Skip to main content
PLOS Medicine logoLink to PLOS Medicine
. 2020 Aug 18;17(8):e1003213. doi: 10.1371/journal.pmed.1003213

Residential green space and child intelligence and behavior across urban, suburban, and rural areas in Belgium: A longitudinal birth cohort study of twins

Esmée M Bijnens 1,2,*, Catherine Derom 2,3, Evert Thiery 4, Steven Weyers 2, Tim S Nawrot 1,5
Editor: Iana Markevych6
PMCID: PMC7446904  PMID: 32810193

Abstract

Background

Exposure to green space has beneficial effects on several cognitive and behavioral aspects. However, to our knowledge, no study addressed intelligence as outcome. We investigated whether the level of urbanicity can modify the association of residential green space with intelligence and behavior in children.

Methods and findings

This study includes 620 children and is part of the East Flanders Prospective Twin Survey (EFPTS), a registry of multiple births in the province of East Flanders, Belgium. Intelligence was assessed with the Wechsler Intelligence Scale for Children-Revised (WISC-R) in 620 children (310 twin pairs) between 7 and 15 years old. From a subset of 442 children, behavior was determined based on the Achenbach Child Behavior Checklist (CBCL). Prenatal and childhood residential addresses were geocoded and used to assign green space indicators. Mixed modeling was performed to investigate green space in association with intelligence and behavior while adjusting for potential confounding factors including sex, age, parental education, neighborhood household income, year of assessment, and zygosity and chorionicity.

We found that residential green space in association with both intelligence and behavior in children was modified by the degree of urbanicity (p < 0.001). In children living in an urban environment, multivariable adjusted mixed modeling analysis revealed that an IQR increment of residential green space (3,000-m radius) was associated with a 2.6 points (95% CI 1.4–3.9; p < 0.001) higher total intelligence quotient (IQ) and 2.0 points (95% CI −3.5 to −0.4; p = 0.017) lower externalizing behavioral score. In children residing in a rural or suburban environment, no association was found. A limitation of this study is that no information was available on school location and the potential for unmeasured confounding (e.g., time spend outdoors).

Conclusions

Our results indicate that residential green space may be beneficial for the intellectual and the behavioral development of children living in urban areas. These findings are relevant for policy makers and urban planners to create an optimal environment for children to develop their full potential.


Esmée M Bijnens and colleagues reveal beneficial effects of a green environment on a child's development and behaviour.

Author summary

Why was this study done?

  • This study examines residential surrounding green space in association with intelligence and behavior in a study area that includes a spectrum of urban to rural environments.

  • Previous studies mainly focused on urban areas, whereas only a few studies have explored differences in the effect of green space on cognition in nonurban settings.

  • Understanding the health disparities that exist between urban and rural environments is essential for maintaining and improving human well-being in a rapidly urbanizing world.

What did the researchers do and find?

  • This longitudinal birth cohort study of twins assessed intelligence in 620 children between 7 and 15 years old.

  • Our results indicate that residential green space is especially beneficial for intellectual and behavioral development of children living in an urban environment.

What do these findings mean?

  • We show that low residential green space in urban children is associated with a “shift” towards a higher incidence of low IQ demonstrating the public health impact of our findings.

Introduction

The brain develops steadily during prenatal and early postnatal periods, which are considered as the most vulnerable windows for effects of environmental exposures [1]. Prenatal exposure to air pollutants may have persisting effects on brain development in children with consequences on cognition and behavior [2]. Later, across childhood and adolescence, crucial changes in cognition, emotion, and behavior occur [3], and some specific functions related with learning and school achievement develop [4]. Environmental factors such as air pollution are associated with neuroinflammation and may impact these performances in healthy children [3]. Children residing in a highly polluted urban environment exhibited deficits in fluid intelligence, memory, executive functions, and also impairments in subscales of intelligence quotient (IQ), relative to children living in a less polluted urban environment [3]. Moreover, airborne polycyclic aromatic hydrocarbons (PAHs) are generated by the incomplete combustion of fossil fuels, and prenatal exposure is significantly associated with a reduced IQ at 5 years of age [57]. However, the effects of green space on intelligence are currently unknown.

