The Impact of Prenatal Alcohol Exposure on Longitudinal Growth, Nutritional Status, and IGF-1 in Early Childhood in Leyte, The Philippines

Abstract

Objective

To assess the impact and potential mechanistic pathways of prenatal alcohol exposure (PAE) on longitudinal growth and nutritional status in early childhood.

Study design

A cohort of 296 mother-infant dyads (32% with PAE vs. 68% unexposed) were recruited in Leyte, the Philippines, and followed from early gestation through 24 months of age. PAE was assessed using serum phosphatidylethanol (PEth) captured twice prenatally and in cord blood and supplemented with self-reported alcohol consumption. Linear mixed models were used to examine longitudinal effects of PAE on growth from birth through 2 years including key potential mediating factors (placental histopathology, and infant serum leptin and IGF-1).

Results

After adjusting for potential confounders, we found that PAE was significantly associated with a delayed blunting of linear growth trajectories (height-for-age z-score, body length) and weight (weight-for-age z-score, body weight) that manifested between 4 and 6 months and continued through 12 to 24 months. PAE was also associated with a decreased rate of mid-upper-arm circumference growth from birth to 12 months, and a lower mean IGF-1 levels at birth and 6 months.

Conclusion

This study demonstrates a delayed impact of PAE on growth that manifested around 6 months of age, underscoring the importance of routine clinical monitoring in early childhood. Furthermore, the findings supported prior animal model findings that suggest a mechanistic role for IGF-1 in PAE-induced growth delay.

Fetal alcohol spectrum disorder (FASD) comprises a wide range of preventable adverse effects on child development. Most current FASD research focuses specifically on birth outcomes and perinatal anthropometry (1–4), small for gestational age (SGA), low birth weight (LBW), and prematurity (3,5,6), timepoints at which only the most severely affected children may demonstrate apparent growth or developmental delay. However, research has found growth trajectory may serve as an important prognostic factor for the severity of neurodevelopmental deficits in FASD (7,8). Many clinicians now propose a combined approach that incorporates facial dysmorphology, neurocognitive indicators, and growth patterns to facilitate a more accurate and timely diagnosis by 18–24 months as a delayed diagnosis may miss the critical window for early intervention (9).

Furthermore, the biological mechanisms for the effects of prenatal alcohol on growth and development are not well characterized, though research suggests dysregulation of the Insulin-like Growth Factor-1 (IGF-1) signaling pathway given its key role in both pre-and postnatal growth and neurodevelopment (10–15). IGF-1 is critical for neurodevelopment (11,15–23), stimulates growth independent of pituitary growth hormone, and is the predominant driver of linear growth in early childhood (21). Several animal and in vitro models have implicated IGF signaling disruptions in alcohol-related growth delay (24–28), but this pathway is understudied in humans.

Notable research gaps exist to understand PAE and postnatal growth delay, particularly in low-and-middle-income (LMIC) countries, each with unique competing risks for adverse child development. Our prior studies in this Filipino cohort revealed widespread prenatal consumption of “tuba”, a local coconut wine containing 7–8% ethanol by mass spectroscopy (29). Using an objective biomarker of prenatal ethanol exposure and routine monitoring of a cohort of rural Filipino children from birth through 2 years, we sought to better categorize growth trajectories and examine potential mediating roles of fetal/infant IGF-1 and leptin on impaired growth in the context of PAE.

METHODS

This prospective cohort includes 296 mother-infant pairs who returned for follow-up after delivery, with 76% retention (225/296) through 24-months in April 2021. Pregnant women aged 18 or older and with a singleton pregnancy at <20 weeks’ gestation were recruited by community health workers in Leyte, The Philippines. Women with significant maternal illness at enrollment determined by history/ physical exam by a study obstetrician or ultrasound evidence of major congenital anomaly were excluded. Iron supplementation was provided to all women. All participants with helminth infection were offered, but all declined, anthelmintic treatment.

Biomarker Sample Collection and Processing

Maternal serum was collected at enrollment and 32 weeks’ gestation and stored at −80°C. At delivery, placentas were weighed after cleaning and trimming umbilical cords to 10 cm. Umbilical artery blood was sampled for newborn IGF-1 and leptin levels using ELISA (30), and 6- and 12-month serum was analyzed via a multiplexed sandwich antibody-based assay(31).

