- Email: [email protected]
Neighbourhood greenness and birth outcomes in a Swedish birth cohort – A short communication
Health and Place 57 (2019) 200–203
Contents lists available at ScienceDirect
Health & Place journal homepage: www.elsevier.com/locate/healthplace
Short Communication
Neighbourhood greenness and birth outcomes in a Swedish birth cohort – A short communication
T
Charlotta Erikssona,b, Tomas Linda, Sandra Ekströma,b, Olena Gruzievaa,b, Antonios Georgelisa,b, Anna Bergströmb, Mare Lõhmusa,b,∗ a b
Centre for Occupational and Environmental Medicine, Stockholm County Council, Solnavägen 4, 11365 Stockholm, Sweden Institute of Environmental Medicine, Karolinska Institute, Box 210, 17177, Stockholm, Sweden
A R T I C LE I N FO
A B S T R A C T
Keywords: Urban greenness Birthweight Birth outcomes Small-for-gestational-age
The present study investigated whether associations between greenness and birth outcomes can be detected in children belonging to a Swedish birth cohort (BAMSE). Normalized difference vegetation index (NDVI) within a 500 m buffer zone around maternal address was used as estimate of greenness. Ordinary least squares and quantile regression models were performed to investigate associations between neighbourhood NDVI and birthweight (n = 2619), birth length (n = 2490) and head circumference (n = 2243). Logistic regression analyses were used to detect the association between NDVI and odds of being born as “small-” or “large-for-gestational-age”. There were no clear associations between NDVI and birth weight in the total sample. However, in a suburban sub-sample, increased NDVI levels were significantly associated with elevated birthweight of small new-borns (β2nd percentile = 276 g, 95% CI 61 to 492, p = 0.012), and significantly reduced the odds ratio (OR) for children being born as small-for-gestational-age (OR = 0.31 95% CI 0.1 to 1, p = 0.049). No significant associations were found between NDVI and birth length or head circumference. In conclusion, neighbourhood greenness appears not to be associated with birthweight as such, but rather decrease the odds of being born underweight, in particular in suburban areas.
1. Introduction Prior research has built a strong case for a positive association between maternal exposure to greenness during pregnancy and birth weight (Fong et al., 2018). Exposure to urban vegetation is generally believed to affect human health by inducing relaxation, promoting exercise and socialising as well as reducing harmful environmental exposures (Markevych et al., 2017). As various birth outcomes have been previously related to maternal stress (Bussières et al., 2015), levels of physical activity (da Silva et al., 2017), ambient temperatures (Carolan-Olah and Frankowska, 2014), environmental air pollution (Fleischer et al., 2014), and, possibly, noise pollution (Nieuwenhuijsen et al., 2017), it is likely that maternal exposure to urban greenness may influence birth outcomes via reduction of stress and harmful environmental exposures and increased physical activity (Fong et al., 2018). While the evidence for an association between maternal greenness exposure and risk for low birthweight is consistent in previous literature, findings for other birth outcomes, such as risk for preeclampsia and preterm birth and associations with head circumference are less frequent (Fong et al., 2018). Socioeconomic status (SES) is suggested to ∗
affect the relationship between greenness and birth outcomes, with stronger effects among individuals with lower SES (Fong et al., 2018; James et al., 2015). In addition, parental ethnic background has been found to affect the greenness-birth outcome relationship (Dadvand et al., 2014). The present study investigated the associations between greenness and birth outcomes using a prospective birth cohort in Stockholm, Sweden. Stockholm is one of the greenest capitals in Europe (Stockholm.se, ). However, the levels of greenness, and the direction of the associations between greenness and SES in Stockholm County have been previously reported to differ between suburban and urban areas (Persson et al., 2018). According to previous studies, we expected to find a positive association between the quantity of residential greenness and the birth outcomes such as the birth weight, length and head circumference in new-borns in Stockholm County. 2. Methods The BAMSE cohort consists of 4089 children born in pre-defined areas of Stockholm County, Sweden between February 1994 and
Corresponding author. Centre for Occupational and Environmental Medicine, Stockholm County Council, Solnavägen 4, 11365 Stockholm, Sweden. E-mail address: [email protected] (M. Lõhmus).
https://doi.org/10.1016/j.healthplace.2019.04.012 Received 6 November 2018; Received in revised form 23 April 2019; Accepted 26 April 2019 Available online 14 May 2019 1353-8292/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
