Hair samples for cortisol analysis were provided by 152 participants. Characteristics of the study population are described in Table 1. We excluded 3 participants due to inexplicable high HCC values (> 3 times the interquartile range above the third quartile). As a result, 149 pregnant women were included in this study, of which 133 women donated a sample at the end of the 2nd trimester (week 26 ± 1.6) and 81 women shortly after delivery (week 39 ± 1.6), 65 women donated a sample twice. Almost half of the 149 mothers (48%) was aged between 26 and 30 years, 61% of participants were primigravid. Most of the study participants were of European origin (75%, 21% data missing), enjoyed higher education (57%, 23% data missing) and were employed prior to their pregnancy (72%, 23.5% data missing).
Table 1. Basic characteristics of the study participants
Characteristic
|
n (%)
|
Age
|
|
≤ 25
|
17 (11.4)
|
26-30
|
71 (47.7)
|
31-35
|
44 (29.5)
|
>35
|
17 (11.4)
|
Missing
|
0
|
Parity
|
|
0
|
91 (61.1)
|
1
|
43 (28.9)
|
≥ 2
|
15 (10)
|
Missing
|
0
|
Pre-existing chronic diseases
|
|
No
|
135 (90.6)
|
Yes
|
12 (8.1)
|
Missing
|
2 (1.3)
|
Pre-pregnancy Body Mass Index (kg/m2)
|
|
Underweight (<18.5)
|
8 (5.4)
|
Normal (18.5-24.9)
|
84 (56.4)
|
Overweight (25-29.9)
|
20 (13.4)
|
Obese (≥30)
|
12 (8.1)
|
Missing
|
25 (16.7)
|
Smoking before pregnancy
|
|
Never
|
94 (63.1)
|
Former smoker
|
21 (14.1)
|
Missing
|
34 (22.8)
|
Alcohol consumption before pregnancy
|
No
|
16 (10.7)
|
Yes
|
99 (66.4)
|
Missing
|
34 (22.8)
|
Ethnic background
|
|
European
|
112 (75.2)
|
Non-European
|
4 (2.7)
|
Missing
|
33 (21.1)
|
Educational attainment
|
|
Low (Basic level)
|
15 (10.1)
|
Intermediate (Secondary school)
|
15 (10.1)
|
High (Higher education)
|
85 (57.0)
|
Missing
|
34 (22.8)
|
Pre-pregnancy employment
|
|
No
|
7 (4.7)
|
Yes
|
107 (71.8)
|
Missing
|
35 (23.5)
|
Daily hair washing
|
|
No
|
97 (65.1)
|
Yes
|
18 (12.1)
|
Missing
|
34 (22.8)
|
Season of 2nd trimester sampling (n=133)
|
Autumn
|
28 (21.2)
|
Winter
|
23 (17.3)
|
Spring
|
38 (28.6)
|
Summer
|
44 (33.1)
|
Missing
|
0
|
Season of 3rd trimester sampling (n=81)
|
|
Autumn
|
23 (28.4)
|
Winter
|
21 (25.9)
|
Spring
|
12 (14.8)
|
Summer
|
25 (30.9)
|
Missing
|
0
|
Note: pre-existing chronic diseases include diabetes, asthma, cardiovascular diseases
Residential characteristics are described in Table 2. Study participants lived in cities (38%), towns and suburbs (62%) in Flanders, none of the participants lived in a rural area. The mean ADI of our study population was 16.4% (95% CI: 14.6, 18.1) whereas the mean 2017 ADI for the study region Antwerp was 17.6% (Statistics Flanders, n.d.). We tested the significance of the association between ADI as an area-level SES indicator and maternal educational attainment as a personal SES-indicator. We did not observe a significant association between neighborhood SES and personal SES (Spearman rank r = -0.074, p = 0.404). A small neighborhood greenspace was accessible for 94% of participants, 76% had residential access to a large greenspace. Three-month geometric mean PM2.5 was 11.61 (95% CI: 11.06, 12.21) µg/m3 and 11.55 (95% CI: 10.95, 12.18) µg/m3 for 2nd trimester and 3rd trimester sampling respectively. Geometric mean NO2 concentrations 3 months before sampling was 23.03 (95% CI: 21.67, 24.47) µg/m3 for the 2nd trimester and 23.19 (95% CI: 21.5, 24.98) µg/m3 for the 3rd trimester, 3-month geometric mean BC concentration 1.13 (95% CI: 1.05, 1.21) µg/m3 for the 2nd trimester and 1.17 (95% CI: 1.07, 1.28) µg/m3 for the 3rd trimester. Noise exposure was covered by the Flemish strategic noise map for 144 participants, 31.5% of participating mothers was exposed to noise levels ≥ 53 dB. Geometric mean 2nd trimester HCC was 3.94 (95% CI: 3.49, 4.45) pg/mg hair, geometric mean 3rd trimester HCC was 6.12 (95% CI: 4.96, 7.56) pg/mg hair. The coefficient of variance (CV) of 2nd trim HCC was 51.5%, CV of 3rd trimester HCC was 51.8%. Second and third trimester cortisol concentrations of participants that donated a hair sample twice were moderately correlated (n = 65, p < 0.01, Pearson's r = 0.571).