Phosphatidylethanol (PEth), a direct biomarker of alcohol metabolism with a detection window of 3 to 4 weeks (32) was collected on dried blood spot (DBS) cards at 3 time points in gestation, stored at 4°C, and sent to United States Drug and Testing Laboratories for quantifying (lower detection limit = 8 ng/dL) (33–36). Weekly alcohol intake was assessed via self-report at each pre- and postnatal visit (0, 1–2, 3–5, 6–8, 9–10 or 11+ drinks) and 3+ drinks/week was significantly associated with detectable PEth levels at corresponding visits (X², p=0.001)

We used a tiered approach to capture Prenatal Alcohol Exposure (PAE) that included serum PEth levels and self-reported drinking frequency when PEth was missing (16% at enrollment, 11% at later gestation). PEth values of ≥ 20.0 ng/mL indicated “high” exposure based on current recommendations (37). PAE included any of the following criteria: (1) any “high” PEth ≥ 20.0 ng/mL during gestation; (2) a detectable PEth (≥ 8.0 ng/mL) at BOTH enrollment and later gestation (32wks or delivery); (3) a detectable PEth (≥ 8.0 ng/mL) at either enrollment and/or late gestation AND self-reported 3+ drinks/wk at BOTH enrollment and 32-week visit; and, (4) if missing all PEth data, self-reported of 3+ drinks/wk at BOTH enrollment and 32-week visits.

We measured infant anthropometry at birth, 1, 6, 12, 18, and 24 months of age using a length board and supine scale for body length (cm) and weight (kg), respectively. We used ultrasound to estimate gestational age (GA) and prematurity was defined as birth at <37 weeks of gestation. Low birth weight (LBW) was defined as ≤ 2.50 kg, and small-for-gestational-age (SGA) was defined as size <10^th percentile using sex and gestational age adjusted birth percentiles using INTERGROWTH-21^st healthy reference standards (38). We measured mid-upper arm circumference (MUAC) using a non-stretch, “Zerfuss” insertion tape to 0.1 cm. For repeat growth metrics, age- and sex-standardized z-scores for length (HAZ) and weight (WAZ) were derived from 2006 WHO Anthro Growth Standards (39).

Mediating and Confounding Covariates

Potential confounders of growth were considered a priori (40–42). At enrollment and 32 weeks of gestation, maternal age, nutritional status/anthropometry (height, weight, MUAC, BMI, thigh circumference and skin fold thickness), smoking status, and obstetric history/parity were recorded. Anemia was defined as hemoglobin ≤11.0 g/dL using a Horiba ABX hematology analyzer (Horiba Ltd.) (43). Maternal infection (yes/no) included evidence of parasitic infection by the Kato-Katz method (44) with S. japonicum, A. lumbricoides, T. trichiura, or hookworm. Maternal vascular malperfusion (MVM) (45) defined by Redline et. al.(46), and parity, were included as potential mediators of the ethanol-fetal IGF-1 relationship. Breastmilk protein concentration (mg/mL) and infant feeding practices were captured at 6 and 12 months postnatally. Socioeconomic status (SES) was determined using a previously validated tool from this study population (29,47–49). We used principal components analysis to assign numeric weights to SES indicators from 4 domains described elsewhere (50); a weighted combination of indicators was dichotomized (≥ median). Highest maternal education was dichotomized as high school or below.

Statistical Analysis

We summarized continuous characteristics using mean/standard deviation or median/IQR where appropriate, categorical variables using relative frequencies n (%), and compared distributions between exposure groups using two sample t-tests (Wilcoxon rank sum) or chi-square tests. We fit linear mixed models to evaluate the longitudinal impact of PAE on infant anthropometry and IGF-1 levels over time using a random intercept to allow for subject-specific baseline values, random effects for age and time-varying covariates, and assumed an unstructured covariance to account for intrasubject heterogeneity. To improve model fit, we used a likelihood-based stepwise selection procedure (51) to include the best subset of covariates, time polynomials, and two-way interaction terms. For comparison, we modeled a 12- and 24-month trajectory separately for each outcome using identical covariates, except 12-month models also include IGF-1 and leptin as they were measured at birth, 6-and 12-months only. An additional multiple linear regression model assessed PAE on IGF-1 levels at birth, 6- and 12-months. Margins plots present mean outcomes by ETOH exposure averaging over model covariates at specified 2- or 4-month age intervals. Analyses were done using Stata v.17 (52) and with an α = .05 level of significance. Analysis used complete visit data with little-to-no missing survey or outcome data. Sensitivity analyses determined missingness was non-differential and not associated with exposure or outcome data, including key covariates used in this analysis.