Greenness refers to the vegetation level, whereas green spaces are covered partly by grass, trees, or other vegetation, and include city parks, community gardens, sports fields, as well as natural and forested areas in rural environments [8,9]. Green spaces provide environmental benefits by reducing air and noise pollution. Higher residential green space is associated with lower ambient air pollution [10] and noise exposure [11]. Both have been associated with diminished cognitive development [1215]. Furthermore, from a health perspective, green space promotes physical activity [16,17] and stress reduction [18]. Apart from the promotion of physical activity, green space in the living environment could also lead to more social contacts [19]. In addition, higher levels of neighborhood green space were associated with significantly lower levels of symptoms of depression, anxiety, and stress.[20] Besides self-reported stress, also salivary cortisol, an indicator of chronic stress, is negatively associated with exposure to green space in the living environment [21]. Higher residential green space during pregnancy is even associated with longer telomere length in placental tissue, an important marker of ageing [22]. Finally, according to the biodiversity hypothesis of health, the reduced contact with a natural environment and biodiversity is associated with adverse implications for the human microbiome and its immunomodulating capacity [23]. Microbial input from the environment may drive brain regulation and is a possible major component of the beneficial effect of greenness [24]. Several animal studies suggest that the human microbiome may have profound effects on brain development and linked the gut microbiota to neurodevelopmental processes and behavior [2527]. Natural space-induced environmental biodiversity has been related to enriched skin and gut microbiota that, in turn, has been shown to improve brain development, probably through regulation of immune response [28].

In a previous study of 2,593 children attending primary school in Barcelona [29], exposure to surrounding green space at the home and school area was associated with greater progress in working memory and attention [29]. The literature on green space around schools and academic performance shows mixed results [30]. Although positive associations were observed for 3 types of green space: green land cover (45% of the findings), greenness (32%), and tree cover (30%), whereas no positive findings were found for agriculture, grass, or shrub cover. Furthermore, access to nearby natural outdoor play spaces may enhance children’s executive functioning [31], and a meta-analysis revealed that the greening of schoolyards across multiple sites in North America and Western Europe has been associated with better psychological well-being and improved school performance among pupils [32].

Evidence shows that green not only results in a beneficial effect on cognitive and school performances but also in improved psychological well-being with positive effects on emotions and behavior [3335]. Higher residential surrounding green space was significant inversely associated with behavioral development in 2,111 schoolchildren from 36 schools in Barcelona [36]. Furthermore, residential green space was associated with reduced scores for hyperactivity and inattention symptoms [36]. A study in 1,932 children residing in the city of Munich and its surrounding areas showed that a higher distance between a child's residence and the nearest urban green space was positively associated with hyperactivity/inattention (odds ratio = 1.20 per 500-m increase in distance) [37]. In addition, reduced aggressive behavior was noted in association with more surrounding residential green space in urban-dwelling adolescents [38]. Few studies on green space and cognition [39] or behavior [37,40] have investigated effect modification by urbanicity but did not find statistically significant differences by urbanization of residence.

The research mentioned previously investigated green space in association with cognitive and behavioral aspects; in connection to this, we investigated the association between green space and intelligence. A major limitation is that most studies mainly focused on urban areas, whereas only a few studies have explored differences in the effect of green space on cognition in nonurban settings. Understanding the disparities that exist between urban and rural environments is important for building and sustaining an effective policy. This study examines the relation of environmental green space to intelligence and behavior in a study area that includes a spectrum of urban to rural environments across the province of East Flanders in Belgium. We hypothesize that the level of urbanicity can modify the outcome on intelligence and behavior in children associated with green space exposure in early life.

Methods

Subject recruitment

This study includes 620 children born between 1980 and 1991. This subgroup is part of the East Flanders Prospective Twin Survey (EFPTS), a registry of multiple births in the province of East Flanders, Belgium. Since its start in 1964, over 10,000 twin pairs have been registered [41,42]. The province East Flanders is located in the North-West of Belgium, and the capital Ghent is centrally located. Besides Ghent and a few urban areas, most of the province is rural (S1 Fig). Similar to the rest of Flanders, this province is densely populated with strong urban metropolitan areas and includes a lot of arable land. The capital Ghent has many parks spread throughout the city, readily accessible or within close proximity to home for most residents [43]. Parks in Flanders are used mostly as a place to relax rather than to engage in intensive sports [43,44].

