Determinants of birthweight in rural Sri Lanka; a cohort study

Birthweight is the first weight of the fetus obtained after birth. Globally, 15% to 20% of babies are born with low birthweight (less than 2500 g), and the issue disproportionately affects babies from low and middle-income countries [1]. In either setting, marginalized and disadvantaged population subgroups show a higher prevalence of low birthweight and its associated complications, including mortality and ill health, demonstrating a cycle of health inequity [2‐4].

Birthweight is unique since it reflects the cumulative impact of numerous factors on the baby's health. These include quality of maternal care services[5, 6], preconceptional and antenatal nutritional status of the mother [6, 7], maternal comorbidities [8], socioeconomic conditions [4, 6], genetics and epigenetics [9]. Intermediary social determinants of health such as the living and working environment [8, 10‐14], food security [15], social capital and experiencing violence [11, 16, 17], and mental health of the mother [18, 19] can affect the birthweight of their babies. Preterm deliveries, either spontaneous or provider-initiated (due to medical, obstetric, or other reasons), also are a major determinant of lower birthweight [6, 8, 18, 19]. Birthweight can predict many short- and long-term adverse health outcomes in children. Neonates with low birthweight are 20 times more likely to die during infancy than their counterparts [14]. Being very low birthweight can affect children's cognitive function and increase the risk for diseases such as asthma and chronic kidney disease [20‐22]. An association has been shown between birthweight and non-communicable diseases such as hypertension and type 2 diabetes [23, 24]. Birthweights at the higher end of the spectrum are associated with the risk of obesity and cancers [25].

Global nutrition targets aim for a 30% reduction in the global incidence of low birthweight by 2025 [26]. However, progress toward this goal is slower than expected [27]. Similar to the rest of the underperforming nutritional indicators albeit having a robust public health system performing well in other aspects of maternal and child health, low birthweight prevalence is comparatively high in Sri Lanka [28]. Many interventions have already been implemented to control the low birthweight incidence, including almost all evidence-based interventions recommended by the WHO in the policy brief on the world health assembly's low birthweight targets. However, similar to the global situation, little progress has been achieved over the past decades [29]. Shifts in the epidemiology of diseases in pregnancy, such as gestational diabetes, may further complicate the picture, increasing the number of births with higher birthweights [30].

Further improvements in birthweight require detailed assessments of contributory factors, especially those often overshadowed by prominent nutritional issues. Going by the overarching principle of sustainable development goals, "leaving no one behind", addressing issues concentrated in specific sub-population groups is also important. Therefore, we adopted a bio-psycho-social approach to study the determinants of birthweight in a mother-infant cohort from a rural setting in a middle-income country.

Out of 532 pregnant women enrolled, the baby's birthweight data was available in 497 (Lost to follow-up 6.5%). Participants were between 16 to 44 years of age. The mean age was 28.5 years (SD 5.6). The median monthly family income was 205.6 USD (Minimum 19.7 USD, Maximum 2951 USD). The majority of the participants were ethnic Sinhalese (n = 423, 80.4%), were unemployed (n = 437, 82.1%), and had completed education at least up to grade 11 (n = 298, 57.2). One-fifth were from farming families (n = 114, 21.5%). Of the participants, 18 (3.4%) reported that during the month prior to the baseline assessment, their husband or another family member had verbally/physically, or sexually done things that made them feel unhappy/sad on a daily basis. Another 25 (4.8%) reported they faced such incidents several times a week, while 75 (14.3%) and 263 (50.3%) had similar "abuse" several times a month and less often, respectively. Only 142 (25.2%) reported not experiencing such violence.

The gestation period at the time of birth varied between 29 weeks plus one day to 42 weeks and four days, with a mean of 38 weeks and six days. There was an entry with 43 weeks and two days PoG with an ultrasound confirmation. Of all births, 61(11.5%, 95% CI 9.06,14.41) were preterm ( born at 36 weeks plus six days or earlier PoG).

Birthweight was approximately normally distributed (Skewedness = -0.21) and ranged from 1020 to 4200 g. The mean birthweight was 2912 g (Standard deviation 456.6 g). Of the babies, 72 (14.6%, 95% CI 11.66,17.85) had low birthweight, while 46 (9.3%, 95% CI 7.01,12.13) had higher birthweight (≥ 3500 g). The distribution of birthweight according to the PoG at birth is shown in Fig. 2.