This study was approved by IRBs at Rhode Island Hospital and the Research Institute of Tropical Medicine (Philippines). Midwife-administered surveys at each prenatal visit included education about teratogenic effects of alcohol and advised participants that drinking during pregnancy can irreversibly harm an unborn fetus. All participants provided informed consent.

RESULTS

Maternal characteristics (Table I) and birth outcomes (Table II) were similar between exposure groups (PAE, n=93; unexposed, n=203). Mothers in the PAE group had higher parity (p=0.02) and tended to be less well educated compared with controls (p=0.07), though we observed no notable differences in gestational length, prematurity rates, breastfeeding practices, breastmilk quality (protein) or birth outcomes between groups. Body length, weight, HAZ, WAZ, and infant MUAC were similar between groups from birth through 6-months, diverging between 6- and 12-months, during which time infants with ETOH-exposure experienced a more stunted growth trajectory and remained smaller than unexposed infants through 2 years. Head circumference was similar at all time points. Although leptin levels were similar, serum IGF-1 was greater in unexposed infants at birth (p=0.002) and 6-months (p=0.05).

Preliminary analyses demonstrated no relationship between prenatal or postnatal alcohol exposure and head circumference, breastfeeding practices, or breastmilk quality. Further, breastfeeding practices and breastmilk protein were not associated with infant length or weight at any follow-up visit, and therefore neither covariate was retained in final models. The adverse effect of PAE on longitudinal growth was non-differential regarding sex when examined separately, though males had slightly greater growth rates overall, and therefore sex was retained in all final models.

Effects of Prenatal Ethanol on Growth Trajectories and IGF-1 Levels

Model results and predicted growth plots were adjusted for all indicated model covariates.

Table III and Figure 1(A–D) present effects of PAE on linear growth. Children with ETOH-exposure had a steeper decline in HAZ through 12-months (PAE x age interaction term, ß= − 0.041, 95% CI: −0.072, −0.011) and demonstrated a more negative slope for absolute length (“blunted growth trajectory”) from birth through 12-months (ß = −0.100, 95% CI: −0.166, −0.034) and through 24-months (ß = −0.040, 95% CI: −0.071, −0.008). IGF-1 and maternal height were strongly associated with greater HAZ and body length over time, though infant leptin, maternal MUAC or SES did not influence linear growth.

Table 3:

Linear Mixed Models for Effects of Prenatal Alcohol Exposure on Height-for-Age Z-score (HAZ) and Body Length (cm)

Outcome Model	HAZ to 12-mos		HAZ to 24-mos		Length to 12-mos		Length to 24-mos
	ß	95% CI	ß	95% CI	ß	95% CI	ß	95% CI
PAE	0.168	−0.118 to 0.454	−0.064	−0.281 to 0.153	0.366	−0.183 to 0.915	−0.061	−0.485 to 0.363
Age (months)	0.011	−0.038 to 0.060	−0.050 **	−0.087 to −0.013	3.633 ***	3.257 to 3.738	3.504 ***	3.420 to 3.589
Age2 (months)	−0.004 *	−0.007 to −0.001	−0.004 *	−0.008 to −0.001	−0.134 ***	−0.140 to −0.127	−0.173 ***	−0.181 to −0.164
PAE x Age b	−0.041 **	−0.072 to −0.011	−0.010	−0.021 to 0.002	−0.100 **	−0.166 to −0.034	−0.040 **	−0.071 to −0.008
Sex (male)	−0.206 *	−0.391 to −0.020	−0.230 **	−0.399 to −0.060	0.935 ***	0.533 to 1.337	0.871 ***	0.493 to 1.249
IGF-1 (ng/mL)	0.263 ***	0.144 to 0.381			0.576 ***	0.323 to 0.828
Leptin (pg/mL)a	−0.005	−0.030 to 0.021			0.000	−0.059 to 0.060
Maternal MUAC at enrollment (cm)	0.005	−0.027 to 0.036	0.014	−0.014 to 0.043	0.007	−0.061 to 0.075	0.026	−0.037 to 0.090
Maternal Height (cm)	0.032 **	0.014 to 0.050	0.036 ***	0.019 to 0.052	0.057 **	0.018 to 0.097	0.061 **	0.025 to 0.098
SES (≥ median)	0.084	−0.104 to 0.271	0.123	−0.049 to 0.294	0.153	−0.254 to 0.559	0.235	−0.147 to 0.617
Intercept	−5.572	−8.261 to −2.882	−5.623	−8.081 to −3.165	38.377	32.54 to 44.21	39.794	34.33 to 45.26
N obs	764		1,498		764		1,498

ß-estimates (excluding intercept) represent mean change in outcome per 1 unit change in covariate over time, adjusted for covariates.

