Ambient and household air pollution on early-life determinants of stunting – a systematic review and meta-analysis

DOI: https://doi.org/10.21203/rs.3.rs-23018/v1

Abstract

Background Stunting is an important risk factor for early growth and development with health implications throughout the life course. While maternal exposure to particulate matter (PM) has been linked to early determinants of stunting, existing evidence has rarely captured the most vulnerable populations.

Methods We conducted a systematic review and meta-analysis of the peer-reviewed literature to assess the evidence of the association between ambient and household PM pollution exposure and postnatal stunting (height-for-age z-score), and prenatal determinants (i.e., intrauterine growth restriction and small for gestational age) that would greatly increase children’s risk of stunting. Relevant manuscripts published from 2000 to 2019 were reviewed. Random effect models were used to estimate pooled odds ratios (OR).

Results Thirty-two studies conducted in 18 countries met our inclusion criteria. We found significant positive associations between prenatal determinants of stunting and a 10 μg/m 3 increase in PM 2.5 during the first trimester (OR=1.02; 95% confidence interval (CI): 1.00–1.04) and second trimester (OR=1.04; 95% CI, 1.01–1.07). Similar associations were found for prenatal determinants of stunting of high versus low quartiles of PM 2.5 exposure during the whole pregnancy. Postnatal stunting was found to be positively, though insignificantly, associated with postnatal exposure to household air pollution.

Conclusions Our analysis shows evidence of increased risk of prenatal determinants of stunting with ambient particulate exposure, especially during first trimester, and suggestive evidence of elevated stunting risk with postnatal exposure to household air pollution. This evidence reinforces the importance of promoting clean air as part of an integrated approach to preventing stunting.

Background

Air pollution is a major environment-related health threat and an important determinant of child health globally. The World Health Organization (WHO) recognizes air pollution as “an overlooked health emergency for children around the world,” noting that the issue can be especially severe for children living in low- or middle-income countries [1]. In 2017, air pollution exposure was responsible for nearly 372,000 deaths in children under 5, making it the third leading risk factor for premature death after malnutrition and unclean water [2]. Unlike many more prominent risk factors, air pollution is pervasive. WHO (2018) estimates that 93% of children and teens ≤ 15 years globally are exposed to ambient air pollution levels higher than 10 µg/m3, the health-based limit in WHO’s Air Quality Guidelines. Burning or fuel combustion is the main source of air pollutants most harmful to human health, the most important being particulate matter less than 2.5 microns in diameter (PM2.5). PM2.5 at high concentrations can manifest as smoke or haze, but it causes health effects at concentrations even too low to be visible. Exposure to elevated levels of particulate matter during childhood is associated with increases in pneumonia, asthma, bronchitis and other respiratory infections and diseases [3, 4]. In recent years, many researchers have turned to elucidating the relationship between in utero exposures and adverse pregnancy outcomes, such as low birthweight and preterm delivery.

Stunting, or linear growth failure in childhood, is a largely irreversible outcome with long-term impacts on children and their communities. In addition to height and physical development concerns, stunted children often achieve lower developmental test scores and suffer from diminished cognitive development and reduced economic activity [49]. Stunting begins in utero, and manifests mostly during the first two years of postnatal life, with increasing prevalence until age 5 [1012]. Globally, an estimated 149 million (22%) children under 5 in 2018 were stunted, with a height-for-age z score (HAZ) of − 2 standard deviations (SD) below the WHO child growth standards median [8]. The highest prevalence of stunting is in Oceania (38%), followed by Southern Asia (33%) and Southeast Asia (25%) [8].

Study Purpose and Rationale

Two previous meta-analyses have suggested positive associations between ambient PM2.5 pollution and adverse pregnancy outcomes, defined as small for gestational age (SGA) or intrauterine growth retardation (IUGR) [13], and between household air pollution from cooking using solid fuels and postnatal stunting measured by HAZ [14]. However, considerable gaps in the evidence remain, especially regarding the strength of evidence for household air pollution or ambient pollution concentrations several-fold greater than the WHO air quality guidelines, both of which occur in regions with the highest stunting prevalence. We therefore aimed to provide an up-to-date systematic review and meta-analysis of the evidence of association of ambient and household air pollution on stunting-related outcomes. Specifically, we reviewed all studies that examined adverse pregnancy outcomes (i.e. SGA, IUGR) that also serve as prenatal determinants of stunting. Based on a meta-analysis of 19 longitudinal birth cohorts, Christian et al. [15] found that SGA was associated with 2.4 times the odds of postnatal stunting in low- and middle-income countries, and a recent study using a birth cohort in the United States even reported a 3.55 odds of stunting among SGA newborns at age 5 years as compared to newborns appropriate for gestational age [16]. In addition, we included studies that evaluate the impact of air pollution on postnatal stunting (i.e., HAZ) among children aged 0–59 months.

