Effect of Prenatal Exposure to Air Pollution (Particulate Matter and Carbon Monoxide) on Infant Adiposity

Introduction

Childhood obesity is a growing public health concern.¹ Currently, excess adiposity and the accompanying comorbidities are the most common threats to human life expectancy. Globally, 10% of school children in 5-17 years age group are obese or overweight.² About 27.3% children were found to be overweight or obese in the USA and 9.7% of infants and toddlers were overweight.^3,4 According to a recent systematic review, the obesity prevalence rate in India varied from 11.8% to 31.3%.⁵ The findings of the systematic review showed that prevalence of overweight in India among children aged <5 years ranged from 0.4% to 1.4%, among 5-19 years children, ranged between 6.1% and 25.2%, while that of obesity ranged between 3.6 and 11.7%.⁶ The evidence suggests the causal relationship between maternal pregnancy, body mass index (BMI) and glucose that can cause adiposity in the new-borns and infants.^7,8 Other factors such as early introduction of bottle feeding contributes to adiposity in infants.⁹ Development of childhood obesity can be attributed to various reasons and combination of both antenatal and postnatal exposures including environmental factors.¹⁰ The environmental pollutants during pregnancy are found to disrupt the developmental processes in foetus and infant, with possible lifelong consequences.¹¹

Majority of children (98%) under five years of age in low- and middle- income countries are exposed to particulate matter of diameter ≤ 2.5 μm, PM_2.5 levels above WHO air quality guidelines. The biological mechanism that explains the association between the exposure to air pollution and infant adiposity states that the particulate matter induces inflammation, oxidative stress and elevated C-reactive protein (CRP) levels during pregnancy.^11,12 The oxidative stress, elevated levels of CRP and plasma concentration of CRP during pregnancy is found to be associated with adiposity.^13,14

The exposure to the pollutants during in utero and early postnatal period can affect individuals leading to the development of serious chronic pathologies such as cardiovascular and metabolic diseases like obesity and diabetes, respiratory, and neurodegenerative diseases in later life.^15,16 The higher concentrations of ambient air pollutants during pregnancy have been associated with small for gestational age and lower infant birth weight.¹⁷ Low birth weight and subsequent rapid growth have in turn been linked to greater risk of obesity and cardiometabolic diseases in childhood and adulthood.¹⁸

A few human studies with restricted participation suggested that exposure to ambient air pollution during prenatal stages might impact infant growth and body composition. These findings could have implications for the risk of child obesity and related metabolic health, although the outcomes of these studies exhibited inconsistencies.¹⁹ Nevertheless, animal studies have indicated that prenatal exposure to a combination of air pollutants can lead to an elevated proinflammatory condition in the epididymal fat, coupled with increased body weight, inflammation, oxidative stress, and dyslipidemia. Moreover, many prior investigations into prenatal air pollution exposure focused solely on birth weight as an outcome, with only a limited number exploring infant adiposity.²⁰

Very few studies have been conducted in India to study the impact of prenatal exposure to air pollution on infant adiposity and the findings showed lower exposure to the average ambient air pollution. The findings from an Indian study demonstrated negative association of PM_2.5 exposure during pregnancy with adiposity at five-months and reported gender as an effect modifier.¹¹ Hence, we find a wider scope for studying the association between prenatal exposure to air pollutants and infant adiposity. In this regard, the Indian Institute of Public Health, Bengaluru started a prospective cohort study titled ‘Ambient and Indoor Air Pollution in Pregnancy and the risk of Low birth weight and Ensuing Effects in Infants (APPLE): A cohort study in Bengaluru, South India’, and the protocol of this study has been published elsewhere. We hypothesize that the in utero exposure to environmental pollutants such as PM_2.5, PM₁₀ and CO has detrimental effects on infant growth after birth, leading to the development of adiposity among younger children. This article presents the results discussing the effect of prenatal exposure to air pollution on adiposity in infants, mediated through low birth weight.