Table 2. Residential characteristics of the study participants
Variable
|
|
Categorical variables (n=149)
|
n (%)
|
Neighborhood greenspace
|
|
Access to small neighborhood greenspace
|
|
No
|
9 (6.0)
|
Yes
|
140 (94.0)
|
Access to large neighborhood greenspace
|
|
No
|
36 (24.2)
|
Yes
|
113 (75.8)
|
Eurostat urbanization
|
|
Cities
|
56 (37.6)
|
Towns and suburbs
|
93 (62.4)
|
Rural
|
0
|
Noise levels
|
|
Lden < 53 dB
|
97 (65.1)
|
Lden ≥ 53 dB
|
47 (31.5)
|
Missing
|
5 (3.4)
|
Continuous variables
|
Geometric mean (95% CI)
|
Distance to major roads (m) (n=149)
|
290 (240, 349)
|
2nd trimester air pollution ( µg/m3) (n=133)
|
NO2 - 90 days prior to sampling
|
23.03 (21.67, 24.47)
|
NO2 - 1 year prior to sampling
|
24.55 (23.39, 25.76)
|
PM2.5 - 90 days prior to sampling
|
11.61 (11.06, 12.21)
|
PM2.5- 1 year prior to sampling
|
13.09 (12.81, 13.37)
|
BC - 90 days prior to sampling
|
1.13 (1.05, 1.21)
|
BC - 1 year prior to sampling
|
1.29 (1.24, 1.36)
|
3rd trimester air pollution (µg/m3) (n=78)
|
|
NO2 - 90 days prior to sampling
|
23.19 (21.53, 24.98)
|
NO2 - 1 year prior to sampling
|
24.70 (23.27, 26.23)
|
PM2.5 - 90 days prior to sampling
|
11.55 (10.95, 12.18)
|
PM2.5- 1 year prior to sampling
|
12.70 (12.42, 12.99)
|
BC - 90 days prior to sampling
|
1.17 (1.07, 1.28)
|
BC - 1 year prior to sampling
|
1.29 (1.21, 1.37)
|
Ambient temperature (°Celsius)
|
|
90 days prior to 2nd trimester sampling (n=133)
|
10.5 (9.7, 11.4)
|
90 days prior to 3rd trimester sampling (n=81)
|
12.6 (11.7, 13.3)
|
Area deprivation index (%) (n=149)
|
16.4 (14.6, 18.1)
|
Note: Categorical data is described as frequencies (%), continuous data is described by geometric mean with 95% confidence interval (95% CI). All data is based on the maternal residential address. Major roads include E- or N-roads. Access to small neighborhood greenspace is defined as access to > 0.2 hectares (ha) of greenspace within a travel distance of 400 meters (m) from residence, access to large neighborhood greenspace is defined as access to > 10 ha of greenspace within a travel distance of 800 m from residence. Lden, day–evening–night noise level; NO2, nitrogen dioxide; PM2.5, fine particulate matter with an aerodynamic diameter ≤ 2.5 μm; BC, black carbon.