Methods

Search Strategy

A systematic search of peer-reviewed articles that assessed the effect of ambient and household particulate matter exposure during or post-pregnancy on stunting-related outcomes was carried out on PubMed and Google Scholar in May 2019. The following terms were used in the search: “air pollution” or “particulate matter” or “indoor air pollution” or “household air pollution” or “solid fuel” or “biomass” AND “adverse pregnancy outcomes” or “adverse birth outcomes” or “stunting” or “height-for-age” or “HAZ” or “small for gestational age” or “SGA” or “intrauterine growth retardation” or “IUGR”. All original studies published between January 2000 and April 2019 were considered. Reference lists of previous relevant published reviews, meta-analyses, and the identified articles were also searched. The last search was updated on 17 May 2019. Daily time-series studies, case reports, case studies and studies available only in abstract form were excluded. All studies were evaluated independently by two reviewers to ensure identified articles were suitable for inclusion in the meta-analysis. Reporting of findings of this systematic review also complies with the PRISMA statement [17].

Eligibility Criteria and Study Selection

Eligible studies included in the meta-analysis met all the following inclusion criteria: (1) original articles examining exposure to particulate matter with aerodynamic diameter less than 10 µm (PM10), PM2.5, or household air pollution exposure from cooking with solid fuel; (2) definitions of growth faltering: moderate and severe stunting defined as HAZ < – 2 SD, and severe stunting defined as HAZ < – 3 SD of the WHO child growth standards or reference population median; SGA defined as birth weight below the 10th percentile for a given gestational age and gender of the newborn in the study population or similar; or IUGR defined as birth weight below the 10th percentile for gestational age or similar; (3) quantitative evaluation (i.e. linear or logistic regression coefficients) of the relationship; (4) exposure period was calculated as whole pregnancy and/or specific trimesters; (5) articles were written in English. Studies or effect estimates were excluded if one or more of the following occurred: (a) did not meet at least one of the above inclusion criteria; (b) exposure period (e.g. trimester-specific, during or post pregnancy) was not explicitly reported; (c) effect estimates in studies could not be converted into odds ratio (OR) and 95% confidence intervals (CI) in risk of stunting; (d) studies adjusted for two or more pollutants in the same model; (e) reviews and repeat literature including secondary analysis of data. For multiple publications of the same birth population in a region of the same outcome, only the study with the largest number of observations and/or the longest study period was included in the meta-analysis to avoid the overlapping datasets in the same outcome. Likewise, if multiple estimates using various exposure assessment tools (e.g. monitoring, satellite) of the same outcome from the same population were presented in a single study, only estimates with the largest sample size and/or smallest standard errors were included.

Data extraction and statistical analysis

Information extracted from each study included: first author’s surname, year of publication, country, study period, birth population, birth outcomes, exposure type (e.g. PM2.5, solid fuel), exposure assessment (e.g. satellite, monitoring data), exposure period (e.g. trimester-specific and/or entire pregnancy), risk measure (e.g. odds ratio, relative risk), effect size, and adjustment of other factors or covariates. As all eligible studies that examined ambient particulate pollution reported effect estimates either for continuous exposure (when they assumed linear associations) or by categories (e.g. tertiles, quartiles, or quintiles) of exposure, we performed all analyses using both continuous exposure (per 10 µg/m3 increase in PM2.5 or PM10 concentration) and categorical exposure defined as high vs. low, where the high level was defined as the highest study-specific category and the low level as the lowest study-specific category. For household particulate pollution, all analysis was performed using categorical exposure defined as high vs low exposure. To ensure comparability, reported relative risks were converted to ORs using the approximation approach described by Zhang and Yu [18] if information on cases and non-cases were available from the study.

Given that SGA and IUGR are often used interchangeably, both outcomes were analyzed together as SGA in our study. A random effects model was used to estimate the pooled effect measures, with between-study heterogeneity assessed using I2 statistic (25%, 50%, and 75% were used as rules of thumb for low, moderate and high heterogeneity, respectively) [19]. In this meta-analysis, effect estimates were grouped by exposure period (e.g. trimester-specific), and heterogeneities were also analyzed by exposure period. The possible publication bias was assessed with a Begg’s test and the degree of asymmetry was evaluated by Egger’s tests [20, 21]. As part of the sensitivity analysis, each eligible study was removed one-by-one from the meta-analysis to examine the robustness of our pooled ORs. We restricted the analysis to eligible studies that had the same outcome definitions. All statistical analyses were performed using R Software version 3.6.0 (R Foundation for Statistical Computing, Vienna, Austria), and a p-value < 0.05 is considered statistically significant.

Results

Characteristics of included studies

Of the 210 articles identified in the literature search and reference lists of previous relevant reviews, meta-analyses, and identified articles, 57 articles with possibly eligible titles were selected by either reviewer, and 11 were subsequently excluded because the abstracts were irrelevant. Full manuscripts of the remaining 46 articles (37 on ambient, nine on household) were reviewed. Of 37 studies that examined ambient particulate pollution, 28 were restricted to either SGA or IUGR and ambient PM2.5 or PM10 pollution exposure during the entire pregnancy, and/or any of the three trimesters. Nine studies were not considered further due to the small number of effect estimates for a meta-analysis, and these studies fell into three categories: they either 1) evaluated very small for gestational age [22], HAZ [23], or physician-diagnosed fetal growth restriction [24] as outcome measures; 2) examined total suspended particles [25], particles with aerodynamic diameter of 7 microns or less [26], or PM10 emitted from incinerators [27] as exposure measure; or 3) evaluated exposure during first month of pregnancy [28] or post pregnancy [29, 30]. Of nine studies that evaluated household air pollution, six examined the impact of post-pregnancy exposure on HAZ as a dichotomous measure, whereas one employed HAZ as a continuous measure [31], and another compared stunting risk from pollution from indoor cooking with that from outdoor cooking [32]. Additional information was not available from four studies to allow the conversion from relative risks to OR. Therefore, we included in the final meta-analyses 26 articles on ambient air pollution and five articles on household air pollution, stratified by outcome, exposure measure and gestational age. Figure 1 shows the study screening process.