Health and Place 57 (2019) 200–203
C. Eriksson, et al.
November 1996 (read more in (Wickman et al., 2002)). Perinatal data including birthweight, birth length, head circumference and gestational age, as well as weather the child was small- or large-for-gestation-age (≥2 standard deviations lower, respectively higher than the average weight of the reference population) were received from Swedish National Medical Birth Register (The National Board of Health and Welfare, 2018). Data on parental background and environmental factors were collected from parental questionnaires when the children were on average 2 months old. Only children with complete data sets were included in the analyses in present study. Normalized difference vegetation index (NDVI), a green biomass density indicator, was used as the estimate of surrounding greenness. The calculation of NDVI is based on the difference of surface reflectance in visible (0.4–0.7 lm) and near-infrared (0.7–1.1 lm) wavelengths. Values of NDVI range from −1 to +1, where higher values indicate denser green vegetation. Cloud-free satellite images were used to calculate mean NDVI levels in 500 m circular buffers around the maternal address at delivery (read more in (Fuertes et al., 2016)). Only newborns whose mothers had lived at the delivery address more than two months prior birth were included in the study. Linear regression models and quantile regression (applying bootstraps with 800 repetitions) models were performed to investigate associations between neighbourhood NDVI and birthweight (g; n = 2619), birth length (mm; n = 2490) and head circumference (mm; n = 2243). Two separate logistic regression analyses were used to detect the effect of NDVI on odds of being born as either “small-for-gestational-age” (SGA, 55 individuals of 2619) or “not-small-for-gestational-age”, and as “large-for-gestation-age” (LGA, 72 individuals of 2619) or “not-large-for-gestational-age”. All models were adjusted for length of pregnancy (weeks), air pollution (NO2 μg/m3), road traffic noise (Lden, dB), municipality (read more in (Wallas et al., 2018; Gruzieva et al., 2012)), and neighbourhood income level (SCB, 2011), as well as for mother's BMI, age at delivery (years), education (post high school yes/no), smoking during pregnancy (yes/no) and the origin of the parents (both Scandinavian-born, one or both non-Scandinavian). The four municipalities included in the study are traditionally classified as either suburban (Sundbyberg, Solna and Järfälla) or urban (Stockholm). To be able to detect differences related to area type, all regression models were expanded with interaction terms to account for effect modification of municipality on the effect of NDVI. Differences in urban and suburban populations were mirrored by the average NDVI values, which were 0.09 (Standard deviation, SD = 0.06) in the “urban” Stockholm municipality and 0.31 (SD = 0.05), 0.23 (SD = 0.04) and 0.23 (SD = 0.08) in the “suburban” Järfälla, Solna and Sundbyberg municipalities, respectively.
Table 1 The association between one IQR increase in maternal NDVI exposure (average values for a 500 m buffer zone around domestic address at the time of delivery) and birth outcomes. All models are adjusted for confoundersa. a) NDVI and birth weight (g), b) Odds ratios (OR) for being “small for gestation age” (SGA) at birth related to NDVI exposure c) NDVI and other birth outcomes (n = number of individuals, CI = confidence interval). a) Birth weight b
Total (n = 2619) Suburban (n = 1829) Urban (n = 790)
β (g)
p
95% CI
7.30 48.00 −79.00
0.809 0.179 0.126
−51.70 −22.00 −179.00
66.20 119.00 22.00
b) SGA
OR
p
95% CI
Total (n = 2619) Suburban (n = 1829) Urban (n = 790)
0.54 0.31 1.39
0.199 0.049 0.643
0.21 0.10 0.34
Other birth outcomes
β (mm)
p
95% CI
Birth length (n = 2619) Head circumference (n = 2243)
−1.80 −1.80
0.193 0.072
−4.50 −0.20
1.39 1.00 5.67
c)
0.90 3.90
a Length of pregnancy (weeks), air pollution (NO2 μg/m3), road traffic noise (Lden, dB), municipality, neighbourhood income level, mother's BMI, mother's age at delivery (years), mother's education (post high school yes/no), mother smoking during pregnancy (yes/no), and the origin of the parents (both Scandinavian-born, one or both non-Scandinavian). b Interaction coefficient between NDVI and area type significant (p = 0.039).
Table 2 The association between one IQR increase in maternal NDVI exposure (average values for a 500 m buffer zone around domestic address at the time of delivery) and birthweight of “small” and “large” new-borns by using an adjusted quantile regression model stratified by area type. Model is adjusted for confoundersa. Confidence intervals are calculated by applying bootstrap method with 800 repetitions. Area type 2nd percentile Suburban Urban 3rd percentile Suburban Urban 5th percentile Suburban Urban 70th percentile Suburban Urban 95th percentile Suburban Urban
3. Results Background characteristics are presented in supplement Table 1S. One interquartile range, IQR, value of NDVI, in the present study, corresponded to a NDVI estimate of 0.167. The association between NDVI and birthweight did not show any clear trends when all children were included in the linear regression model (Table 1). However, since there was a significant interaction between NDVI and municipality type (urban and suburban, p = 0.039), separate analyses were performed within urban and suburban areas. The quantile regression analyses showed that one IQR increase in NDVI was significantly associated with elevated birthweight (β = 276 g, 95% CI 61 to 492, p = 0.012, Table 2) in suburban children with low birth weight (i.e. the second percentile). One IQR increase in NDVI in suburban areas also significantly reduced the odds ratio of children being born as SGA (OR = 0.31, 95% CI 0.1 to 1, p = 0.049, Table 1), but did not affect the odds ratios for being LGA. We did not find any significant associations between NDVI and birth length or head circumference (Table 1C). Since the interaction coefficients for municipality type in these analyses were not significant, we did not conduct any subgroup-analyses.
β (g)
95% p-value
95% Confidence interval
276 −42
0.012 0.826
61; 492 −416; 332
205 −62
0.009 0.643
50; 360 −323; 200
153 −66
0.015 0.430
30; 277 −230; 98
10 −157
0.830 0.001
−78; 97 −251; −59
99 −47
0.308 0.730
−91; 288 −313; 219
a Length of pregnancy (weeks), air pollution (NO2 μg/m3), road traffic noise (Lden, dB), municipality, neighbourhood income level, mother's BMI, mother's age at delivery (years), mother's education (post high school yes/no), mother smoking during pregnancy (yes/no) and the origin of the parents (both Scandinavian-born, one or both non-Scandinavian).