ß-estimates and robust standard errors (SE) adjusted for within-subject correlation.

^***p< 0.001,

^**p< 0.01,

^*p< 0.05

^aLeptin, IGF-1 ln transformed;

^bPAExAge interaction represents different rate of change in outcome in ETOH exposed vs Unexposed groups

^dCubic (age³) and/or Quadratic terms (age²) improved model fit for non-linear trend over time in 12-month and 24-month models, respectively;

Note: Random effects for age, age^2, age³ inc. in models with corresponding main effect only; Random effect for IGF-1, Leptin in 12-month models only

Figure 1.

Adjusted linear mixed model marginal plots of predicted mean height estimates over time by exposure group (+/− shaded 95% confidence limits). (A) Predictive Mean Height-for-Age z-score (HAZ) over 12-months, and (B) 24-months. (C) Predictive Mean Body Length (cm) over 12-months, and (D) 24-months.

Table IV and Figure 2 (A–D) include similar effects on weight over time. PAE was associated with a steeper decline in WAZ from birth through 12 months (ß = −0.022, 95% CI: −0.039, −0.005) and 24 months (ß = −0.012, 95% CI: −0.021, −0.003), as well as body weight through 12 months (ß = −0.021, 95% CI: −0.037, −0.005) and 24 months (ß = −0.015, 95% CI: −0.025, −0.005). IGF-1 levels correlated with greater mean WAZ and weight, whereas leptin was associated with mean body weight only. Maternal MUAC, height and SES were associated with greater mean WAZ and/or body weight to varying degrees.

Table 4:

Linear Mixed Model for Effects of PAE on Weight-for-Age Z-score (WAZ) and Body Weight (kg)

Outcome Model	WAZ to 12-mos			WAZ to 24-mos			Weight (kg) to 12-mos			Weight (kg) to 24-mos
	ß	SE	95% CI	ß	SE	95% CI	ß	SE	95% CI	ß	SE	95% CI
PAE	0.047	0.085	−0.199, 0.214	0.018	0.089	−0.156, 0.191	0.030	0.039	−0.046, 0.106	0.033	0.048	−0.061, 0.128
Age (months)	0.004	0.021	−0.037, 0.044	−0.075 ***	0.014	−0.103, −0.048	0.903 ***	0.016	0.873, 0.933	0.857 ***	0.013	0.832, 0.882
Age2 (months)	−0.003 *	0.001	−0.006, −0.001	0.001	0.001	−0.001, 0.004	−0.038 ***	0.001	−0.040, −0.036	−0.048 ***	0.001	−0.050, −0.045
PAE x Age b	−0.022 *	0.009	−0.039, −0.005	−0.012 **	0.005	−0.021, −0.003	−0.021 **	0.008	−0.037, −0.005	−0.015 **	0.005	−0.025, −0.005
Sex (male)	0.052	0.075	−0.094, 0.198	0.021	0.074	−0.124, 0.167	0.165 ***	0.037	0.093, 0.237	0.250 ***	0.045	0.161, 0.339
IGF-1 (ng/mL)	0.144 ***	0.041	0.065, 0.224				0.097 ***	0.027	0.045, 0.149
Leptin (pg/mL)a	0.011	0.008	−0.006, 0.027				0.021 **	0.007	0.006, 0.035
Maternal MUAC at enrollment (cm)	0.047 ***	0.013	0.022, 0.072	0.049 ***	0.012	0.025, 0.074	0.017 **	0.006	0.005, 0.030	0.025 ***	0.008	0.010, 0.040
Maternal Height (cm)	0.016 *	0.007	0.002, 0.030	0.016 *	0.007	0.002, 0.030	0.006	0.004	−0.001, 0.013	0.005	0.004	−0.004, 0.014
SES (≥ median)	0.175 *	0.075	0.029, 0.322	0.118	0.075	−0.028, 0.265	0.081 *	0.036	0.009, 0.152	0.039	0.046	−0.050, 0.129
Intercept	−4.996	1.080	−7.113, −2.878	−4.400	1.072	−6.501, −2.300	0.965	0.534	−0.082, 2.011	1.567	0.657	0.279, 2.855
N obs	764			1,498			764			1,498

ß-estimates (excluding intercept) represent mean change in outcome per 1 unit change in covariate over time, adjusted for covariates.

ß-estimates and standard errors (SE) adjusted for within-subject correlation.