Table 1 summarizes the characteristics of the studies included in this review. Of 26 studies on ambient air pollution, most were from countries with low ambient air pollution concentrations (e.g. the United States). Three studies were from Asia and one was from Australia. Study periods ranged from 1975 to 2017, with the majority of the individual studies conducted in the 2000s. The population of participants from individual studies ranged from 1,514 to 3 million. Exposure assessment methods of most studies were based on existing central monitoring data, while a few used advanced monitoring approaches (e.g. satellite data, or modeled pollution measurement from land-use regression models). PM2.5 and PM10 exposures ranged from 4 to 71 3 and from 12 to 102 µg/m3, respectively.

Table 1
Characteristics of the studies included in the meta-analysis.
First Author
Year Published
Country
Data Year
Study design
Sample Size
% of outcome
Exposure Measure
Exposure Period
Outcome
SES adjusted?
Smoking adjusted?
Wang et al. 2019
2019
Guangzhou, China
2015–2017
Retrospective cohort study
506,000
9.7%
PM 2.5, PM 10
Entire pregnancy, trimesters
SGA
No
No
Percy et al. 2019
2019
Ohio, USA
2007–2010
Retrospective cohort study
224,921
10.8%
PM 2.5
Entire pregnancy, trimesters
SGA
No
Yes
Smith et al. 2017
2017
London
2006–2010
Retrospective cohort study
540,365
9.5%
PM 2.5, PM 10
Entire pregnancy
SGA
Yes
Yes
Lavigne et al. 2016
2016
Canada
2005–2012
Retrospective cohort study
818,400
9.3%
PM 2.5
Entire pregnancy, trimesters
SGA
Yes
Yes
Stieb et al. 2016
2016
Canada
1999–2008
Retrospective cohort study
2,966,705
8.3%
PM 2.5
Entire pregnancy, trimesters
SGA
No
No
Pereira et al. 2016
2016
Connecticut, USA
1988–2008
Retrospective cohort study
9,587
na
PM 10
Entire pregnancy, trimesters
SGA
No
Yes
Qian et al. 2016
2016
Wuhan, China
2010–2013
Prospective cohort study/case-control study
95,911
8.8%
PM 2.5, PM 10
Entire pregnancy, trimester 1
IUGR
Yes
Yes
Capobussi et al. 2016
2016
Italy, Europe
2005–2012
Retrospective cohort study
27,128
10.2%
PM 10
Entire pregnancy, trimesters
SGA
Yes
No
Winckelmans et al. 2015
2015
Belgium
1999–2009
Retrospective cohort study
494,653
9.4%
PM 10
Entire pregnancy, trimesters
SGA
No
No
Hannam et al. 2014
2014
UK
2004–2008
Retrospective cohort study
5,721
8.6%
PM 2.5, PM 10
Entire pregnancy
SGA
Yes
Yes
Hyder et al. 2014
2014
Connecticut and Massachusetts, USA
2000–2006
Retrospective cohort study
643,839
10.0%
PM 2.5
Entire pregnancy, trimesters
SGA
No
Yes
Gray et al. 2014
2014
North Carolina, USA
2002–2006
Retrospective cohort study
457,642
9.3%
PM 2.5
Entire pregnancy
SGA
No
Yes
Lee et al. 2013
2013
Pennsylvania, USA
1997–2002
Retrospective cohort study
34,705
8.5%
PM 2.5, PM 10
Trimester 1
SGA
No
Yes
Sathyanarayana et al. 2013
2013
USA
1997–2005
Retrospective cohort study
298,835
10.2%
PM 2.5
Trimesters
SGA
Yes
Yes
Salihu, et al. 2012
2012
USA
2000–2007
Retrospective cohort study
12,356
12.9%
PM 2.5, PM 10
Entire pregnancy
SGA
No
Yes
van den Hooven, et al. 2012
2012
Netherlands
2001–2006
Prospective cohort study
6,997
5.5%
PM 10
Entire pregnancy
SGA
No
Yes
Le, et al. 2012
2012
Michigan, USA
1990–2001
Retrospective cohort study
24,096
8.3%
PM 10
Trimesters
SGA
No
Yes
Madsen, et al. 2010
2010
Norway
1999–2002
Retrospective cohort study
25,229
9.6%
PM 2.5, PM 10
Entire pregnancy
SGA
No
Yes
Rich, et al. 2009
2009
New Jersey, USA
1999–2003
Retrospective cohort study
88,678
8.2%; 2.4%
PM 2.5
Trimesters
SGA; VSGA
Yes
Yes
Brauer, et al. 2008
2008
Canada
1999–2002
Retrospective cohort study
70,249
9.0%
PM 2.5, PM 10
Entire pregnancy
SGA
Yes
Yes
Liu, et al. 2007
2007
Canada
1985–2000
Retrospective cohort study
386,202
10.9%
PM 2.5
Trimesters
IUGR
No
No
Hansen, et al. 2007
2007
Australia
2000–2003
Retrospective cohort study
26,617
7.1%
PM 10
Trimesters
SGA
Yes
No
Kim, et al. 2007
2007
Seoul, Korea
2001–2004
Prospective cohort study
1,514
0.9%
PM 10
Trimesters
IUGR
No
No
Parker, et al. 2005
2005
California, USA
2000
Retrospective cohort study
18,247
8.4%
PM 2.5
Entire pregnancy
SGA
No
No
Mannes, et al. 2005
2005
Australia
1998–2000
Retrospective cohort study
138,056
9.7%
PM 2.5, PM 10
Trimesters
SGA
Yes
Yes
Salam, et al. 2005
2005
California, USA
1975–1987
Retrospective cohort study
3,901
15.0%
PM 10
Entire pregnancy, trimesters
IUGR
Yes
Yes
Kim, et al. 2017
2017
South Asia (pooled)
2005–2014
Cross-sectional study
18,586
38.4%
Use of solid fuel
Post-pregnancy (6–23 months)
HAZ
Yes
No
Machisa, et al. 2013
2013
Swaziland
2006–2007
Cross-sectional study
1,150
27.6%
Use of biomass fuel
Post-pregnancy (6–36 months)
HAZ
Yes
No
Tielsch, et al. 2009
2009
India
1998–2000
Prospective cohort study
10,437
89.4%; 57.5%
Use of wood/dung fuel
Post-pregnancy (0–6 months)
HAZ; SGA
No
Yes
Kyu, et al. 2009
2009
Cambodia, DR, Haiti, Jordan, Moldova, Namibia, Nepal
2005–2007
Cross-sectional study
28,439
14.8%; 8%
Exposure to biofuel smoke
Post-pregnancy (0–59 months)
HAZ
Yes
No
Fenske, et al. 2013
2013
India
2005–2006
Cross-sectional study
12,176
36.8%; 17.1%
Use of diff cooking fuels
Post-pregnancy (0–24 mouths)
HAZ
No
No

