increases in varying degree with increase in marriage with close relatives1,3,24,26,27.

In India, low levels of education, lower socioeconomic status, and religious beliefs increase the prevalence of consanguineous marriage10 and often limit the use of medical and health care facilities28. We observed an increase in spontaneous abortion rates among women with higher education levels. But, the authors cannot rule out the possibility of underreporting of cases of spontaneous abortion from those with lower level of education. Exogenous factors like environmental, medical factors, religious beliefs, lack of awareness among the couples and their family members regarding child birth8,25,26,28 significantly affect adverse pregnancy outcome. In Islam, it is believed that the spirit of the foetus resides in heaven, under the care of the Almighty, and prays for the betterment of their parents29. This type of worldview of the family members towards pregnancy and health of the child could be a reason for high intra-uterine mortality outcomes, especially among the Muslims. The family members of the pregnant women often do not want to disclose the news of pregnancy to their peers fearing the effect of evil eye (Nujur) which may bring adverse pregnancy outcome30,31. Thus, it becomes difficult for the health service providers to closely monitor the intra-uterine development of the fetus and diagnose abnormality, if any, which might lead to spontaneous abortion or stillbirth. Cultural beliefs about ritualistic disease prevention practices, choice of therapies and practitioners for sickness care, abstinence to prevent illness in the suckling child often impact the survival of the offspring30,32,33. Based on the trend of the findings, it may be argued that the genetic adversity owing to consanguinity may pose challenges to pre-birth (intra-uterine) life, in terms of pregnancy wastage since it is more guided by cultural prescriptions and proscriptions than public health intervention.

Studies argue that premarital genetic counselling can be used as an important tool in the increase of awareness and selecting marital partners in improving pregnancy outcomes and to meet psychosocial and cultural challenges34,35. In countries like Jordan36, Saudi Arabia37 and Iran38, having high rates of consanguinity and inherited disorders, community programmes for premarital genetic counselling (e.g., thalassemia and sickle cell anaemia) are in progress. India needs to introduce adequate facilities for genetic counselling and initiate aggressive campaigning to avail this facility in order to reduce adverse pregnancy outcome resulting from consanguinity.

In general, alloparenting (support from extended kin and community members) has been a widespread phenomenon in human societies39. The degree of alloparenting varies with factors like social structure, wealth and mode of subsistence31,39. Adverse neonatal and post-neonatal birth outcome can be reduced with increased level of child care from the extended kin and availing of health care facilities. It was observed that the pregnancy outcomes of DC were likely to have an enhanced risk of adverse post-birth outcome. The authors do not have immediate biological explanation for such results. But, it can be argued that support from immediate kin members in rearing and caring of new born is lesser in case of couples who practised distant consanguinity (DC) than those who practised close consanguinity (CC). Earlier studies have found that caring and support from family members is one of the major reasons behind the practice of consanguinity31,39. For example, Kalam (2021) found that neonates receive better care from the relatives of close cousin couples29. In-laws of distant consanguine (DC) couples are reluctant to extend support during crucial phases, and parents of women rarely visit them due to social prestige29,40.

The Accredited Social Health Activists (ASHA) workers in India receive government funding to provide medical aid and support for neonatal care23. So, during the post-neonatal period the door step care for the new born from these public health service providers becomes reduced, thereby increasing the risk of post-neonatal mortality.

Conclusion

It appears from the study that the endogenous effect1,8 of consanguinity still pose a serious challenge to the survival of fetus and new born; but the exogenous effect, i.e., the development in socio-demographic, maternal and child health-care-utilization and economic conditions in India and alloparenting reduces the risk of death from child. The study has certain limitations. Since the findings of the study are based on survey data, collected by field investigators, there might be inter-observer bias. The survey data do not provide information on lineages, congenital abnormalities arising out of consanguinity.

Based on this study, the authors propose to incorporate socially entrenched practice of consanguinity explicitly into Mosley and Chen’s (1984) framework for the aid in understanding child survival in developing countries9. Policies focusing only on the improvement in the provision and utilization of maternal and child healthcare, and socioeconomic condition of the people may not be sufficient to reduce pregnancy wastage owing to consanguinity. It is equally important to develop policies to emancipate people regarding the detrimental effect consanguineous marriage.

Future studies could explore the potential mechanisms underlying the observed associations, such as socioeconomic factors, to provide a precise understanding of the pathways linking consanguinity to pregnancy outcomes and offspring mortality. Future studies can also pool multiple datasets to identify vulnerable population sub-groups who are at higher risk of adverse outcomes associated with consanguinity. Small-scale qualitative investigation would also be appreciated to understand further nuances in relation to consanguinity, pregnancy wastage and offspring mortality. This would help in designing targeted intervention through appropriate policy recommendations. (Supplementary Table S1).