Spearman rank correlations of residential exposure characteristics are presented in Table 3 (2nd trimester study population) and Table 4 (3rd trimester study population).
Table 3. Spearman rank correlations between residential exposures in the 2nd trimester study population (n = 133)
|
BC
|
NO2
|
Distance to major road
|
Small greenspace
|
Large greenspace
|
Noise
|
ADI
|
PM2.5
|
0.68*
|
0.68*
|
-0.15
|
-0.02
|
-0.04
|
0.08
|
0.15
|
|
BC
|
0.89*
|
-0.29*
|
-0.09
|
-0.14
|
0.28*
|
0.38*
|
|
|
NO2
|
-0.37*
|
-0.08
|
-0.12
|
0.31*
|
0.46*
|
|
|
|
Distance to major road
|
0.01
|
0.17
|
-0.32*
|
-0.20*
|
|
|
|
|
Small greenspace
|
0.43*
|
-0.14
|
-0.11
|
|
|
|
|
|
Large greenspace
|
-0.20*
|
-0.18*
|
|
|
|
|
|
|
Noise
|
0.21
|
*Significant correlations (p < 0.05)
Note: Air pollutants were modelled at the maternal home address, 3-month mean concentrations were calculated. Major roads include E- or N-roads. Access to small neighborhood greenspace is defined as access to > 0.2 hectares (ha) of greenspace within a travel distance of 400 meters (m) from residence, access to large neighborhood greenspace is defined as access to > 10 ha of greenspace within a travel distance of 800 m from residence. Noise exposure is evaluated as exposure above the WHO health-based guideline of 53 dB Lden (day–evening–night noise level). ADI, area deprivation index; NO2, nitrogen dioxide; PM2.5, fine particulate matter with an aerodynamic diameter ≤ 2.5 μm; BC, black carbon.
Table 4. Spearman rank correlations between residential exposure in the 3rd trimester study population (n = 81)
|
BC
|
NO2
|
Distance to major road
|
Small greenspace
|
Large greenspace
|
Noise
|
ADI
|
PM2.5
|
0.61*
|
0.55*
|
-0.06
|
-0.19*
|
-0.31*
|
0.01
|
0.04
|
|
BC
|
0.89*
|
-0.24*
|
-0.22*
|
-0.30*
|
0.18
|
0.31*
|
|
|
NO2
|
-0.32*
|
-0.21*
|
-0.28*
|
0.24*
|
0.39*
|
|
|
|
Distance to major road
|
0.13
|
0.26*
|
-0.31*
|
-0.25*
|
|
|
|
|
Small greenspace
|
0.41*
|
-0.37*
|
-0.08
|
|
|
|
|
|
Large greenspace
|
-0.15*
|
-0.19
|
|
|
|
|
|
|
Noise
|
0.25*
|
*Significant correlations (p < 0.05)
Note: Air pollutants were modelled at the maternal home address, 3-month mean concentrations were calculated. Major roads include E- or N-roads. Access to small neighborhood greenspace is defined as access to > 0.2 hectares (ha) of greenspace within a travel distance of 400 meters (m) from residence, access to large neighborhood greenspace is defined as access to > 10 ha of greenspace within a travel distance of 800 m from residence. Noise exposure is evaluated as exposure above the WHO health-based guideline of 53 dB Lden (day–evening–night noise level). ADI, area deprivation index; NO2, nitrogen dioxide; PM2.5, fine particulate matter with an aerodynamic diameter ≤ 2.5 μm; BC, black carbon.
We observed strong positive correlations between 3-month mean air pollutant concentrations (r ranged from 0.61 to 0.89). Distance to major roads was negatively correlated with NO2 and BC concentrations (r ranged from -0.24 and -0.37), but not with PM2.5 concentrations.