Conversely, most studies on household air pollution included in the meta-analysis were from areas where a high proportion of the population uses solid fuels: Asia (n = 4), followed by South America (n = 1) and Africa (n = 1). One study contained data from seven developing countries across the globe [33]. The number of participants from individual studies ranged from 202 to nearly 28,500, and the average percentage of population using solid fuels in countries from those studies was 62% (range from 21–91%). Overall, covariate adjustments varied considerably though key adjustments included maternal age, maternal education, parity and infant sex. Thirteen studies adjusted for socioeconomic status and 19 studies adjusted for tobacco use.

Pooled estimates of ambient air pollution

A summary of the results from adverse pregnancy outcomes is shown in Table 2. Figure 2 shows forest plots of the pooled OR estimates for SGA with a 10 µg/m3 increase in PM2.5 based on the whole pregnancy period and/or one of the three gestational periods (the first, second, or third trimesters). The estimates are sorted by the mean exposure levels reported. The pooled ORs indicated significant positive associations between SGA and a 10 µg/m3 increase in PM2.5 during the first (OR = 1.02; 95% CI: 1.00–1.04) and second trimesters (OR = 1.04; 95% CI: 1.01–1.07). Elevated pooled odds ratios were also observed for the whole pregnancy period (OR = 1.04; 95% CI: 1.00–1.10) and third trimester (OR = 1.02; 95% CI: 0.98–1.07), though the associations were not statistically significant at the alpha level of 0.05. A similar pattern of association was observed for stunting risks of high versus low quartiles of PM2.5 exposure during the whole pregnancy (Table 2). Likewise, significant elevated odds of SGA associated with a 10 µg/m3 increase in PM10 for the first trimester (OR = 1.03; 95% CI: 1.00–1.06) was observed (S1 Figure), as well as risk associated with high versus low quartile of PM10 exposure during the whole pregnancy (OR = 1.10; 95% CI: 1.02–1.18).

Table 2
Summary of combined effect estimates on stunting associated with particulate exposures
Outcome and PM type
Exposure measure
Trimester
No. of studies
Test of associations (random effect)
Test of heterogeneity
Publication bias
OR [95% CI]
p-value
I2 (%)
p-value
Begg’s test (p-value)
Egger’s test (p-value)
Ambient air pollution
SGA and PM2.5
Continuous
(per 10 µg/m3)
Entire pregnancy
10
1.044 [0.986, 1.104]
0.1372
92%
< 0.0001
0.3252
0.4056
Trimester 1
9
1.023 [1.003, 1.045]
0.0258
81%
< 0.0001
0.9999
0.1883
Trimester 2
8
1.041 [1.011, 1.072]
0.0064
80%
< 0.0001
0.9999
0.7292
Trimester 3
8
1.025 [0.980. 1.073]
0.2797
90%
< 0.0001
0.6207
0.2550
Categorical
(high vs low)
Entire pregnancy
4
1.048 [0.982, 1.119]
0.1565
0.0%
0.7702
0.4969
0.6590
SGA and PM10
Continuous
(per 10 µg/m3)
Entire pregnancy
7
1.041 [0.993, 1.092]
0.0961
87%
< 0.0001
0.0985
0.2984
Trimester 1
8
1.029 [1.001, 1.058]
0.0433
77%
< 0.0001
0.4579
0.4007
Trimester 2
7
1.029 [0.987, 1.072]
0.1770
80%
< 0.0001
0.2931
0.8763
Trimester 3
6
1.012 [0.981, 1.044]
0.4582
60%
< 0.0001
0.5730
0.3575
Categorical
(high vs low)
Entire pregnancy
5
1.097 [1.018, 1.181]
0.0146
35%
0.1854
0.2672
0.3272
Household air pollution
HAZ < − 2SD*
Use of solid fuel
(high vs. low)
Post pregnancy
5
1.129 [0.969, 1.314]
0.1202
63%
0.0297
0.9999
0.8769
HAZ < − 3SD
Post pregnancy
3
1.123 [0.996, 1.265]
0.0548
18%
0.2953
0.6015
0.3265
* SD refers to standard deviation. Bold font indicates statistical significance at alpha = 0.05 level.