4. Discussion We found that the odds of being born as small for gestation age were associated with maternal exposure to residential greenness during pregnancy. However, associations between neighbourhood greenness and birthweight were apparent in suburban areas only, and 201
Health and Place 57 (2019) 200–203
C. Eriksson, et al.
be associated with birthweight as such, but rather decrease the odds of being born small-for-gestational-age. Moreover, type of residential area (urban or suburban) was a significant effect modifier of the association between greenness and birth weight.
predominantly affected new-borns with low birth weight. No associations between maternal greenness exposure and birth length or head circumference were detected. The association between greenness and birth outcomes has been studied in multiple countries and positive associations have been reported consistently across the majority of the studies (Fong et al., 2018; James et al., 2015; Kondo et al., 2018). Occasionally, higher greenness exposure has also been linked to lower risks for preterm births (Hystad et al., 2014; Casey et al., 2016; Grazuleviciene et al., 2015), larger head circumference (Dadvand et al., 2012), and lower infant mortality risk (Kihal-Talantikite et al., 2013), however, these finding are seldom replicated across studies (Twohig-Bennett and Jones, 2018). There may be several underlying reasons for the significant associations between birthweight and greenness in the present study only being apparent in suburban neighbourhoods and not in the urban ones. Firstly, fewer individuals belonged to the “urban” group than to the suburban one, and almost no individuals among the urban population were exposed to above median levels of residential greenness (see Table 1S), which affects the data available for high/low SES and high/ low NDVI exposed population groups in different areas. However, it still raises the question of why we do not find any negative effects associated with low levels of neighbourhood greenness in the urban group. Unfortunately, we can only speculate about the answer. While the SES and residential area greenness in many places is assumed to be highly and positively correlated (WHO, 2016), according to our recent publication, in Stockholm County the direction of the SES/ greenness association differs between area types (Persson et al., 2018). In urban neighborhoods of Stockholm, higher income is associated with less greenness, whereas in suburban areas the association is the opposite, linking higher income to more greenness. In the sensitivity analyses of the present study, we did not find that living in a low SES area significantly modified the effect of NDVI (data not shown). Despite this, differences in income distribution between urban and suburban municipalities were still apparent: while the population within the lowest income quartile was as good as missing in the urban municipality, very few individuals residing in suburban areas belonged to the group of the highest income quartile (see also Table 1S for income distribution). Furthermore, significantly higher number of mothers had post-high school education in urban area than in suburban ones (Table 1S). It is possible that there are certain behavioral/habitual differences between the study subjects living in urban and suburban areas, respectively, that we have not been able to measure. Health benefits associated to greenness exposure have repeatedly seen to be less obvious or absent in population groups with high SES/education levels and only detected in groups with low SES/education levels (Dadvand et al., 2012, 2014; Agay-Shay et al., 2014; Markevych et al., 2014; Cusack et al., 2017). Consequently, it has been hypothesized that high-income takers may be able to compensate in several ways for the lack of residential green structure. Thus, in our study, it is possible that the urban population to a higher degree compensate their lack of residential greenness by, for example, travelling or owning/visiting summerhouses. That kind of compensatory behavior would, however, not be available for low-income takers, which may make them more dependable on health benefits from greenness close to their neighbourhood. Some limitations of our study should be mentioned. Despite being one of the best birth cohorts in Europe regarding the background information of the study subjects, the BAMSE cohort was primarily designed to investigate the effect of environmental exposures after birth on the development of allergies and is therefore not entirely optimal for a study about pre-birth exposure effects on birth outcomes. Furthermore, the children included in the cohort were all declared to be healthy new-borns, which most likely excluded cases with very low birth weight. Finally, as the birth of the BAMSE cohort children took place in the beginning of 90s, the exposure data that is available is not as detailed as it would be if the study were done today. To conclude, in our study, neighbourhood greenness appears not to