Materials and Methods

This was an ongoing prospective study nested within a cohort, ‘Maternal Antecedents of Adiposity and Studying the Transgenerational role of Hyperglycemia and Insulin (MAASTHI) cohort study’. MAASTHI characterizes the impact of diet, psychosocial stress, and pregnancy complications during pregnancy, in the intrauterine milieu on the foetus. Pregnant women inhaling polluted air are at a higher risk of having infants with adiposity. The current study measured the exposure of air pollution including diet, psychosocial stress to assess its impact on low birth weight and risk of adiposity among infants. The protocol of the study was published in the Wellcome Open Research Journal in 2018. The study targeted 516 pregnant women aged 18 years and above, residing in the East and West zones of urban slums of Bangalore. Thus by evaluating the putative role of air pollution during pregnancy on infant adiposity, we intend to reduce the risk for non-communicable diseases later by addressing the environmental factors which are leading to non-communicable diseases (NCD) risk factor later in adulthood.

The study was conducted in the urban slums of Bangalore from September 2017 to Sep 2019. Written informed consent was obtained from the participants. Pregnant women residing in slums belonging to the age group of 18-45 years were approached, and women with a gestational age of fewer than 18 weeks, and no coexisting illness were recruited in the study.

Exposure assessment

We used personal sampling monitors to measure the concentration of PM_2.5, PM₁₀, and CO at the individual level. The pregnant women were requested to wear a personal sampling device for an entire day (24-hours) in the second and third trimesters. The data collected for the two measurements of PM_2.5, PM₁₀, and CO for each pregnant woman were stored in the server and downloaded by the account manager. The ambient data were extracted from the nearby ambient air quality monitoring station situated within a 5 km radius from the study site, installed by the Karnataka State Pollution Control Board which was considered as a proxy measure for ambient exposure. The total average of personal exposure levels was computed by adding up the second and third-trimester average measurements for the final analysis.

Outcome

Infant adiposity was measured in terms of infant weight for age Z-score. The weight of the infant was measured using weighing scale SECA 354 and recorded to the nearest 0.1 kg. SECA 354 weighing scale was used for baby weighing at six months age. In each of these visits, the height and the crown-rump length was measured with SECA 417 Infantometer. The circumference measurements, such as waist and hip, is a sensitive marker for abdominal adiposity.²¹ Skinfold thickness is the most accurate measure of adiposity. We used the sum of skinfold thickness (biceps + triceps + subscapular) to estimate overall adiposity. Chasmors’ body circumference tape was employed for assessing measurements such as chest, waist, hip, mid-upper arm circumference, while skinfold thickness (biceps, triceps, and subscapular) was determined using the Holtain Caliper from Holtain, UK. The research team adhered to standardized procedures for all measurements, undergoing regular testing and certification for internal validity through assessments conducted at St. John’s Research Institute in Bengaluru. We measured weight in kilograms, and readings were taken to the nearest 0.5 g. Height and circumferences in centimetres, and skinfold thickness in millimetres were measured. We obtained three readings for each measurement. We created dichotomous variables of all anthropometric markers using the above 85th percentile cut-off.

Covariates

Age, socioeconomic status (SES), parity, depression status during pregnancy, mother’s weight and height, spouse’s smoking status, birth weight and gender of the infants, were collected at the follow up visit using a structured questionnaire. The weight and height of the pregnant women were recorded using SECA 213 stadiometer and Omron HN-283 weighing scale by the trained research staff. We used a cut-off of 13 and more to indicate the depressive symptoms in the postpartum period, as suggested by several studies. The prevalence of depressive symptoms were assessed using Edinburgh Postnatal Depression Scale (EPDS) and cutoff level of ≥12/13 were used to categorise them as ‘no depressive symptoms’ and ‘depressive symptoms’. The socioeconomic status of the participants was assessed using the Kuppuswamy scale. The total score of this scale ranges from 3-29, and it classifies families into three groups - 3 to 10 as lower class, 11-25 as middle class, and above 25 as upper class. World Health Organization cut off measurement was used to diagnose gestational diabetes mellitus. The fasting blood glucose level (≥92 mg/dL) and two-hour sample (≥153 mg/dL) were considered as the cut off measurement.

