Did Covid-19 pandemic increase incidence of low-birth-weight among newborns?

Globally, one in four babies are born with low birth weight, with long-term impacts on health and economic outcomes. This article uses Indian data from 2019-2021 to analyse the impact of the Covid-19 pandemic on birth outcomes, given the disruption of essential maternal and neonatal services. It finds that infants born post-pandemic had 45 grams lower birth weight, with 3% greater incidence of low birth weight than infants born before the pandemic.

The Covid-19 pandemic significantly disrupted the delivery of essential maternal and neonatal health services worldwide. Lockdowns, travel restrictions, and fear of infection hindered many women from accessing prenatal care, delivery services, and postpartum care. Consequently, these disruptions heightened the risk of maternal and neonatal morbidity and mortality. Additionally, the increased risk of maternal morbidity and amplified maternal stress may have adversely affected birth weight, a crucial determinant of infant health and survival.

Globally, approximately one in every four babies, or 30 million infants per year, are born with low birth weight (LBW), which is defined as a birth weight of less than 2,500 grams. About 95% of these LBW births occur in low- and middle-income countries. In South Asia, up to 27% of all births are LBW, and the region accounts for about 50% of all LBW babies worldwide. The recent National Family Health Survey (NFHS-5) conducted in 2019-2021 reported an LBW incidence of 18% in India (International Institute for Population Sciences (IIPS) and ICF, 2021).

Such a high incidence of LBW in India is a major public health challenge for policymakers due to the adverse impacts of poor health at birth on long-term health and economic outcomes. LBW infants have an increased risk of morbidity and mortality, and this has long-term implications for human capital development. LBW babies are more likely to experience developmental delays, cognitive impairments, and behavioural problems, affecting their ability to learn and succeed in school. This results in lower human capital accumulation, which adversely impacts their future employment opportunities, labour force participation, and wages (Currie and Vogl 2013, Kumar et al. 2022).

Given the nuanced impacts of Covid-19 lockdowns on access to antenatal care, maternal nutrition, and maternal stress, we study the association between the pandemic and LBW infants in India. Specifically, we compared the birth weight of infants born before the pandemic (July 2014-December 2019) and after the pandemic (April 2020-April 2021).

Data and methodology

We use micro-data from NFHS-5. The survey was conducted in two phases: phase I from 17 June 2019 to 30 January 2020 and phase II from 2 January 2020 to 30 April 2021. The sample included 636,699 households, comprising 724,115 women and 101,839 men, drawn from 707 districts in 36 states and union territories of the country (IIPS and ICF, 2021). The two phases of data collection in the pre-pandemic and pandemic periods are well-suited for estimating the effects of the Covid-19 pandemic on birth outcomes. The primary outcomes included in the analysis are birth weight (measured in grams) and a binary indicator of LBW – that is, whether or not the birth weight was below the threshold of 2,500 grams.

The survey provided pregnancy details for the mothers of 232,920 children. We excluded observations with missing values for birth weight and other relevant variables, resulting in a final sample of 204,615 children for our study.

We employ ‘multivariate linear regression’ to analyse birth weight and ‘multivariate logistic regression’ for the low-birth-weight model. For the logistic models, we calculate the odds of LBW for children born during the pandemic compared to those born before the pandemic. We report regression coefficients from the linear regression models and odds ratios from the logistic models.

To minimise biases from potential confounding variables, we include several factors in our empirical analysis: the child's gender and birth order, household caste, an indicator for being Hindu, rural residence, mother's age and education, and a household wealth index. We adjust for district-specific time-invariant characteristics, such as health infrastructure and development level, that may affect birth outcomes.

Main findings

We find statistically significant differences in birth weight and the prevalence of LBW between infants born before and after the pandemic. Infants born after the pandemic had 45 grams lower birth weight and a 3% greater probability of being born with LBW.

After adjusting for confounding variables and district-specific, time-invariant characteristics, the regression-based results show that births during the pandemic are associated with an 11 gram decrease in birth weight. The LBW results from the logistic regression model show that compared with infants born before the pandemic, infants born during the pandemic period had 1.08 times greater odds of LBW. Children born to educated mothers and in wealthier households had higher birth weight compared to reference groups.¹

Further, we observe significant variability in the associations between LBW and household and maternal characteristics. Among the nine subsamples analysed, the association is statistically significant in eight, except for the middle wealth groups (Figure 1). Compared to the subsample with secondary maternal education (schooling of more than five years), pandemic cohorts born to mothers with primary or less than primary education (schooling of less than or equal to five years) have slightly higher odds of LBW. Specifically, in the less-than-primary-schooling sample, the adjusted odds ratio for LBW is higher (1.09 versus 1.06). The analysis by household caste shows that the effect of the Covid-19 pandemic on LBW probability is higher for children from scheduled caste/scheduled tribe (SC/ST) households, which are historically disadvantaged communities, than the children from other castes. The analysis by religion shows that the pandemic had adverse impacts on LBW for Hindu as well as non-Hindu households, but the effect size is smaller for Hindus as compared with non-Hindu groups. Moreover, the effects of the pandemic are significant only for the poor households and are statistically insignificant for the middle and rich wealth groups; however, the odds ratios are more than one for all three groups.

Figure 1. Stratified analysis by mother's education, caste, religion, and household wealth

Notes: (i) Coefficients from linear regression are reported; the bar shows 95% confidence interval (CI). A 95% confidence interval means that if you were to repeat the experiment over and over with new samples, 95% of the time the calculated confidence interval would contain the true effect. (ii) All models adjust for confounding variables and district-specific, time-invariant characteristics.

Policy implications

In summary, our results provide insights into the short-term impacts of the Covid-19 pandemic on birth outcomes in India. We demonstrate that there was an increased risk of LBW incidence during the pandemic, underscoring the need for targeted interventions to support maternal and neonatal health during potential future crises. The adverse impacts of the Covid-19 pandemic on birth weight in India highlight several policy implications to address and mitigate these effects. Strengthening and ensuring continuous maternal, antenatal, neonatal and postnatal healthcare services during the pandemic and other emergencies are important for reducing the incidence of LBW. Services such as expanded telehealth and mobile health clinics can be effective in covering vulnerable populations in rural and underserved areas. Targeted interventions for socioeconomically disadvantaged groups, such as SC/ST communities and those with lower educational attainment, could be crucial for reducing disparities in birth outcomes. Future research should focus on understanding the long-term effects of the pandemic on child health and further investigate the mechanisms behind the observed trends.

Notes:
1. Sensitivity analysis using a random-effects model confirmed similar results, supporting the main findings.