Proximity to Methyl Bromide Use and Birth Outcomes
Proximity to Methyl Bromide Use and Birth Outcomes
Most of the women in the study were Latina (96%) and born in Mexico (85%) (Table 1). About half were recent immigrants to the United States, having lived in the country for ≤ 5 years. Most mothers lived in families with incomes < 200% of the federal poverty level (96%), and 78% had not graduated from high school. The median age was 25 years (SD = 5), and few women smoked during pregnancy (6%); consumption of ≥ 1 alcoholic drink per week (1%) and illegal drug use (2%) were also low (data not shown). more than half of the sample was either overweight (40%) or obese (22%). Approximately 4% of deliveries were born with low birth weight (< 2,500 g) and 7% were preterm (< 37 weeks gestational age). Characteristics of women in each trimester subsample were similar to those of the total sample [see Supplemental Material, Table S1 (http://dx.doi.org/10.1289/ehp.1205682)].
Table 2 shows the distribution of methyl bromide use for each trimester for the buffer distance of interest (5 km) and the three additional buffer distances used in sensitivity analyses (1, 3, and 8 km). The range of methyl bromide use near residences was similar for trimesters 2 and 3 for each buffer distance, whereas values for trimester 1 were consistently larger due to the presence of 12 outliers (defined as > Q3 + (1.5 × IQR), where Q3 is the value of the 75th percentile (or 3rd quartile) and IQR is the interquartile range). Not surprisingly, the proportion of women living near methyl bromide applications increased as the buffer distance increased. For example, 85% of women lived within 8 km of some amount of methyl bromide use during the first trimester compared to 78%, 60%, and 16% of women for the 5-km, 3-km, and 1-km distances respectively. For the primary buffer distance used in this analysis (5 km), the estimated median amounts of methyl bromide use ranged from 6,010 to 8,201 kg, and the 95th percentile values ranged from 53,685 to 68,893 kg.
Table 3 presents estimated associations between methyl bromide use within 5 km of the home as a continuous variable (log10) and the four birth outcomes of interest. In minimally adjusted analyses, increases in methyl bromide use near residences in the second trimester were associated with decreases in birth weight, birth length, and head circumference. After adjusting for confounders, associations remained; each 10-fold increase of methyl bromide use in the second trimester was associated with a 21.4-g (95% CI: –43.2, 0.4) decrease in birth weight, a 0.16-cm (95% CI: –0.28, –0.04) decrease in birth length, and a 0.08-cm (95% CI: –0.15, –0.01) decrease in head circumference. Model estimates did not support associations between methyl bromide use in the first or third trimesters and fetal growth. Residential proximity to methyl bromide use was also positively associated with gestational age in the first trimester in both crude and adjusted analysis (adjusted β = 0.13 week; 95% CI: 0.03, 0.22).
Proximity to methyl bromide use was also analyzed as a three-level categorical variable (Table 4). The findings are similar to those from the continuous variable analysis. Using the 5-km buffer, moderate use of methyl bromide (vs. no use) nearby during the first trimester was associated with a 0.70-week increase in gestational age (95% CI: 0.21, 1.18), and high use was associated with a 0.59-week increase (95% CI: 0.12, 1.06). Moderate and high methyl bromide use within 5 km of the home during the second trimester were negatively associated with birth weight (β = –93.1 g; 95% CI: –198.0, 11.7 and β = –113.1 g; 95% CI: –218.1, –8.1, respectively) and birth length (β = –0.62 cm; 95% CI: –1.20, –0.03 and β = –0.85 cm; 95% CI: –1.44, –0.27, respectively), with stronger associations with high versus moderate use. In addition, proximity to moderate or high use in the second trimester was negatively associated with head circumference (β = –0.42 cm; 95% CI: –0.76, –0.08 and β = –0.33 cm; 95% CI: –0.67, 0.01, respectively). There was little evidence of associations between proximity to methyl bromide use during the first and third trimesters and birth weight, length, or head circumference.
Sensitivity analyses were conducted using different buffer distances around the maternal residence. Results for second trimester methyl bromide use (any vs. none) are presented in Table 5 using 1-km, 3-km, 5-km, and 8-km distances. Proximity to methyl bromide use within 1 km and 3 km during the second trimester was negatively associated with body length. Additionally, the 3-km distance yields marginally significant associations with reduced birth weight (p = 0.06), reduced head circumference (p = 0.08), and increased gestational age (p = 0.05). Unlike the smaller buffer distances, the 8-km analysis was more consistent with the 5-km buffer analysis; proximity to methyl bromide use during the second trimester was negatively associated with birth weight, length, and head circumference, and associations were stronger than estimates from the 5-km analysis. Sensitivity analyses with other trimesters yielded few statistically significant associations: In the 3-km analysis there was a significant positive association with first trimester exposure and gestational age [see Supplemental Material, Table S2 (http://dx.doi.org/10.1289/ehp.1205682)].