Statistical methods

We included data of 519 subjects for the analysis. All the participants irrespective of the exposure assessment were followed up at delivery and infant follow-ups were done at birth, six months, and one year. Among the total recruited, we found three stillbirths and six abortions. We were able to track 350 subjects at six months follow-up and while the remaining participants could not be tracked due to the non-response, or change in contact number/non existing contact. Among those who had completed six month follow up, we were able to record anthropometric details for only 190 infants. The final analysis was performed using SPSS statistical software version 24. Descriptives were shown in terms of mean and standard deviation. The outcome variable considered was adiposity; so weight for age Z-score, skinfold thickness were considered. The linear regression analysis was done to examine the association between pollution parameters and infant weight for age Z-score and skinfold thickness. Both adjusted and unadjusted correlation coefficients have been reported along with P-value and 95% confidence interval.

Results

The study recruited 519 pregnant women and among them, 297 participants underwent an individual pollution assessment; indoor and ambient exposure data were available for all of them. We could track 456 subjects at birth (87.5%), and among them, data of 297 children with complete anthropometric details were recorded. The remaining could not be completed due to COVID-19 pandemic. We completed follow-up assessment of 362 (80%) children among 449 children at six months and among them, 160 children had complete anthropometric data. We completed 356 (85%) annual assessments against 418 intended and could obtain data with complete anthropometric details from only 95 children.

With the available data of 160 subjects along with anthropometric assessments at six months, we estimated the adiposity prevalence among children. Weight for age Z-score above the 85^th percentile and sum of skinfold thickness above 90^th percentile were considered to be at risk for adiposity. We observed that nearly 25.2% of children were at risk for adiposity when subscapular skinfold thickness (SSFT) above 90th percentile was considered, whereas weight-for-age Z-score (WAZ) above 85^th percentile gave the adiposity prevalence of 13.8%.

The case summaries of the important variables have been listed in Table 1. We observed that the mean values for the mother’s weight, height, pregnancy exposure to indoor PM_2.5 and PM₁₀ were higher among children with weight for age Z-scores, whereas pregnancy exposure to indoor and ambient CO was lower. The depression score during pregnancy, social support scores were higher among those children with adiposity.

The model was adjusted for age, socio-economic status, parity, pregnancy depression, gestational diabetes (GDM) and status during pregnancy, mother weight, height, spouse’s smoking status, gender of the child and child age at follow up.

The linear regression model was run to evaluate the study hypothesis and we found a statistically significant association between pregnancy exposure to particulate matter and weight for age Z-score and triceps skinfold thickness.

One unit increase in PM_2.5 was associated with an increase in WAZ-score by 0.006 units and an increase in one unit of PM₁₀ was associated with an increased WAZ-score by 0.003 units. We did not find any significant association between indoor carbon monoxide exposure, ambient PM, and CO exposure and net exposure of CO with weight for age Z-score. We did found significant association between exposure to PM_2.5, PM₁₀ and net PM and triceps, whereas findings related to sum of skinfold thickness were not significantly associated with any of the exposure parameters considered here.

Discussion

The findings of our study found significant evidence for an association between prenatal exposure to particulate matter and adiposity among children at six months of age. We did not find any significant results associated with CO exposure during pregnancy and adiposity. We hypothesized based on the existing evidence that on exposure to air pollution during pregnancy, birth weight will decrease and these effects are usually seen in later infancy among those infants highly exposed to air pollution in utero.

The findings are supportive to the existing evidence. According to Fleisch AF et al., newborns exposed to greater levels of traffic-related pollution during pregnancy have a negative impact on postnatal weight gain in infants aged 0 to 6 months and revealed that children whose mothers lived closest to major highways during pregnancy had a 1.9 mm higher sum of skinfold thickness in early childhood.^22,23 As per the findings of Jerette M et al., traffic pollution is significantly associated with an increase in BMI among children aged 5 to 11 years.24 According to Dong DH et al.’s findings, an increase in PM₁₀ in the interquartile range is associated with a 1.19 times higher risk of obesity in children aged 2 to 14 years.²⁵ The findings of Bont J de et al., based on a longitudinal study conducted among children aged 2 to 5 years, showed that early childhood exposure to PM₁₀ and PM_2.5 is related to a 2% increase in the risk of overweight and obesity.26 According to Wilding S et al., increased mean annual PM₁₀ exposure was linked to a 2% rise in childhood overweight or obesity.²⁷ Schembari A et al. showed that pregnancy exposure to ambient PM_2.5 is associated with larger triceps and subscapular skinfold thickness.²⁰