Declarations of interest for all authors None. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.healthplace.2019.04.012. References Agay-Shay, K., Peled, A., Crespo, A.V., Peretz, C., Amitai, Y., Linn, S., et al., 2014. Green spaces and adverse pregnancy outcomes. Occup. Environ. Med. 71 (8), 562–569. Bussières, E.-L., Tarabulsy, G.M., Pearson, J., Tessier, R., Forest, J.-C., Giguère, Y., 2015. Maternal prenatal stress and infant birth weight and gestational age: a meta-analysis of prospective studies. Dev. Rev. 36, 179–199. Carolan-Olah, M., Frankowska, D., 2014. High environmental temperature and preterm birth: a review of the evidence. Midwifery 30 (1), 50–59. Casey, J.A., James, P., Rudolph, K.E., Wu, C.-D., Schwartz, B.S., 2016. Greenness and birth outcomes in a range of Pennsylvania communities. Int. J. Environ. Res. Public Health 13 (3), 311. Cusack, L., Larkin, A., Carozza, S., Hystad, P., 2017. Associations between residential greenness and birth outcomes across Texas. Environ. Res. 152, 88–95. da Silva, S.G., Ricardo, L.I., Evenson, K.R., Hallal, P.C., 2017. Leisure-time physical activity in pregnancy and maternal-child health: a systematic review and meta-analysis of randomized controlled trials and cohort studies. Sports Med. 47 (2), 295–317. Dadvand, P., Sunyer, J., Basagaña, X., Ballester, F., Lertxundi, A., Fernández-Somoano, A., et al., 2012. Surrounding greenness and pregnancy outcomes in four Spanish birth cohorts. Environ. Health Perspect. 120 (10), 1481. Dadvand, P., Wright, J., Martinez, D., Basagaña, X., McEachan, R.R., Cirach, M., et al., 2014. Inequality, green spaces, and pregnant women: roles of ethnicity and individual and neighbourhood socioeconomic status. Environ. Int. 71, 101–108. Fleischer, N.L., Merialdi, M., van Donkelaar, A., Vadillo-Ortega, F., Martin, R.V., Betran, A.P., et al., 2014. Outdoor air pollution, preterm birth, and low birth weight: analysis of the world health organization global survey on maternal and perinatal health. Environ. Health Perspect. 122 (4), 425. Fong, K.C., Hart, J.E., James, P., 2018. A review of epidemiologic studies on greenness and health: updated literature through 2017. Curr. Environ. Health Rep. 5 (1), 77–87. Fuertes, E., Markevych, I., Bowatte, G., Gruzieva, O., Gehring, U., Becker, A., et al., 2016. Residential greenness is differentially associated with childhood allergic rhinitis and aeroallergen sensitization in seven birth cohorts. Allergy 71 (10), 1461–1471. Grazuleviciene, R., Danileviciute, A., Dedele, A., Vencloviene, J., Andrusaityte, S., Uzdanaviciute, I., et al., 2015. Surrounding greenness, proximity to city parks and pregnancy outcomes in Kaunas cohort study. Int. J. Hyg Environ. Health 218 (3), 358–365. Gruzieva, O., Bellander, T., Eneroth, K., Kull, I., Melén, E., Nordling, E., et al., 2012. Traffic-related air pollution and development of allergic sensitization in children during the first 8 years of life. J. Allergy Clin. Immunol. 129 (1), 240–246. Hystad, P., Davies, H.W., Frank, L., Van Loon, J., Gehring, U., Tamburic, L., et al., 2014. Residential greenness and birth outcomes: evaluating the influence of spatially correlated built-environment factors. Environ. Health Perspect. 122 (10), 1095. James, P., Banay, R.F., Hart, J.E., Laden, F., 2015. A review of the health benefits of greenness. Curr. Epidemiol. Rep. 2 (2), 131–142. Kihal-Talantikite, W., Padilla, C.M., Lalloué, B., Gelormini, M., Zmirou-Navier, D., Deguen, S., 2013. Green space, social inequalities and neonatal mortality in France. BMC Pregnancy Childbirth 13 (1), 191. Kondo, M.C., Fluehr, J.M., McKeon, T., Branas, C.C., 2018. Urban green space and its impact on human health. Int. J. Environ. Res. Public Health 15 (3), 445. Markevych, I., Fuertes, E., Tiesler, C.M., Birk, M., Bauer, C.-P., Koletzko, S., et al., 2014. Surrounding greenness and birth weight: results from the GINIplus and LISAplus birth cohorts in Munich. Health Place 26, 39–46. Markevych, I., Schoierer, J., Hartig, T., Chudnovsky, A., Hystad, P., Dzhambov, A.M., et al., 2017. Exploring pathways linking greenspace to health: theoretical and methodological guidance. Environ. Res. 158, 301–317. Nieuwenhuijsen, M.J., Ristovska, G., Dadvand, P., 2017. WHO environmental noise guidelines for the European region: a systematic review on environmental noise and adverse birth outcomes. Int. J. Environ. Res. Public Health 14 (10), 1252. Persson, Å., Eriksson, C., Lõhmus, M., 2018. Inverse associations between neighborhood socioeconomic factors and green structure in urban and suburban municipalities of Stockholm County. Landsc. Urban Plann. 179, 103–106. SCB, 2011. SCB-data för forskning. Innehållsbeskrivning av olika register. Statistics Sweden. Register Unit, Örebro. Stockholm.se Stockholm. http://www.stockholm.se/KulturFritid/Park-och-natur/ Stockholms Stad, Accessed date: 28 December 2016. The National Board of Health and Welfare, 2018. Swedish Medical Birth Register. The National Board of Health and Welfare, Sweden. http://www.socialstyrelsen.se/
202
Health and Place 57 (2019) 200–203
C. Eriksson, et al.
Wallas, A., Ekström, S., Eriksson, C., Gruzieva, O., Pyko, A., Bergström, A., et al., 2018. Road Traffic Noise Exposure and Development of Body Mass Index from Birth to Adolescence. Wickman, M., Kull, I., Pershagen, G., Nordvall, S.L., 2002. The BAMSE project: presentation of a prospective longitudinal birth cohort study. Pediatr. Allergy Immunol. 13, 11–13.
register/halsodataregister/medicinskafodelseregistret/inenglishStockholm. Twohig-Bennett, C., Jones, A., 2018. The health benefits of the great outdoors: a systematic review and meta-analysis of greenspace exposure and health outcomes. Environ. Res. 166, 628–637. WHO, 2016. Urban Green Spaces and Health. A Review of Evidence. WHO Regional office for Europe, Copenhagen.