Results
Most of the women in the study were Latina (96%) and born in Mexico (85%) (Table 1). About half were recent immigrants to the United States, having lived in the country for ≤ 5 years. Most mothers lived in families with incomes < 200% of the federal poverty level (96%), and 78% had not graduated from high school. The median age was 25 years (SD = 5), and few women smoked during pregnancy (6%); consumption of ≥ 1 alcoholic drink per week (1%) and illegal drug use (2%) were also low (data not shown). more than half of the sample was either overweight (40%) or obese (22%). Approximately 4% of deliveries were born with low birth weight (< 2,500 g) and 7% were preterm (< 37 weeks gestational age). Characteristics of women in each trimester subsample were similar to those of the total sample [see Supplemental Material, Table S1 (http://dx.doi.org/10.1289/ehp.1205682)].
Table 2 shows the distribution of methyl bromide use for each trimester for the buffer distance of interest (5 km) and the three additional buffer distances used in sensitivity analyses (1, 3, and 8 km). The range of methyl bromide use near residences was similar for trimesters 2 and 3 for each buffer distance, whereas values for trimester 1 were consistently larger due to the presence of 12 outliers (defined as > Q3 + (1.5 × IQR), where Q3 is the value of the 75th percentile (or 3rd quartile) and IQR is the interquartile range). Not surprisingly, the proportion of women living near methyl bromide applications increased as the buffer distance increased. For example, 85% of women lived within 8 km of some amount of methyl bromide use during the first trimester compared to 78%, 60%, and 16% of women for the 5-km, 3-km, and 1-km distances respectively. For the primary buffer distance used in this analysis (5 km), the estimated median amounts of methyl bromide use ranged from 6,010 to 8,201 kg, and the 95th percentile values ranged from 53,685 to 68,893 kg.
Table 3 presents estimated associations between methyl bromide use within 5 km of the home as a continuous variable (log10) and the four birth outcomes of interest. In minimally adjusted analyses, increases in methyl bromide use near residences in the second trimester were associated with decreases in birth weight, birth length, and head circumference. After adjusting for confounders, associations remained; each 10-fold increase of methyl bromide use in the second trimester was associated with a 21.4-g (95% CI: –43.2, 0.4) decrease in birth weight, a 0.16-cm (95% CI: –0.28, –0.04) decrease in birth length, and a 0.08-cm (95% CI: –0.15, –0.01) decrease in head circumference. Model estimates did not support associations between methyl bromide use in the first or third trimesters and fetal growth. Residential proximity to methyl bromide use was also positively associated with gestational age in the first trimester in both crude and adjusted analysis (adjusted β = 0.13 week; 95% CI: 0.03, 0.22).
Proximity to methyl bromide use was also analyzed as a three-level categorical variable (Table 4). The findings are similar to those from the continuous variable analysis. Using the 5-km buffer, moderate use of methyl bromide (vs. no use) nearby during the first trimester was associated with a 0.70-week increase in gestational age (95% CI: 0.21, 1.18), and high use was associated with a 0.59-week increase (95% CI: 0.12, 1.06). Moderate and high methyl bromide use within 5 km of the home during the second trimester were negatively associated with birth weight (β = –93.1 g; 95% CI: –198.0, 11.7 and β = –113.1 g; 95% CI: –218.1, –8.1, respectively) and birth length (β = –0.62 cm; 95% CI: –1.20, –0.03 and β = –0.85 cm; 95% CI: –1.44, –0.27, respectively), with stronger associations with high versus moderate use. In addition, proximity to moderate or high use in the second trimester was negatively associated with head circumference (β = –0.42 cm; 95% CI: –0.76, –0.08 and β = –0.33 cm; 95% CI: –0.67, 0.01, respectively). There was little evidence of associations between proximity to methyl bromide use during the first and third trimesters and birth weight, length, or head circumference.
Sensitivity analyses were conducted using different buffer distances around the maternal residence. Results for second trimester methyl bromide use (any vs. none) are presented in Table 5 using 1-km, 3-km, 5-km, and 8-km distances. Proximity to methyl bromide use within 1 km and 3 km during the second trimester was negatively associated with body length. Additionally, the 3-km distance yields marginally significant associations with reduced birth weight (p = 0.06), reduced head circumference (p = 0.08), and increased gestational age (p = 0.05). Unlike the smaller buffer distances, the 8-km analysis was more consistent with the 5-km buffer analysis; proximity to methyl bromide use during the second trimester was negatively associated with birth weight, length, and head circumference, and associations were stronger than estimates from the 5-km analysis. Sensitivity analyses with other trimesters yielded few statistically significant associations: In the 3-km analysis there was a significant positive association with first trimester exposure and gestational age [see Supplemental Material, Table S2 (http://dx.doi.org/10.1289/ehp.1205682)].
Source...