Our findings are contradicting to those of Starling AP et al. They found no association between pregnancy exposure to outdoor PM_2.5, traffic and infant obesity in a prospective cohort analysis of mother and child pairs.¹¹ Fioravanti S et al. reported that there was no association between vehicular traffic exposure and overweight among children in their longitudinal follow up study.²⁸

There were no studies conducted in the developing countries reporting the association between pregnancy exposure to air pollution and obesity, taking into consideration the individual exposure levels. Most of the studies analysed the risk of obesity among children of school going age or among adults, considering ambient exposure as an proxy measure.

Inconsistent findings were observed in the previous studies reporting the effect of prenatal exposure to environmental pollutants on weight gain among children. A systematic review done in the area of environmental aspects stressed on the need for research to evaluate the impact of environmental exposure levels on child weight outcomes. Apart from this, there are several significant findings regarding the prenatal maternal smoking increasing the odds of offspring overweight. The majority of evidence came from the high income countries and only few prospective studies were reported from the lower and middle income countries. Our study findings can contribute to the significant findings to support evidence regarding the impact of environmental exposure on child obesity.

Dr. Paula Baillie-Hamilton first proposed that higher exposure to chemicals can cause growth reduction, while lower exposure levels can increase the weight. The change in environment due to exponential production and synthesis of organic and inorganic chemicals, at the higher exposure levels cause weight loss, while lower exposure will cause weight promotion. She explained that the same principle will be used to increase the body weight by administering pharmaceutical drugs in humans. She proposed that even low levels of certain chemicals damage the body’s natural weight control mechanism.²⁹ The damaged weight control mechanism might lead to weight promoting actions, thus leading to obesity.

Thus it is important to understand the explanation on how these airborne particles cause metabolic dysfunction, in order to understand the role of air pollution of NCD risk factors. The mechanisms behind these can be explained as follows. Exposure to particulate matter induces oxidative stress and inflammation in organ tissues.³⁰ This will lead to the metabolic dysfunction and increase lipid deposition in the adipose tissues and hence can induce damage in the weight control mechanism, leading to obesity. Therefore the health effects in terms of obesity reflect the effects of particulate matter. These findings suggest the urgent need to reduce air pollution exposure, given the growing burden of obesity.

The main strength of the present study is that it was a prospective design study with a follow up from birth to one year of age that allowed for examining the effects of prenatal pollutant exposure on early childhood period. We used strict protocol for collecting anthropometric variables and research staff were trained in anthropometry by the St. John’s research team. SOP was followed for data collection using standardized questionnaire, and also for BP and blood collection, ensuring quality data. We recorded the exposure measurement at individual level and also considered the standard monitoring station data for ambient exposure. The study had limited power due to data considered from small sample of children from birth to one year with anthropometric measurements. The COVID-19 pandemic impacted our anthropometric data collection.

Conclusion

The study results indicate a potential correlation between prenatal air pollution exposure and infant obesity, specifically in terms of weight and skinfold thickness at six months of age. These findings suggest that exposure to air pollution during pregnancy may have a lasting impact on postnatal growth. The air pollutants may act as a obesogenic by inducing oxidative stress and metabolic dysfunction and these effects vary according to pollutant type, duration of exposure and timing of exposure. Further large scale and long term follow up studies are needed to better understand the biological mechanism linking air pollution exposure and its role in obesity risk among children.

Funding

This study was funded by the National Network Programme on Human Health, Climate Change Programme-(SPLICE), Department of Science & Technology, Government of India. [DST/CCP/ NHH/108/2017(G) and DST/CCP/NHH/108/2017(C)] and Wellcome Trust/DBT India Alliance Fellowship [Grant No. IA/CPHI/14/1/501499] awarded to Giridhara R. Babu.

Conflict of interest

The authors declare that they have no competing interests