Overall, the heterogeneities of all studies were estimated in random-effect models. We observed a high degree of heterogeneity (I2 ≥ 60%) among studies that used continuous measure of PM2.5 and PM10 as exposure metric (e.g. SGA and PM2.5 for entire pregnancy, I2 = 92%, p < 0.001), but not for studies that reported findings for categorical measures of PM2.5 and PM10 (PM2.5, I2 = 0%, p = 0.77; PM10, I2 = 51%, p = 0.10). Our pooled estimates were robust; removing a particular study did not affect the pooled estimates by > 5%, though the significance of the associations may change in some instances (S3 Figure). For SGA and PM2.5 (continuous) exposure during the first trimester, when either Hyder et al. (2014), Liu et al. (2007), Rich et al. (2009) or Qian et al. (2016) was removed, the observed association was no longer formally statistically significant. Similarly, removing the study by Lavigne et al. (2016) also results in insignificant stunting risk. Nonetheless, there was no significant publication bias for both exposures and outcomes according to both Begg's and Egger's tests (p > 0.05).

Pooled estimates of household air pollution

Two pooled estimates associated with household exposure to solid fuels during the entire pregnancy were calculated for both moderate and severe (defined by HAZ < − 2 SD) and severe (HAZ < − 3 SD) stunting (Fig. 3). Pooled OR for moderate and severe stunting from five epidemiologic studies is 1.13 (95% CI: 0.97–1.31; Table 2), with moderate heterogeneity (I2 = 63%; p = 0.03) and no publication bias (p > 0.05 for both Begg's and Egger's tests). For severe stunting only, there was no evidence of statistical heterogeneity (I2 = 18%; p = 0.2953) nor publication bias, and the pooled estimate is 1.12 (95% CI: 1.00–1.27; S2 Figure). Sensitivity analysis showed that removing the study by Kim et al. (2017) results in a statistically significant increase in odds of moderate and severe stunting (S4 Figure).

Discussion

In this comprehensive quantitative analysis, we summarized the most up-to-date evidence of the associations between prenatal and postnatal exposure to ambient and household air pollution on SGA, IUGR and HAZ from 31 epidemiologic studies published on or before May 2019. The pooled effect estimates consistently suggest a moderate, but noteworthy increase in the risk of SGA (including IUGR). A 10 µg/m3 increase in PM2.5 during the first trimester was associated with a 2.3% increase (95% CI: 0.03–4.5%) in risk of SGA, and between a 2% and 11% elevated risk is observed with increased exposure to PM during various gestational exposure periods, including the first and second trimesters (PM2.5) and over the entire pregnancy (PM10). The direction of the associations between household air pollution and moderate and severe stunting (defined as HAZ < − 2 SD) are consistent with that for ambient air pollution and SGA, though the associations are not statistically significant.

We identified three previous systematic reviews linking air pollution and SGA, IUGR, and HAZ [13, 14, 38]. Two reviews focused on the impact of ambient PM2.5 exposure on various adverse pregnancy outcomes that included SGA. Only one review provided pooled effect estimates [13], while the other offered qualitative observations of the evidence [38]. In a 2015 review, Zhu et al. [13] reported ORs of 1.07 (95% CI: 1.05–1.10) and 1.06 (95% CI: 1.02–1.10) for SGA per increase in PM2.5 exposure during first and second trimesters, as well as statistically significant associations for the third trimester and the entire pregnancy. Our pooled OR estimates of 1.02 and 1.04 for SGA per 10 µg/m3 increase in PM2.5 during the first and second trimesters, respectively, are slightly smaller in magnitude than those from previous meta-analyses. One explanation for this discrepancy is that individual epidemiologic studies that were published after the 2015 meta-analysis tend to report effect estimates that are lower in magnitude and statistical significance as compared to those from earlier studies. Several of the newly included studies containing PM2.5 were from cities in China where much higher PM2.5 levels (e.g. 38–71 µg/m3) were reported [36, 39]. This is in sharp contrast to earlier studies that were conducted in cities with much lower pollution levels (e.g., 4–14 µg/m3).