Access to a neighborhood greenspace did not significantly correlate with air pollutants and distance to major roads in the 2nd trimester. In the 3rd trimester study population, we did find weak negative correlations between access to a large neighborhood greenspace and air pollutants (r ranged from -0.28 to -0.31) and a weak positive association of access to a large neighborhood greenspace with distance to major road (r = 0.26). The ADI was weakly positively correlated with NO2, BC and noise exposure above the WHO guideline (r ranged from 0.21 to 0.46) and negatively correlated with distance to major roads and access to a large neighborhood greenspace (r ranged from -0.18 to -0.25).
In the ANOVA, season of sampling and daily hair washing were identified as significant covariates of 2nd trimester HCC, no significant covariates were identified for 3rd trimester HCC (see Table S1 for details). None of the participants reported the systemic use of glucocorticoids. Residential noise exposure above the WHO guideline (≥53 dB Lden) was not significantly associated with 2nd or 3rd trimester HCC (p = 0.871, p = 0.190 respectively). Nor did we find significant associations between the ADI and 2nd or 3rd trimester HCC (p = 0.661, p = 0.388 resp.).
Results of the associations between air pollution exposure, access to neighborhood greenspace and maternal biological stress are presented in Table 5. We found a significant negative association between 3-month mean PM2.5 concentrations and 2nd trimester HCC in the unadjusted model (β = 0.81 (95% CI: 0.70, 0.95), p = 0.009), the association did not remain significant after adjustment for season of sampling (β = 0.87 (95% CI: 0.70, 1.08), p = 0.200). In the 3rd trimester, 3-month mean PM2.5 concentrations were not significantly associated with HCC in the unadjusted model (β = 1.20 (95%CI: 0.89, 1.62), p = 0.227), we observed a tendency towards a significant positive association after adjustment for season of sampling (β = 1.46 (95% CI: 1.01, 2.11), p = 0.051).
Table 5. Associations between residential exposures and hair cortisol concentrations in the second and third pregnancy trimester
Second trimester (n = 133)
|
Model I
|
Model II
|
Exposure (p25-p75)
|
p-value
|
β (95% CI)
|
p-value
|
β (95% CI)
|
3-month mean PM2.5 (9.66 – 13.64 µg/m3)
|
0.009
|
0.81 (0.70, 0.95)
|
0.200
|
0.87 (0.70, 1.08)
|
3-month mean NO2 (18.37 – 30.22 µg/m3)
|
0.934
|
0.99 (0.83, 1.18)
|
0.287
|
1.10 (0.92, 1.34)
|
3-month mean BC (0.84 – 1.49 µg/m3)
|
0.551
|
0.94 (0.79, 1.13)
|
0.775
|
1.03 (0.84, 1.27)
|
Distance to major road (143 – 642 m)
|
0.016
|
0.82 (0.70, 0.96)
|
0.011
|
0.82 (0.70, 0.95)
|
Access to a small NHGS
|
0.117
|
0.68 (0.42, 1.10)
|
0.061
|
0.63 (0.38, 1.02)
|
Access to a large NHGS
|
0.073
|
0.77 (0.57, 1.03)
|
0.095
|
0.78 (0.59, 1.04)
|
Third trimester (n = 81)
|
Model I
|
Model II
|
Exposure (p25-p75)
|
p-value
|
β (95% CI)
|
p-value
|
β (95% CI)
|
3-month mean PM2.5 (9.59 – 13.27 µg/m3)
|
0.227*
|
1.20 (0.89, 1.62)
|
0.051*
|
1.46 (1.01, 2.11)
|
3-month mean NO2 (18.35 – 30.00 µg/m3)
|
0.016*
|
1.42 (1.07, 1.88)
|
0.008*
|
1.53 (1.12, 2.09)
|
3-month mean BC (0.84 – 1.48 µg/m3)
|
0.032*
|
1.37 (1.03, 1.82)
|
0.017*
|
1.54 (1.08, 2.18)
|
Distance to major road (114 – 598 m)
|
0.040
|
0.74 (0.55, 0.99)
|
0.055
|
0.75 (0.56, 1.01)
|
Access to a small NHGS
|
0.354
|
0.67 (0.28, 1.05)
|
0.169
|
0.53 (0.22, 1.32)
|
Access to a large NHGS
|
0.062
|
0.65 (0.41, 1.02)
|