203
Contents lists available at ScienceDirect
Health & Place journal homepage: www.elsevier.com/locate/healthplace
Short Communication
Neighbourhood greenness and birth outcomes in a Swedish birth cohort – A short communication
T
Charlotta Erikssona,b, Tomas Linda, Sandra Ekströma,b, Olena Gruzievaa,b, Antonios Georgelisa,b, Anna Bergströmb, Mare Lõhmusa,b,∗ a b
Centre for Occupational and Environmental Medicine, Stockholm County Council, Solnavägen 4, 11365 Stockholm, Sweden Institute of Environmental Medicine, Karolinska Institute, Box 210, 17177, Stockholm, Sweden
A R T I C LE I N FO
A B S T R A C T
Keywords: Urban greenness Birthweight Birth outcomes Small-for-gestational-age
The present study investigated whether associations between greenness and birth outcomes can be detected in children belonging to a Swedish birth cohort (BAMSE). Normalized difference vegetation index (NDVI) within a 500 m buffer zone around maternal address was used as estimate of greenness. Ordinary least squares and quantile regression models were performed to investigate associations between neighbourhood NDVI and birthweight (n = 2619), birth length (n = 2490) and head circumference (n = 2243). Logistic regression analyses were used to detect the association between NDVI and odds of being born as “small-” or “large-for-gestational-age”. There were no clear associations between NDVI and birth weight in the total sample. However, in a suburban sub-sample, increased NDVI levels were significantly associated with elevated birthweight of small new-borns (β2nd percentile = 276 g, 95% CI 61 to 492, p = 0.012), and significantly reduced the odds ratio (OR) for children being born as small-for-gestational-age (OR = 0.31 95% CI 0.1 to 1, p = 0.049). No significant associations were found between NDVI and birth length or head circumference. In conclusion, neighbourhood greenness appears not to be associated with birthweight as such, but rather decrease the odds of being born underweight, in particular in suburban areas.
1. Introduction Prior research has built a strong case for a positive association between maternal exposure to greenness during pregnancy and birth weight (Fong et al., 2018). Exposure to urban vegetation is generally believed to affect human health by inducing relaxation, promoting exercise and socialising as well as reducing harmful environmental exposures (Markevych et al., 2017). As various birth outcomes have been previously related to maternal stress (Bussières et al., 2015), levels of physical activity (da Silva et al., 2017), ambient temperatures (Carolan-Olah and Frankowska, 2014), environmental air pollution (Fleischer et al., 2014), and, possibly, noise pollution (Nieuwenhuijsen et al., 2017), it is likely that maternal exposure to urban greenness may influence birth outcomes via reduction of stress and harmful environmental exposures and increased physical activity (Fong et al., 2018). While the evidence for an association between maternal greenness exposure and risk for low birthweight is consistent in previous literature, findings for other birth outcomes, such as risk for preeclampsia and preterm birth and associations with head circumference are less frequent (Fong et al., 2018). Socioeconomic status (SES) is suggested to ∗
affect the relationship between greenness and birth outcomes, with stronger effects among individuals with lower SES (Fong et al., 2018; James et al., 2015). In addition, parental ethnic background has been found to affect the greenness-birth outcome relationship (Dadvand et al., 2014). The present study investigated the associations between greenness and birth outcomes using a prospective birth cohort in Stockholm, Sweden. Stockholm is one of the greenest capitals in Europe (Stockholm.se, ). However, the levels of greenness, and the direction of the associations between greenness and SES in Stockholm County have been previously reported to differ between suburban and urban areas (Persson et al., 2018). According to previous studies, we expected to find a positive association between the quantity of residential greenness and the birth outcomes such as the birth weight, length and head circumference in new-borns in Stockholm County. 2. Methods The BAMSE cohort consists of 4089 children born in pre-defined areas of Stockholm County, Sweden between February 1994 and
Corresponding author. Centre for Occupational and Environmental Medicine, Stockholm County Council, Solnavägen 4, 11365 Stockholm, Sweden. E-mail address: [email protected] (M. Lõhmus).
https://doi.org/10.1016/j.healthplace.2019.04.012 Received 6 November 2018; Received in revised form 23 April 2019; Accepted 26 April 2019 Available online 14 May 2019 1353-8292/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