This phenomena of decreasing effect size with increasing PM2.5 concentrations has been previously documented [40], which indicates a non-linear association between concentration and response. Nonetheless, our findings are supported by existing literature on the impact of ambient air pollution on stunting defined as HAZ. A recent study found that prenatal exposure to the 1997 Indonesian forest fires is associated with a 0.41 in HAZ (or 3.4 cm) at age 17, which implies a loss of 4% of average monthly wages for approximately 1 million Indonesian workers born during this period [41]. Another study reported that in Bangladesh, where stunting prevalence is as high as 36%, children with a high quartile of PM2.5 exposure (52–73 µg/m3) had 1.13 times the risk of stunting (HAZ < − 2) than that of children in the lowest quartile of exposure [23].

For household air pollution, we identified five epidemiologic studies for moderate and severe stunting defined by HAZ, including three studies published after a previous meta-analysis [14]. Bruce et al. [14] reported a pooled OR of 1.27 (95% CI: 1.12–1.43) for moderate and severe stunting associated with household air pollution from cooking with solid fuels. Similar to the pooled estimates for ambient air pollution, our pooled effect estimates for household air pollution from cooking with solid fuels are smaller, but nonetheless in the same direction of association. The volume of evidence from both ambient and household particulate exposure suggests a high level of confidence regarding causality between ambient particulate pollution and prenatal determinants of stunting.

There is a strong biological basis for the relationship between air pollution exposure and low birthweight. Kannan et al. [42] reviewed the evidence from existing literature and determined there are potentially five distinct mechanisms at work: oxidative stress, inflammation, coagulation, endothelial function and hemodynamic responses. While precise biological mechanisms connecting air pollution with impaired fetal growth are unknown, it is commonly hypothesized that transplacental and postnatal exposure to particulate matter may result in oxidative stress leading to DNA damage. Induced acute placental and pulmonary inflammation, increased possibility for coagulation, and triggered endothelial dysfunction are also hypothesized biological mechanisms.

Our meta-analysis had several limitations. First, we observed a moderate to high degree of heterogeneity across gestational exposure periods and exposure metrics. Such heterogeneity may be explained by differences in study design methods and exposure assessment, covariate adjustment, study population and/or PM chemical composition that varies by study location. Further studies are needed to explore the independent and joint effects of early life exposures to air pollution and nutrition, and the effect of PM constituents on stunting-related risks. Second, we did not evaluate studies with other exposure periods (e.g., months) due to scarcity of relevant studies. Hence our observed effects on stunting-related risks based on the entire pregnancy period and specific trimester periods may not infer biological significance as for timing over gestation. Moreover, while majority of the included studies evaluated stunting-related outcomes as categorical measures (e.g., HAZ < − 2 SD), it is important to note that growth faltering is a gradient, and children above the conventional cut-off points for SGA/IUGR/HAZ may still experience suboptimal linear growth, especially in low-resource settings [7, 12], and thus the actual impact of air pollution on stunting or growth failure may be underestimated. Last but not least, we did not assess the quality of each included study.

These limitations are balanced by the substantial strengths of our study. While much of the material in this review has been published previously, additional value derives from updating existing reviews (seven versus 26 for ambient PM pollution in the current review, three versus nine for household air pollution) using comparable methods, including effects from both the continuous and categorical measures of PM2.5 and PM10 exposures. Prior to this review, an effort to join all available childhood stunting-related outcomes, as well as both the ambient particulate pollution exposure and household air pollution from solid fuel, had not been attempted. Our pooled estimates were robust in sensitivity analyses, as demonstrated by using stricter definitions of stunting, or removing one study from the main analysis. There was also no significant publication bias according to Begg’s and Egger’s tests.

Conclusions

Based on a meta-analysis of 31 eligible studies on the association of air pollution and stunting-related outcomes, we found new evidence of increased risk of prenatal determinants of stunting associated with ambient PM2.5 exposure during pregnancy, especially the first trimester, as well as suggestive evidence of elevated stunting risk with postnatal exposure to household air pollution. Eligible studies reflect a much greater range of exposures than those examined in the previous systematic reviews, and also include studies conducted in countries where both adverse pregnancy outcomes and stunting are major public health challenges. Findings suggest the importance of promoting clean air as part of an integrated approach to preventing stunting. Given the critical importance of prenatal determinants of stunting, emerging evidence provides an additional public health imperative to improve air quality. For example, in India, a country with one of the highest stunting prevalence (38%) among children under 5 and the highest SGA prevalence (37%) among live births [43, 44], reducing the annual mean PM2.5 concentration from 91 µg/m3 in 2017 to WHO air quality guideline levels (10 µg/m3) would result in more than 1.8 million or 20% fewer children born SGA in India.

Abbreviations

CI: Confidence interval; HAZ: Height-for-age z score; IUGR: Intrauterine growth retardation; OR: Odds ratio; PM: Particulate matter; PM10: particulate matter with aerodynamic diameter less than 10 μm; PM2.5: Particulate matter less than 2.5 μm in diameter; SD: Standard deviations; SGA: Small for gestational age; WHO: World Health Organization

Declarations

Ethics approval and consent to participate

Not applicable

 

Consent for publication

Not applicable

 

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

 

Competing interests

The authors declare that they have no competing interests.

 

Funding

No funding was received.

 

Authors' contributions

VP conducted literature review, prepared the datasets, interpreted the results and drafted and finalized the manuscript; RD contributed to literature review, dataset preparation, and results interpretation; SM contributed to the conception of the meta-analysis, interpreted the findings and assisted in drafting the manuscript. All authors have read and approved the manuscript.