^***p< 0.001,

^**p< 0.01,

^*p< 0.05

^aLeptin, IGF-1 ln transformed;

^bPAExAge interaction represents different rate of change in outcome in ETOH exposed vs Unexposed groups;

^dQuadratic term (age²) improved model fit for non-linear trend over time;

Note: Random effects for age, age² inc. in models with corresponding main effect; Random effect for IGF-1, Leptin in 12-month models only

Figure 2.

Adjusted linear mixed model marginal plots of predicted mean weight estimates over time by exposure group (+/− shaded 95% confidence limits). (A) Predictive Mean Weight-for-Age z-score (WAZ) over 12-months, and (B) 24-months. (C) Predictive Mean Body Weight (kg) over 12-months, and (D) 24-months.

IGF-1 levels were lower at birth and 6months in infants with ETOH-exposure (Table V; online). Although IGF-1 levels declined over time in both groups, there was no difference in IGF-1 rates of decline over 12- months (not shown). Infant MUAC growth rate over the first 12 months was more negative in the ETOH-exposed group (ß = −0.060, 95% CI: −0.108, −0.012) (Table VI; online). Maternal MUAC and SES were associated with greater infant MUAC and/or WHZ to varying degrees, though growth rates for infant MUAC and WHZ were similar between groups.

DISCUSSION

We longitudinally assessed prenatal ethanol exposure on growth trajectories in an LMIC setting with an objective biomarker of ethanol exposure and repeated measures of validated, standardized growth metrics throughout infancy. Our results demonstrated that teratogenic effects of even mild to moderate levels of prenatal ethanol contributed to adverse growth and nutritional status that manifested most apparently around 6-months of age and appeared to be partially explained through disruptions to fetal and infant IGF-1 signaling beginning in utero.

In this prospective cohort, we observed a delayed adverse effect of PAE on growth and nutrition among infants with and without ETOH exposure who had similar birth outcomes. Infants with PAE appeared to “drop off” growth curves between 4 and 8 months and tracked below unexposed counterparts through 24 months. These findings coincide with earlier PAE studies from the 1990’s that were among the first to report a lag in growth rates in the first 8 months of life (53). Even among children without an FASD diagnosis, those exposed to alcohol prenatally demonstrated a diminished growth capacity from gestation through 36 months (54) and 14 years (55) after controlling for nutritional and environmental factors. This delayed effect is supported by another cohort study which found that, even among children with similar birth characteristics, those with the most apparent growth deficiency over time were more likely to present with severe neurocognitive effects at FASD diagnosis (56). Kalberg et al assessed early life predictors of FASD phenotypes at age 5 and found that developmental differences between children with and without a FASD diagnosis were not readily apparent by 9 to 18 months of age (9). Comparable rates of head growth were less surprising as prenatal alcohol often has a less pronounced effect on head circumference (57–59) – termed the “head sparing phenomenon”– implying that tracking other growth parameters such as length or weight may prove more valuable in guiding diagnosis. These findings emphasize the utility of longitudinal growth trajectories as a more sensitive method of risk stratification versus cross-sectional infant anthropometry alone.

We observed both direct and indirect adverse effects of PAE on growth via the downregulation of IGF-1 in early childhood as evidenced by similar conclusions using longitudinal models with and without controlling for IGF-1 at 12 and 24 months, respectively. Our results are consistent with other animal and human studies which link PAE with IGF downregulation (60,61). One human study found that IGF dysregulation in early childhood was associated with adverse growth and neurocognitive scores among 8–12-year-olds (62). These findings lend credence to the hypothesis that developmental delays in FASD are due, in part, to IGF-1 signaling disruption in early life. It is worth noting, however, the multifactorial causes of birth outcomes such as genetics, placental health, and the fetal environment, which this study did not fully account for and was not powered to explain the discrepancy between fetal IGF-1 levels and birth outcomes between groups.

Interestingly, we observed a lag in IGF-1 with low levels in the exposed group at birth that equalized by 12 months. This could be due to toxic in-utero effects of alcohol on the fetal liver as the primary source of fetal IGF-1 (63) which, after birth, no longer contributes to placental insufficiency or direct toxic effects as feeding and environmental factors take over. IGF-1 is critical in many synergistic metabolic pathways, namely insulin metabolism and bone and muscle growth (64), which may partially explain the few months delay in growth lagging behind the decline in IGF-1 levels.