In 1992 and between 1996 and 1999, we assessed intelligence in 663 twin pairs at a mean age of 10.4 years (age range 7–15) using the Wechsler Intelligence Scale for Children-Revised (WISC-R). Of the 663 pairs, we invited 599 eligible mothers to complete a questionnaire (S1 Text) on residential history. In total, 64 mother/twin pairs were lost for the present study: mothers with missing contact details (n = 7), mothers who indicated not willing to further participate (n = 20), late mothers (n = 19), and twin pairs of which one of the twins died (still born or neonatal, n = 18). The questionnaire was completed by 324 mothers (54%), and 275 mothers must be classified as nonresponders, i.e., 126 could not be reached, 126 did not fill out the questionnaire, and 23 indicated not willing to participate. We excluded 14 twin pairs from our analysis because of missing data (on residential location [n = 5] and parental educational level [n = 9]). The number of children included in our analysis was 620 individuals (310 pairs). From a subset of 442 individuals (221 twin pairs), aged 6–15 years, data are available on behavioral analysis. An overview is presented in the flowchart as supplement (S2 Fig). We compared the characteristics of those that we were able to collect information on residential history compared with nonresponders. Children included in the study differed slightly from those of which we were not able to collect data in terms of parental educational level, IQ, and externalizing behavior. However, given that the differences between responders and nonresponders were rather small, it is unlikely that this would have had a major impact on the results. This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 STROBE Checklist). The method and study design used in this study were described before the analyses were conducted, and the prospective analysis plan is present in supplement (S2 Text). Written informed consent was obtained from all participants, and ethical approval was given by the Ethics Committee of University Hospital Ghent and Hasselt University (Registration number: B670201730788).

Data collection

Data recorded by the obstetrician at birth included gestational age, birth weight, and sex of the twins. Gestational age was calculated as the number of completed weeks of pregnancy and was estimated on the basis of last menstrual period combined with real-time ultrasonography in early pregnancy. At time of birth, placentas were examined within 24 hours after delivery by a trained midwife following a standardized protocol [45]. Cord insertion categorized into 2 groups: central insertion (central, paracentral, paramarginal) and peripheral insertion (marginal, membrane septum, and membrane peripheral). Zygosity was determined by sequential analysis based on sex, choriontype, blood group determined on umbilical cord blood, placental alkaline phosphatase, and since 1982, DNA fingerprints [46]. After DNA fingerprinting, a zygosity probability of 99.9% was reached. Based on chorionicity, monozygotic twins where further divided into 2 groups, monozygotic dichorionic and monozygotic monochorionic. Monochorionic twins share only 1 placenta [47].

Further, we collected information on parental education. The highest educational level of both parents as a proxy of socioeconomic status (SES) was categorized into 2 groups according to the Belgian education system: (1) no education, primary school, lower secondary education, or higher secondary education, and (2) tertiary education. We gathered information on neighborhood socioeconomic status. Based on the home addresses, all twins were assigned to statistical sectors (average area = 1.55 km2), the smallest administrative entity for which statistical data are produced by the Belgian National Institute of Statistics (NIS). Belgian census data (FOD Economie/DG Statistiek) derived from the NIS were used to define neighborhood household income based on annual household income in the year 1994.

Green space and traffic-related exposure

Residential addresses of the children during pregnancy and during childhood at time of the IQ measurement were geocoded. All analyses were carried out using Geographic Information System (GIS) functions with ArcGIS 10 software. Residential location was categorized into urban, suburban, and rural based on a map (Flemish Government–Department Environment) containing all statistical sectors in Flanders classified as urban, suburban, and rural areas depending on their population density, employment, location, and spatial planning (S1 Fig). Seminatural, forested, blue, and urban green areas (green space) in several radius distances (5,000, 4,000, 3,000, 2,000, 1,000, and 500 m) around the residential address during childhood and during pregnancy were estimated based on CORINE (Coordination of Information of the Environment) Land Cover 1990 (European Environment Agency). Green space does not include agricultural areas. Moreover, we calculated residential green space based on a high-resolution (1 m2) land cover data set (Green Map of Flanders 2012 from the Agency for Geographic Information Flanders [AGIV]), which is available. We estimated high green (vegetation height higher than 3 m), within several radii (2,000, 1,000, 500, 300, 100, and 50 m) around the residence.

Residential distances to the nearest major road were determined. Major roads were defined by 2 types of roads, namely, freeways and national roads. We calculated the regional background levels of NO2 for the residential address during childhood using a kriging interpolation method [48,49] that uses land cover data obtained from satellite images. This model provides interpolated NO2 values from the Belgian telemetric air quality networks in 4 × 4 km grids. We calculated individual NO2 concentrations (micrograms per cubic meter) 1 year before the measurement.