Acknowledgements

We would like to thank Nahid Tahrima Rashid for her support during the initial phase of this review, and Daniel Kass and Karen Schmidt for their insightful comments and edits on the manuscript. 

References

  1. World Health Organization. Air pollution and child health: prescribing clean air - summary. 2018. https://www.who.int/ceh/publications/Advance-copy-Oct24_18150_Air-Pollution-and-Child-Health-merged-compressed.pdf?ua=1. Accessed 9 Sep 2019.
  2. IHME. GBD Compare. 2019. https://vizhub.healthdata.org/gbd-compare/#. Accessed 8 Jan 2019.
  3. UNICEF. Clear the air for children. 2016. https://www.unicef.org/publications/files/UNICEF_Clear_the_Air_for_Children_30_Oct_2016.pdf. Accessed 9 Sep 2019.
  4. UNICEF. One is too many. Ending child deaths from pneumonia and diarrhea. 2016. http://data.unicef.org/topic/child-health/pneumonia/. Accessed 9 Sep 2019.
  5. Hoddinott J, Behrman JR, Alderman H, Horton S, Haddad L. The economic rationale for investing in stunting reduction. Matern Child Nutr. 2015;9:69–82.
  6. Woldehanna T, Behrman JR, Araya MW. The effect of early childhood stunting on children’s cognitive achievements: Evidence from young lives Ethiopia. Ethiop J Heal Dev. 2017;31:75–84. http://www.ncbi.nlm.nih.gov/pubmed/29249889. Accessed 9 Sep 2019.
  7. de Onis M, Branca F. Childhood stunting: A global perspective. Maternal and Child Nutrition. 2016;12 Suppl Suppl 1:12–26. doi:10.1111/mcn.12231.
  8. Unicef/ WHO/The World Bank. Levels and Trends in Child malnutrition - Unicef WHO The World Bank Joint Child Malnutrition Estimates, key findings pf the 2019 edition. 2019. doi:10.1016/S0266-6138(96)90067-4.
  9. Victora CG, De Onis M, Hallal PC, Blössner M, Shrimpton R. Worldwide timing of growth faltering: Revisiting implications for interventions. Pediatrics. 2010;125:e473–80. doi:10.1542/peds.2009-1519.
  10. Goudet SM, Griffiths PL, Bogin BA, Madise NJ. Nutritional interventions for preventing stunting in children (0 to 5 years) living in urban slums. Cochrane Database Syst Rev. 2015;2015. doi:10.1002/14651858.CD011695.
  11. World Health Organization. World Health Assembly Global Nutrition Targets 2025: Stunting policy brief. 2013. https://apps.who.int/iris/bitstream/handle/10665/149019/WHO_NMH_NHD_14.3_eng.pdf?ua=1.
  12. Prendergast AJ, Humphrey JH. The stunting syndrome in developing countries. Paediatr Int Child Health. 2014;34:250–65.
  13. Zhu X, Liu Y, Chen Y, Yao C, Che Z, Cao J. Maternal exposure to fine particulate matter (PM2.5) and pregnancy outcomes: a meta-analysis. Environ Sci Pollut Res. 2015;22:3383–96.
  14. Bruce NG, Dherani MK, Das JK, Balakrishnan K, AdairRohani H, Bhutta ZA, et al. Control of household air pollution for child survival: estimates for intervention impacts. BMC Public Health. 2013;13:S8. http://0-ovidsp.ovid.com.wam.city.ac.uk/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed11&AN=24564764%5Cnhttp://openurl.ac.uk/ukfed:city.ac.uk/?sid=OVID:embase&id=pmid:24564764&id=doi:10.1186/1471-2458-13-S3-S8&issn=1471-2458&isbn=&volume=13&issue=&spa.
  15. Christian P, Lee SE, Donahue Angel M, Adair LS, Arifeen SE, Ashorn P, et al. Risk of childhood undernutrition related to small-for-gestational age and preterm birth in low- and middle-income countries. Int J Epidemiol. 2013;42:1340–55. doi:10.1093/ije/dyt109.
  16. Xie C, Epstein LH, Eiden RD, Shenassa ED, Li X, Liao Y, et al. Stunting at 5 years among SGA newborns. Pediatrics. 2016;137.
  17. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P, Altman D, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg. 2010;8:336–41.
  18. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. J Am Med Assoc. 1998;280:1690–1. doi:10.1001/jama.280.19.1690.
  19. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–58. doi:10.1002/sim.1186.
  20. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50:1088–101. http://www.ncbi.nlm.nih.gov/pubmed/7786990.
  21. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–34. http://www.ncbi.nlm.nih.gov/pubmed/9310563%0Ahttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC2127453.
  22. Rich DQ, Demissie K, Lu S-E, Kamat L, Wartenberg D, Rhoads GG. Ambient air pollutant concentrations during pregnancy and the risk of fetal growth restriction. J Epidemiol Community Heal. 2009;63:488–96. doi:10.1136/jech.2008.082792.
  23. Goyal N, Canning D. Exposure to ambient fine particulate air pollution in utero as a risk factor for child stunting in Bangladesh. Int J Environ Res Public Health. 2018;15.