The current study showed that 14.6% of babies in this rural setting are born with low birthweight. Preterm birth, maternal BMI, and mid-pregnancy psychosocial wellbeing of mothers affect the birthweight of babies born in this community.

Birthweight is both an indicator and a predictor of perinatal health, and its effects are carried on further across the lifespan [8]. In Sri Lanka, the routine surveillance of public health data gives a reliable estimate of the low birthweight incidence, 11.8% in 2017 [29]. However, a detailed assessment of the distribution of birthweight or its determinants cannot be ascertained with surveillance data. Even though previous studies have shown marked geographical variations in average birthweight and low birthweight incidences at the district level, variations at micro-geographical levels are also scarce [29]. The role of the psychosocial environment of pregnant women in perinatal health in low and middle-income settings is often overshadowed by stronger determinants such as nutritional issues. Conducted in a homogenous, rural community in a middle-income country, the current study contributes to addressing these gaps in research evidence.

Marked micro-geographical variations in marital and family cohesion (including the pregnant women's perceptions regarding care received from their husbands and family members and family conflicts) have been described within this district [33]. This particular community displayed a possibly high prevalence of "abuse" compared to the rates reported in other communities. For example, 72.8% of the participants in the current study reported perceived physical, verbal, or sexual "abuse" from their husband or another household member at least once during the past month. Even though the assessment methods may vary, the previously reported prevalence of physical or sexual violence during pregnancy was 4.2% [34]. Demographic and health survey 2016 reported that.4% and 17% of ever-partnered women in Anuradhapura and Sri Lanka experienced intimate partner violence [34]. Surprisingly, despite hostile social conditions, this region has the third-lowest poverty headcount index in the district, showing that general economic status is comparatively good [31].

Stress experienced by pregnant women leads to changes in the hypothalamo-pitutary axis resulting in higher cortisol levels[35]. Higher maternal cortisol levels are associated with preterm births and lower birthweights [35]. The presence of depression or anxiety [18, 36, 37], inadequate support received from the family structure [38], experiencing violence [11], and facing stressful circumstances [19, 39] have all been associated with low birthweight or preterm births or both, emphasizing the importance of maternal psychosocial wellbeing during pregnancy. The current study has shown that the factor score for psychosocial wellbeing; the composite variable for frequency of experiencing abuse, perceived care from husband and family, perceived wellbeing, and perceived happiness during pregnancy was significantly associated with birthweight even after other possible confounders such as the economic status and medical conditions were controlled for. This highlights the importance of improving domestic cohesion, preventing violence, and improving maternal mood during pregnancy to reduce low birthweight and preterm births in this community. However, the health system will need major reorientations in the existing human and physical resources to cater to these demands [40, 41].

Maternal early pregnancy BMI is believed to be closely representative of the pre-pregnancy BMI and hence used as an indicator of the mother's nutritional status [42]. Maternal BMI was a strong predictor of birthweight in this model (4% of unique variance explained, p = 0.001) even after controlling for several factors, including the PoG at which the measurement was taken. However, to observe a 120 g increase in birthweight, BMI should be increased by 4.8 kg/m². Though the mean birthweight of babies born to overweight and obese pregnant women was significantly higher than in other BMI categories, there was no significant difference between babies born to women with low BMI and women with normal BMI. Therefore, popular nutritional interventions aiming at pre-pregnancy BMI to prevent low birthweight will not be effective in this community.

Preterm birth was the strongest predictor of birthweight in the regression model. This was after controlling for hypertensive disorders and gestational or chronic diabetes in pregnancy which are among the commonest causes of provider-initiated preterm deliveries [43]. Therefore, why preterm births occur and how to reduce the number of preterm births need to be investigated.

Proxy indicators used during field care provision to monitor fetal growth cannot effectively predict birthweight. Total weight gain during pregnancy showed a very weak correlation with birthweight (r = 0.094, P 0.038). Third-trimester SFH being compatible with the PoG and having gained weight according to national recommendations were not shown to be significantly associated with the birthweight.

In a setting where a comparatively higher proportion of pregnant women suffer 'abuse' in their homes, psychosocial wellbeing during pregnancy, a largely neglected aspect in routine maternal care provision, was an important determinant of birthweight of babies. Strengthening the health system approach beyond routine nutrition-focused interventions to address the social determinants and psychological wellbeing in pregnant women would help improve perinatal health in developing countries.