Health and Place 57 (2019) 200–203
C. Eriksson, et al.
November 1996 (read more in (Wickman et al., 2002)). Perinatal data including birthweight, birth length, head circumference and gestational age, as well as weather the child was small- or large-for-gestation-age (≥2 standard deviations lower, respectively higher than the average weight of the reference population) were received from Swedish National Medical Birth Register (The National Board of Health and Welfare, 2018). Data on parental background and environmental factors were collected from parental questionnaires when the children were on average 2 months old. Only children with complete data sets were included in the analyses in present study. Normalized difference vegetation index (NDVI), a green biomass density indicator, was used as the estimate of surrounding greenness. The calculation of NDVI is based on the difference of surface reflectance in visible (0.4–0.7 lm) and near-infrared (0.7–1.1 lm) wavelengths. Values of NDVI range from −1 to +1, where higher values indicate denser green vegetation. Cloud-free satellite images were used to calculate mean NDVI levels in 500 m circular buffers around the maternal address at delivery (read more in (Fuertes et al., 2016)). Only newborns whose mothers had lived at the delivery address more than two months prior birth were included in the study. Linear regression models and quantile regression (applying bootstraps with 800 repetitions) models were performed to investigate associations between neighbourhood NDVI and birthweight (g; n = 2619), birth length (mm; n = 2490) and head circumference (mm; n = 2243). Two separate logistic regression analyses were used to detect the effect of NDVI on odds of being born as either “small-for-gestational-age” (SGA, 55 individuals of 2619) or “not-small-for-gestational-age”, and as “large-for-gestation-age” (LGA, 72 individuals of 2619) or “not-large-for-gestational-age”. All models were adjusted for length of pregnancy (weeks), air pollution (NO2 μg/m3), road traffic noise (Lden, dB), municipality (read more in (Wallas et al., 2018; Gruzieva et al., 2012)), and neighbourhood income level (SCB, 2011), as well as for mother's BMI, age at delivery (years), education (post high school yes/no), smoking during pregnancy (yes/no) and the origin of the parents (both Scandinavian-born, one or both non-Scandinavian). The four municipalities included in the study are traditionally classified as either suburban (Sundbyberg, Solna and Järfälla) or urban (Stockholm). To be able to detect differences related to area type, all regression models were expanded with interaction terms to account for effect modification of municipality on the effect of NDVI. Differences in urban and suburban populations were mirrored by the average NDVI values, which were 0.09 (Standard deviation, SD = 0.06) in the “urban” Stockholm municipality and 0.31 (SD = 0.05), 0.23 (SD = 0.04) and 0.23 (SD = 0.08) in the “suburban” Järfälla, Solna and Sundbyberg municipalities, respectively.
Table 1 The association between one IQR increase in maternal NDVI exposure (average values for a 500 m buffer zone around domestic address at the time of delivery) and birth outcomes. All models are adjusted for confoundersa. a) NDVI and birth weight (g), b) Odds ratios (OR) for being “small for gestation age” (SGA) at birth related to NDVI exposure c) NDVI and other birth outcomes (n = number of individuals, CI = confidence interval). a) Birth weight b
Total (n = 2619) Suburban (n = 1829) Urban (n = 790)
β (g)
p
95% CI
7.30 48.00 −79.00
0.809 0.179 0.126
−51.70 −22.00 −179.00
66.20 119.00 22.00
b) SGA
OR
p
95% CI
Total (n = 2619) Suburban (n = 1829) Urban (n = 790)
0.54 0.31 1.39
0.199 0.049 0.643
0.21 0.10 0.34
Other birth outcomes
β (mm)
p
95% CI
Birth length (n = 2619) Head circumference (n = 2243)
−1.80 −1.80
0.193 0.072
−4.50 −0.20
1.39 1.00 5.67
c)
0.90 3.90
a Length of pregnancy (weeks), air pollution (NO2 μg/m3), road traffic noise (Lden, dB), municipality, neighbourhood income level, mother's BMI, mother's age at delivery (years), mother's education (post high school yes/no), mother smoking during pregnancy (yes/no), and the origin of the parents (both Scandinavian-born, one or both non-Scandinavian). b Interaction coefficient between NDVI and area type significant (p = 0.039).
Table 2 The association between one IQR increase in maternal NDVI exposure (average values for a 500 m buffer zone around domestic address at the time of delivery) and birthweight of “small” and “large” new-borns by using an adjusted quantile regression model stratified by area type. Model is adjusted for confoundersa. Confidence intervals are calculated by applying bootstrap method with 800 repetitions. Area type 2nd percentile Suburban Urban 3rd percentile Suburban Urban 5th percentile Suburban Urban 70th percentile Suburban Urban 95th percentile Suburban Urban
3. Results Background characteristics are presented in supplement Table 1S. One interquartile range, IQR, value of NDVI, in the present study, corresponded to a NDVI estimate of 0.167. The association between NDVI and birthweight did not show any clear trends when all children were included in the linear regression model (Table 1). However, since there was a significant interaction between NDVI and municipality type (urban and suburban, p = 0.039), separate analyses were performed within urban and suburban areas. The quantile regression analyses showed that one IQR increase in NDVI was significantly associated with elevated birthweight (β = 276 g, 95% CI 61 to 492, p = 0.012, Table 2) in suburban children with low birth weight (i.e. the second percentile). One IQR increase in NDVI in suburban areas also significantly reduced the odds ratio of children being born as SGA (OR = 0.31, 95% CI 0.1 to 1, p = 0.049, Table 1), but did not affect the odds ratios for being LGA. We did not find any significant associations between NDVI and birth length or head circumference (Table 1C). Since the interaction coefficients for municipality type in these analyses were not significant, we did not conduct any subgroup-analyses.
β (g)
95% p-value
95% Confidence interval
276 −42
0.012 0.826
61; 492 −416; 332
205 −62
0.009 0.643
50; 360 −323; 200
153 −66
0.015 0.430
30; 277 −230; 98
10 −157
0.830 0.001
−78; 97 −251; −59
99 −47
0.308 0.730
−91; 288 −313; 219
a Length of pregnancy (weeks), air pollution (NO2 μg/m3), road traffic noise (Lden, dB), municipality, neighbourhood income level, mother's BMI, mother's age at delivery (years), mother's education (post high school yes/no), mother smoking during pregnancy (yes/no) and the origin of the parents (both Scandinavian-born, one or both non-Scandinavian).