  24. Nobles CJ, Grantz KL, Liu D, Williams A, Ouidir M, Seeni I, et al. Ambient air pollution and fetal growth restriction: Physician diagnosis of fetal growth restriction versus population-based small-for-gestational age. Sci Total Environ. 2019;650:2641–7. doi:10.1016/j.scitotenv.2018.09.362.
  25. Bobak M. Outdoor air pollution, low birth weight, and prematurity. Environ Health Perspect. 2000;108:173–6.
  26. Michikawa T, Morokuma S, Fukushima K, Kato K, Nitta H, Yamazaki S. Maternal exposure to air pollutants during the first trimester and foetal growth in Japanese term infants. Environ Pollut. 2017;230:387–93. doi:10.1016/j.envpol.2017.06.069.
  27. Candela S, Ranzi A, Bonvicini L, Baldacchini F, Marzaroli P, Evangelista A, et al. Air pollution from incinerators and reproductive outcomes: A multisite study. Epidemiology. 2013;24:863–70.
  28. Dejmek J, Solanský I, Beneš I, Leníček J, Šrám RJ. The impact of polycyclic aromatic hydrocarbons and fine particles on pregnancy outcome. Environ Health Perspect. 2000;108:1159–64.
  29. Rossner PR, Tabashidze N, Dostal M, Novakova Z, Chvatalova I, Spatova M, et al. Genetic, biochemical, and environmental factors associated with pregnancy outcomes in newborns from the Czech Republic. Environ Health Perspect. 2011;119:265–71.
  30. Rogers JF, Dunlop AL. Air pollution and very low birth weight infants: A target population? Pediatrics. 2006;118:156–64.
  31. Saxton J, Rath S, Nair N, Gope R, Mahapatra R, Tripathy P, et al. Handwashing, sanitation and family planning practices are the strongest underlying determinants of child stunting in rural indigenous communities of Jharkhand and Odisha, Eastern India: a cross-sectional study. Matern Child Nutr. 2016;12:869–84.
  32. Ghosh R, Amirian E, Dostal M, Sram RJ, Hertz-Picciotto I. Indoor coal use and early childhood growth. Arch Pediatr Adolesc Med. 2011;165:492–7.
  33. Kyu HH, Georgiades K, Boyle M. Maternal smoking, biofuel smoke exposure and child height-for-age in seven developing countries. Int J Epidemiol. 2009;38:1342–50.
  34. Hyder A, Lee HJ, Ebisu K, Koutrakis P, Belanger K, Bell ML. PM2.5 Exposure and Birth Outcomes. Epidemiology. 2014;25:58–67. doi:10.1097/EDE.0000000000000027.
  35. Liu S, Krewski D, Shi Y, Chen Y, Burnett RT. Association between maternal exposure to ambient air pollutants during pregnancy and fetal growth restriction. J Expo Sci Environ Epidemiol. 2007;17:426–32.
  36. Qian Z, Zhang B, Liang S, Wang J, Yang S, Hu K, et al. Ambient Air Pollution and Adverse Pregnancy Outcomes in Wuhan, China. Res Rep Health Eff Inst. 2016;:1–65.
  37. Lavigne E, Yasseen AS, Stieb DM, Hystad P, van Donkelaar A, Martin R V., et al. Ambient air pollution and adverse birth outcomes: Differences by maternal comorbidities. Environ Res. 2016;148:457–66. doi:10.1016/j.envres.2016.04.026.
  38. Yuan L, Zhang Y, Gao Y, Tian Y. Maternal fine particulate matter (PM2.5) exposure and adverse birth outcomes: an updated systematic review based on cohort studies. Environ Sci Pollut Res. 2019;26:13963–83. doi:10.1007/s11356-019-04644-x.
  39. Wang Q, Benmarhnia T, Li C, Knibbs LD, Bao J, Ren M, et al. Seasonal analyses of the association between prenatal ambient air pollution exposure and birth weight for gestational age in Guangzhou, China. Sci Total Environ. 2019;649:526–34. doi:10.1016/j.scitotenv.2018.08.303.
  40. Vodonos A, Awad YA, Schwartz J. The concentration-response between long-term PM 2.5 exposure and mortality; A meta-regression approach. Environ Res. 2018;166 August:677–89. doi:10.1016/j.envres.2018.06.021.
  41. Tan-Soo J-S, Pattanayak SK. Seeking natural capital projects: Forest fires, haze, and early-life exposure in Indonesia. Proc Natl Acad Sci. 2019;116:201802876. doi:10.1073/pnas.1802876116.
  42. Kannan S, Misra DP, Dvonch JT, Krishnakumar A. Exposures to airbone particulate matter and adverse perinatal outcomes: A biologically plausible mechanistic framework for exploring potential effect modification by nutrition. Environ Health Perspect. 2006;114:1636–42.
  43. Lee ACC, Kozuki N, Cousens S, Stevens GA, Silveira MF, Sania A, et al. Estimates of burden and consequences of infants born small for gestational age in low and middle income countries with INTERGROWTH-21st standard : analysis of CHERG datasets. 2017;:1–11.
  44. UNICEF/WHO/World Bank. Expanded global databases - stunting. 2019. https://data.unicef.org/topic/nutrition/malnutrition/.