Assessment of intelligence and behavior during childhood

Intelligence was assessed by the WISC-R. This test was administered separately and at random by 2 trained research workers [50]. The methods of assessment of intelligence has been previously described [51]. In brief, the WISC-R consists of 6 verbal and 6 performance subscales and has been validated for use in this population [52]. The scores on the subscales are standardized for age and added up to verbal IQ (VIQ), performance IQ (PIQ), and total IQ (TIQ). The minimum and maximum scores that can be obtained are 50 and 150 points. In this study, the VIQ, PIQ, and TIQ scores were analyzed.

Behavior was assessed by the Achenbach Child Behavior Checklist (CBCL). This checklist was developed by Achenbach (1991) to examine the extent to which children have behavioral and emotional problems as perceived by their parents [53]. Good reliability and validity have been demonstrated for the Dutch CBCL [54]. Parents were asked to answer each item in 3 levels: 0 = not true, 1 = sometimes/somewhat true, and 2 = very true/often true, with higher score indicating more adverse behavior. The sum of all questions served as a global score for children's problematic behavior (the total problems behavior score). Separate scores were calculated for internalizing problems and externalizing problems. CBCL T-scores equal to 64 or higher on the internalizing, externalizing, or total problem scales are considered in the clinical range [55]. In this study, we examined the total behavioral score, and the externalizing and internalizing subdomains. The internalizing behavioral score is formed by combining emotionally reactive, anxious/depressed, somatic complaints, and withdrawn. The externalizing behavioral score is formed by combining attention problems and aggressive behavior.

Statistical analysis

For data management and statistical analyses, we used SAS software, version 9.4 (SAS Institute, Cary, NC). All reported p-values are 2-sided and were considered statistically significant when p < 0.05. The distribution of all variables was inspected. Mixed modeling was performed to investigate intelligence/behavior in association with residential green space. The twins were analyzed as individuals in a multilevel regression analysis to account for relatedness between twin members by adding a random intercept to the model. The variance–covariance structure was allowed to differ between the 3 zygosity and chorionicity groups, including dizygotic dichorionic, monozygotic dichorionic, and monozygotic monochorionic. Mixed modeling was performed adjusted for covariates selected a priori. The models investigating the association between intelligence/behavior and green space during childhood included the covariates sex, age, parental education, neighborhood household income, year of assessment WISC-R/CBCL, and zygosity and chorionicity. We expressed the estimates for an IQR change in percentage green space for buffers from 1,000 to 5,000 m. Because of the skewedness of the data for the 500-m radius, we expressed the effect size for a dichotomized variable representing the presence or absence of green space within this buffer. To capture potential nonlinear effects of year of assessment, the quadratic terms of this variable was tested but did not reach significance. Further, to allow for nonlinearity, green space variables were categorized into tertiles. We constructed scatterplots (Fig 1), and in addition, we plotted IQ (TIQ, VIQ, and PIQ) for each tertile of green space (3,000-m buffer). We did not find evidence for deviation from linearity. Studying green space during pregnancy additional adjustment were made for perinatal factors such as birth weight, gestational age, birth year (linear and quadratic), cord insertion, and maternal smoking during pregnancy. We tested the interaction of urbanicity-by-green space interaction on IQ. Because of a significant interaction, we stratified the analysis in children residing in an urban, suburban, and rural area at time of the IQ measurements. To address the public health significance, we compared the distribution of total IQ for children residing in “high” (above median) versus “low” (below median) residential green space across the 3 urbanization levels.

Fig 1. The association between intelligence and residential green space is modified by the degree of urbanicity (p for interaction < 0.001).

Fig 1

Intelligence is shown in association with green space in a 3,000-m radius around the current residence in twins living in an urban (n = 232), suburban (n = 126), and a rural area (n = 254). Data points (n = 8) with green space >20% were excluded from the figure for visibility of the figure. (A) Total IQ, (B) verbal IQ, and (C) performance IQ. Correlation coefficients are shown for data points in figure (excluding points with green space >20%) and for all data points. IQ, intelligence quotient.