4. Discussion We found that the odds of being born as small for gestation age were associated with maternal exposure to residential greenness during pregnancy. However, associations between neighbourhood greenness and birthweight were apparent in suburban areas only, and 201
Health and Place 57 (2019) 200–203
C. Eriksson, et al.
be associated with birthweight as such, but rather decrease the odds of being born small-for-gestational-age. Moreover, type of residential area (urban or suburban) was a significant effect modifier of the association between greenness and birth weight.
predominantly affected new-borns with low birth weight. No associations between maternal greenness exposure and birth length or head circumference were detected. The association between greenness and birth outcomes has been studied in multiple countries and positive associations have been reported consistently across the majority of the studies (Fong et al., 2018; James et al., 2015; Kondo et al., 2018). Occasionally, higher greenness exposure has also been linked to lower risks for preterm births (Hystad et al., 2014; Casey et al., 2016; Grazuleviciene et al., 2015), larger head circumference (Dadvand et al., 2012), and lower infant mortality risk (Kihal-Talantikite et al., 2013), however, these finding are seldom replicated across studies (Twohig-Bennett and Jones, 2018). There may be several underlying reasons for the significant associations between birthweight and greenness in the present study only being apparent in suburban neighbourhoods and not in the urban ones. Firstly, fewer individuals belonged to the “urban” group than to the suburban one, and almost no individuals among the urban population were exposed to above median levels of residential greenness (see Table 1S), which affects the data available for high/low SES and high/ low NDVI exposed population groups in different areas. However, it still raises the question of why we do not find any negative effects associated with low levels of neighbourhood greenness in the urban group. Unfortunately, we can only speculate about the answer. While the SES and residential area greenness in many places is assumed to be highly and positively correlated (WHO, 2016), according to our recent publication, in Stockholm County the direction of the SES/ greenness association differs between area types (Persson et al., 2018). In urban neighborhoods of Stockholm, higher income is associated with less greenness, whereas in suburban areas the association is the opposite, linking higher income to more greenness. In the sensitivity analyses of the present study, we did not find that living in a low SES area significantly modified the effect of NDVI (data not shown). Despite this, differences in income distribution between urban and suburban municipalities were still apparent: while the population within the lowest income quartile was as good as missing in the urban municipality, very few individuals residing in suburban areas belonged to the group of the highest income quartile (see also Table 1S for income distribution). Furthermore, significantly higher number of mothers had post-high school education in urban area than in suburban ones (Table 1S). It is possible that there are certain behavioral/habitual differences between the study subjects living in urban and suburban areas, respectively, that we have not been able to measure. Health benefits associated to greenness exposure have repeatedly seen to be less obvious or absent in population groups with high SES/education levels and only detected in groups with low SES/education levels (Dadvand et al., 2012, 2014; Agay-Shay et al., 2014; Markevych et al., 2014; Cusack et al., 2017). Consequently, it has been hypothesized that high-income takers may be able to compensate in several ways for the lack of residential green structure. Thus, in our study, it is possible that the urban population to a higher degree compensate their lack of residential greenness by, for example, travelling or owning/visiting summerhouses. That kind of compensatory behavior would, however, not be available for low-income takers, which may make them more dependable on health benefits from greenness close to their neighbourhood. Some limitations of our study should be mentioned. Despite being one of the best birth cohorts in Europe regarding the background information of the study subjects, the BAMSE cohort was primarily designed to investigate the effect of environmental exposures after birth on the development of allergies and is therefore not entirely optimal for a study about pre-birth exposure effects on birth outcomes. Furthermore, the children included in the cohort were all declared to be healthy new-borns, which most likely excluded cases with very low birth weight. Finally, as the birth of the BAMSE cohort children took place in the beginning of 90s, the exposure data that is available is not as detailed as it would be if the study were done today. To conclude, in our study, neighbourhood greenness appears not to