Although breastfeeding and breastmilk quality (protein) was not associated with PAE or infant growth in this cohort, it is possible that the impact of PAE on decreased linear growth and nutrition status in early childhood could be explained by other postnatal maternal factors. Specifically, it is possible that women who drink during pregnancy may also be more likely to consume alcohol postnatally, which could have affected both quantity and quality of breast milk, in turn impacting infant growth (65). In contrast to what many women incorrectly believe based on surveys globally, one study reported that breast milk production was decreased by 9.3% if women drank alcohol within two hours breast feeding(66). Other studies suggest decreased milk intake overall (65), whereas others have not (67). Further, although some studies demonstrate an impact of alcohol use on earlier discontinuation of breast feeding (at 6 months) (68), other studies do not support this (69). A recent study in South Africa sought to understand whether postnatal differences in infant feeding practices, including duration and exclusivity of breastfeeding and introduction of complementary foods, could explain the delayed impact on infant growth. It concluded that fetal alcohol-related postnatal growth restriction was not attributable to differences in postnatal feeding practices or alcohol exposure in breast milk and thus is likely a direct teratogenic effect of PAE (70). With respect to breastmilk quality, some studies have examined the impact of alcohol intake on immunologic factors in breastmilk (71), but few if any have found differences in the nutrient quality of breast milk.

In this population, postnatal factors appeared to have less effect on growth and nutrition for several reasons. First, very few women had high prenatal PEth levels, and smoking and substance use are highly uncommon among women in this population. Though some may drink enough to adversely impact their own nutritional status while breastfeeding, or hinder securing nutritious foods, this is less likely to be the case in this population than in other settings where women have significant alcohol use or other co-morbid substance use disorders. Finally, adjusting for SES likely removes effects of confounding due to nutrition education and food resources, which in most LMIC settings is a significant predictor of child nutritional status.

Our study has several strengths, including a relatively high retention. A notable absence of smoking and drug use in this cohort removes the most common confounders in perinatal research, and out cohort is relatively homogenous both genetically and environmentally, which limits common sources of residual confounding. Robust measures of SES and breastfeeding practice controls for key drivers of infant growth in LMICs, and use of an objective biomarker of alcohol exposure limits exposure misclassification vs. self-report alone. Thus, we are confident these results are likely due to PAE rather than other competing risks.

Our study has some limitations. First, multiple mother-baby dyads missed follow-up visits due to the SARS-CoV-2 pandemic. Also, the utility of PEth is limited by its 4-week window and lower limit of detection, and we presume some participants with low-to-moderate drinking may be misclassified as unexposed. Lastly, our results are limited to a 2-year follow-up, which may not capture FASD phenotypes that remain sub-clinical until later childhood or adolescence.

FASD studies have increasingly focused on identifying sensitive biomarkers to improve timely diagnosis, though these are costly and unavailable in many LMICs. Alternatively, standardized growth curves are widely accessible and represent a more economically feasible alternative in resource-limited settings for early identification of children at risk for FASD. In our cohort and several others, PAE did not predict prematurity or other common birth parameters, suggesting that infants with PAE may be misclassified as “low risk” for FASD if birth parameters alone are used to risk stratify. Lastly, in many LMICs, there is limited public health messaging about the harms of alcohol in pregnancy. Despite education by midwives, we found that alcohol exposure continued throughout gestation and impacted early childhood growth and nutrition, manifesting most apparently between 6 and 12 months of age, with IGF-1 downregulation playing an indirect mechanistic role. Likewise, another study from this cohort addressed knowledge and beliefs about risks of prenatal alcohol and found many were unaware of potential fetal harm and reported continued use in pregnancy (72). Results were shared with the Ministry of Health and midwives in Leyte. Future studies should consider parsing out long-term effects of other growth and nutritional hormones dysregulated by prenatal ethanol exposure, including IGF-1 and IGF-2, to identify modifiable risk factors and targets for early intervention.

Abbreviations

CI

Confidence Interval

FASD

Fetal Alcohol Spectrum Disorder

HAZ

height/length-for-age z-score

IGF-1

Insulin-like Growth Factor 1

LBW

low birth weight

LMIC

low- and middle-income countries

MUAC

middle-upper-arm circumference

PAE

Prenatal Alcohol Exposure

SES

socioeconomic status

SGA

small-for-gestational age

sTfR

soluble transferrin receptor

WAZ

weight-for-age z-score

WHZ

weight-for-height z-score

Data sharing:

Requests for access to the data should be made to the Center for International Health Research at Rhode Island Hospital, cc: Jennifer F. Friedman, MD, PhD (Jennifer_Friedman@brown.edu).