0.019
|
0.57 (0.36, 0.91)
|
Note: Estimates (β) of the linear regression models are presented with their 95% confidence intervals (95%CI) as a factor increase in hair cortisol concentrations for a factor increase in exposure from the 25th to the 75th percentile (p75/p25). Model I is unadjusted, Model II is adjusted for season of sampling, Significant associations (p < 0.05) are marked in bold. *Associations evaluated for 78 participants. Air pollutants were modelled at the maternal home address, 3-month mean concentrations were calculated. Major roads include E- or N-roads. Access to small neighborhood greenspace is defined as access to > 0.2 hectares (ha) of greenspace within a travel distance of 400 meters (m) from residence, access to large neighborhood greenspace is defined as access to > 10 ha of greenspace within a travel distance of 800 m from residence. CI, confidence interval; NO2, nitrogen dioxide; PM2.5, fine particulate matter with an aerodynamic diameter ≤ 2.5 μm; BC, black carbon; NHGS neighborhood greenspace. Significant associations (p < 0.05) are marked in bold.
We did not observe significant associations between 3-month mean NO2 and BC concentrations and 2nd trimester HCC. We observed a significant positive association between 3-month mean NO2 concentrations and 3rd trimester HCC in the unadjusted model (β = 1.42 (95% CI: 1.07, 1.88), p = 0.016), the association remained significant after adjustment for season of sampling (p = 0.008). For an increase of 3-month mean residential NO2 concentrations with a factor 1.63 (18.35 µg/m3 (p25) to 30 µg/m3 (p75)), an increase of 3rd trimester HCC with a factor 1.53 (95% CI: 1.12, 2.09) was estimated in the model, adjusted for season of sampling. We also observed a significant positive association between 3-month mean BC concentrations and 3rd trimester HCC in the unadjusted model (β= 1.37 (95% CI: 1.03, 1.82), p = 0.032) and after adjustment for season of sampling (p = 0.017). For an increase of 3-month mean residential BC concentrations with a factor 1.76 (0.84 µg/m3 (p25) to 1.48 µg/m3 (p75)), an increase of 3rd trimester HCC with a factor 1.54 (95% CI: 1.08, 2.18) was estimated in the model, adjusted for season of sampling. The model including season of sampling and residential 3-month mean NO2 concentrations explained 7.7% of the variation in 3rd trimester HCC, season of sampling and 3-month mean BC concentrations explained 5.9% of the variation in 3rd trimester HCC.
Residential distance to a major road was negatively associated with second trimester HCC in the unadjusted model (β = 0.82 (95% CI: 0.70, 0.96), p = 0.016) and in the season-adjusted model (p = 0.011). For an increase in distance to major roads with a factor 4.49 (143 m (p25) to 642 m (p75)), a decrease of 2nd trimester HCC with a factor 0.82 (95% CI: 0.70, 0.95) was estimated in the model, adjusted for season of sampling. This model explained 7.8% of the variation in 2nd trimester HCC. Distance to major roads was also negatively associated with 3rd trimester HCC in the unadjusted model (p = 0.040), for an increase of distance to a major road with a factor 5.25 (114 m (p25) to 598 m (p75)), a decrease of 3rd trimester HCC with a factor 0.74 (95% CI: 0.55, 0.99) was estimated. Distance to a major road explained 4% of the variation in 3rd trimester HCC. Adjusting the model for season of sampling slightly attenuated the association (β = 0.75 (95% CI: 0.56, 1.01), p = 0.055).