Declarations of interest for all authors None. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.healthplace.2019.04.012. References Agay-Shay, K., Peled, A., Crespo, A.V., Peretz, C., Amitai, Y., Linn, S., et al., 2014. Green spaces and adverse pregnancy outcomes. Occup. Environ. Med. 71 (8), 562–569. Bussières, E.-L., Tarabulsy, G.M., Pearson, J., Tessier, R., Forest, J.-C., Giguère, Y., 2015. Maternal prenatal stress and infant birth weight and gestational age: a meta-analysis of prospective studies. Dev. Rev. 36, 179–199. Carolan-Olah, M., Frankowska, D., 2014. High environmental temperature and preterm birth: a review of the evidence. Midwifery 30 (1), 50–59. Casey, J.A., James, P., Rudolph, K.E., Wu, C.-D., Schwartz, B.S., 2016. Greenness and birth outcomes in a range of Pennsylvania communities. Int. J. Environ. Res. Public Health 13 (3), 311. Cusack, L., Larkin, A., Carozza, S., Hystad, P., 2017. Associations between residential greenness and birth outcomes across Texas. Environ. Res. 152, 88–95. da Silva, S.G., Ricardo, L.I., Evenson, K.R., Hallal, P.C., 2017. Leisure-time physical activity in pregnancy and maternal-child health: a systematic review and meta-analysis of randomized controlled trials and cohort studies. Sports Med. 47 (2), 295–317. Dadvand, P., Sunyer, J., Basagaña, X., Ballester, F., Lertxundi, A., Fernández-Somoano, A., et al., 2012. Surrounding greenness and pregnancy outcomes in four Spanish birth cohorts. Environ. Health Perspect. 120 (10), 1481. Dadvand, P., Wright, J., Martinez, D., Basagaña, X., McEachan, R.R., Cirach, M., et al., 2014. Inequality, green spaces, and pregnant women: roles of ethnicity and individual and neighbourhood socioeconomic status. Environ. Int. 71, 101–108. Fleischer, N.L., Merialdi, M., van Donkelaar, A., Vadillo-Ortega, F., Martin, R.V., Betran, A.P., et al., 2014. Outdoor air pollution, preterm birth, and low birth weight: analysis of the world health organization global survey on maternal and perinatal health. Environ. Health Perspect. 122 (4), 425. Fong, K.C., Hart, J.E., James, P., 2018. A review of epidemiologic studies on greenness and health: updated literature through 2017. Curr. Environ. Health Rep. 5 (1), 77–87. Fuertes, E., Markevych, I., Bowatte, G., Gruzieva, O., Gehring, U., Becker, A., et al., 2016. Residential greenness is differentially associated with childhood allergic rhinitis and aeroallergen sensitization in seven birth cohorts. Allergy 71 (10), 1461–1471. Grazuleviciene, R., Danileviciute, A., Dedele, A., Vencloviene, J., Andrusaityte, S., Uzdanaviciute, I., et al., 2015. Surrounding greenness, proximity to city parks and pregnancy outcomes in Kaunas cohort study. Int. J. Hyg Environ. Health 218 (3), 358–365. Gruzieva, O., Bellander, T., Eneroth, K., Kull, I., Melén, E., Nordling, E., et al., 2012. Traffic-related air pollution and development of allergic sensitization in children during the first 8 years of life. J. Allergy Clin. Immunol. 129 (1), 240–246. Hystad, P., Davies, H.W., Frank, L., Van Loon, J., Gehring, U., Tamburic, L., et al., 2014. Residential greenness and birth outcomes: evaluating the influence of spatially correlated built-environment factors. Environ. Health Perspect. 122 (10), 1095. James, P., Banay, R.F., Hart, J.E., Laden, F., 2015. A review of the health benefits of greenness. Curr. Epidemiol. Rep. 2 (2), 131–142. Kihal-Talantikite, W., Padilla, C.M., Lalloué, B., Gelormini, M., Zmirou-Navier, D., Deguen, S., 2013. Green space, social inequalities and neonatal mortality in France. BMC Pregnancy Childbirth 13 (1), 191. Kondo, M.C., Fluehr, J.M., McKeon, T., Branas, C.C., 2018. Urban green space and its impact on human health. Int. J. Environ. Res. Public Health 15 (3), 445. Markevych, I., Fuertes, E., Tiesler, C.M., Birk, M., Bauer, C.-P., Koletzko, S., et al., 2014. Surrounding greenness and birth weight: results from the GINIplus and LISAplus birth cohorts in Munich. Health Place 26, 39–46. Markevych, I., Schoierer, J., Hartig, T., Chudnovsky, A., Hystad, P., Dzhambov, A.M., et al., 2017. Exploring pathways linking greenspace to health: theoretical and methodological guidance. Environ. Res. 158, 301–317. Nieuwenhuijsen, M.J., Ristovska, G., Dadvand, P., 2017. WHO environmental noise guidelines for the European region: a systematic review on environmental noise and adverse birth outcomes. Int. J. Environ. Res. Public Health 14 (10), 1252. Persson, Å., Eriksson, C., Lõhmus, M., 2018. Inverse associations between neighborhood socioeconomic factors and green structure in urban and suburban municipalities of Stockholm County. Landsc. Urban Plann. 179, 103–106. SCB, 2011. SCB-data för forskning. Innehållsbeskrivning av olika register. Statistics Sweden. Register Unit, Örebro. Stockholm.se Stockholm. http://www.stockholm.se/KulturFritid/Park-och-natur/ Stockholms Stad, Accessed date: 28 December 2016. The National Board of Health and Welfare, 2018. Swedish Medical Birth Register. The National Board of Health and Welfare, Sweden. http://www.socialstyrelsen.se/
202
Health and Place 57 (2019) 200–203
C. Eriksson, et al.
Wallas, A., Ekström, S., Eriksson, C., Gruzieva, O., Pyko, A., Bergström, A., et al., 2018. Road Traffic Noise Exposure and Development of Body Mass Index from Birth to Adolescence. Wickman, M., Kull, I., Pershagen, G., Nordvall, S.L., 2002. The BAMSE project: presentation of a prospective longitudinal birth cohort study. Pediatr. Allergy Immunol. 13, 11–13.
register/halsodataregister/medicinskafodelseregistret/inenglishStockholm. Twohig-Bennett, C., Jones, A., 2018. The health benefits of the great outdoors: a systematic review and meta-analysis of greenspace exposure and health outcomes. Environ. Res. 166, 628–637. WHO, 2016. Urban Green Spaces and Health. A Review of Evidence. WHO Regional office for Europe, Copenhagen.
203