Access to a small neighborhood greenspace tended towards a significant negative association with 2nd trimester HCC (β = 0.68 (95%CI: 0.42, 1.10), p = 0.117 for the unadjusted model and β = 0.63 (95% CI: 0.38, 1.02, p = 0.061 for the season-adjusted model) and was not significantly associated with 3rd trimester HCC before or after adjustment for season of sampling (β = 0.67 (95%CI: 0.28, 1.05), p = 0.354 and β = 0.53 (95%CI: 0.22, 1.32), p = 0.169 resp.). Access to a large neighborhood greenspace tended towards a significant negative association with 2nd trimester HCC (β = 0.77 (95% CI: 0.57, 1.03), p = 0.073 for the unadjusted model, and β = 0.78 (95% CI: 0.59, 1.04), p = 0.095 for the season-adjusted model). Access to a large neighborhood greenspace also tended towards a significant negative association with 3rd trimester HCC in the unadjusted model (β = 0.65 (95% CI: 0.41, 1.02, p = 0.062), we observed a significant negative association after adjustment for season of sampling (β = 0.57 (0.36, 0.91), p = 0.019).
We tested whether access to a neighborhood greenspace moderated the associations between air pollution exposure, proximity to major roads and maternal biological stress. We found no significant interaction between access to a small or large neighborhood greenspace and air pollution constituents in relation to 2nd or 3rd trimester HCC (see Table S2 for details). We did observe a significant interaction between access to a large neighborhood greenspace (NHGS) and distance to major roads in relation to 2nd trimester HCC in both the unadjusted model, as presented in Figure 1, and in the model adjusted for season of sampling (p-interaction = 0.021, p-interaction = 0.034 resp.). Distance to major roads was significantly associated with 2nd trimester HCC for participants without access to a large NHGS (β = 0.64 (95% CI: 0.48, 0.85) p = 0.003). The association was not significant for participants with access to a large NHGS (β = 0.92 (95% CI: 0.77, 1.11), p = 0.399). The season-adjusted interaction model explained 10.8% of the variations in 2nd trimester HCC.
Figure 1 Interaction between distance to a major road and access to a large neighborhood greenspace in relation to 2nd trimester HCC
The interaction between access to a large neighborhood greenspace (NHGS) and distance to major roads in relation to 3rd trimester HCC tended towards significance in the unadjusted model (p-interaction = 0.073), as presented in Figure 2, and after adjustment for season of sampling (p-interaction = 0.080). For pregnant women without access to a large NHGS, we found a significant association between distance to major roads and 3rd trimester HCC (β = 0.53 (95% CI: 0.30, 0.94), p = 0.030) after adjustment for season of sampling. The association was not significant for pregnant women with access to a large NHGS (β = 0.96 (95% CI: 0.68, 1.36 ), p = 0.816). The interaction model, adjusted for season of sampling, explained 9.6% of the variations in 3rd trimester HCC.
Figure 2 Interaction between distance to a major road and access to a large neighborhood greenspace in relation to 3rd trimester HCC
In a sensitivity analysis, we evaluated the significance of associations between 1-year mean PM2.5, NO2 and BC concentrations and HCC to reflect the participants’ longer-term residential exposure to air pollution. Results are presented in Table S3. Extending the exposure period did not change our results much, we found significant positive associations between 1-year mean NO2 concentrations and 3rd trimester HCC in the unadjusted model and in the model, adjusted for season of sampling ( β = 1.49 (95% CI: 1.09, 2.04), p = 0.013 and β = 1.44 (95% CI: 1.05, 1.99), p = 0.024, resp.) and between 1-year mean BC concentrations and 3rd trimester HCC (β = 1.39 (95% CI: 1.01, 1.93), p = 0.046). After adjustment for season of sampling, the association between 1-year mean BC concentrations and 3rd trimester HCC slightly attenuated (p = 0.083, β = 1.36 (95% CI: 0.96, 1.92). The robustness of our results was also evaluated by additional adjustment of our models with daily hair washing. Our results, presented in Table S4, remained robust. Additional adjustment of the models for maternal age, pre-pregnancy BMI, educational attainment and ADI slightly strengthened the estimated associations of distance to major roads with 2nd and 3rd trimester HCC and of residential exposure to NO2, BC and access to large NHGS with 3rd trimester HCC (see Table S5 for details). Excluding participants of non-European origin did not change the statistical significance of the results (see Table S6 for details).