Finally, we conducted a series of sensitivity analysis to check the robustness of the findings. First, as a sensitivity analysis, we investigated the association between green space and IQ in a subset of children (n = 144) living minimum 10 years (or their whole life if the age of the child <10 years old) at the current address. Second, we adjusted the main models for regional background levels of NO2 exposure and distance to major road in a sensitivity analysis. To check the effect modification by SES, we stratified the children residing in an urban area, based on parental education, into groups of low parental education (n = 82) and high parental education (n = 154). We investigated the association between green space and IQ in children of parents with a low education level and in children of parents with a high education level. As a final sensitivity analysis, we investigated the association between intelligence and residential green space based on the high-resolution model (2012).

Results

Characteristics of the study population

Table 1 summarizes the characteristics of the study population including 310 mothers and the 620 children. Based on the residential address of the twins at time of the IQ measurements, 236 (38%) children were living in an urban environment, 128 (21%) in a suburban area, and 256 (41%) in a rural area. Overall, most parents (61.3%) were highly educated. The population comprised 312 (50.3%) boys and 308 (49.7%) girls. Our analysis included 356 (57.4%) dizygotic twins, 106 (17.1%) monozygotic-dichorionic twins, and 158 (25.5%) monozygotic-monochorionic twins. The children completed the WISC-R test at a mean age of 11.0 years (range: 7–15 years). The averaged IQ scores (±SD) were 105.3 ± 14.2 for total IQ, 106.2 ± 13.5 for verbal IQ, and 103.2 ± 15.2 for the performance IQ. Behavior was assessed by the CBCL in a subset of 442 children. The mean CBCL total score was 46.5 ± 9.9, the mean CBCL externalizing score was 46.3 ± 9.9, and the mean CBCL internalizing score was 47.7 ± 9.4. Table 2 contains the residential distribution of green space for a 500–5,000-m radius. The median residential green space ranges from 1.13% to 3.78%. In the 500-m radius, green space was present for 17% of the participants (respectively, 19%, 20%, and 14% in urban, suburban, and rural) and in the 1,000-m radius for 37% of the participants (respectively, 45%, 42%, and 27% in urban, suburban, and rural).

Table 1. Study population characteristics.

Characteristic All n = 620 Urban n = 236 Suburban n = 128 Rural n = 256
Sex, male 312 (50.3) 114 (48.3) 71 (55.5) 127 (49.6)
Birth year 1,986 ± 2.8 1,987 ± 2.7 1,987 ± 2.5 1,986 ± 2.9
Zygosity-chorionicity
    Dizygotic-dichorionic 356 (57.4) 150 (63.6) 74 (57.8) 132 (51.6)
    Monozygotic-dichorionic 106 (17.1) 40 (16.9) 14 (10.9) 52 (20.3)
    Monozygotic-monochorionic 158 (25.5) 46 (19.5) 40 (31.3) 72 (28.1)
Parental education
    Low 120 (38.7) 41 (34.8) 4 (6.3) 59 (46.1)
    High 190 (61.3) 77 (65.2) 44 (68.8) 69 (53.9)
Neighborhood income, euro 18,831 ± 2,481 18,898 ± 2,426 19,755 ± 2,848 18,308 ± 2,196
Intelligence (WISC-R) n = 620 n = 236 n = 128 n = 256
Age at investigation 11.0 ± 1.8 11.0 ± 1.7 10.9 ± 1.7 11.1 ± 1.9
Year of investigation 1,998 ± 1.6 1,998 ± 1.4 1,998 ± 1.6 1,998 ± 1.7
Intelligence score (IQ)
    Total 105.3± 14.2 106.2 ± 13.2 104.8 ± 14.2 104.8 ± 15.1
    Verbal 106.2 ± 13.5 107.0 ± 12.9 106.1± 13.3 105.5 ± 14.2
    Performance 103.2 ± 15.2 103.9 ± 14.1 102.4 ± 15.7 103.0 ± 15.9
Behavior (CBCL) n = 442 n = 168 n = 92 n = 182
Age at investigation 8.5 ± 2.2 8.4 ± 2.2 8.1 ± 2.0 8.7 ± 2.3
Year of investigation 1,994 ± 0.5 1,994 ± 0.5 1,994 ± 0.5 1,994 ± 0.5
CBCL score
    Total 46.5 ± 9.9 47.0 ± 9.5 46.8 ± 9.7 45.9 ± 10.3
    Externalizing 46.3 ± 9.9 47.2 ± 9.6 46.2 ± 10.1 45.4 ± 10.1
    Internalizing 47.7 ± 9.4 48.0 ± 9.4 48.2 ± 9.4 47.0 ± 9.4

Data presented are means ± standard deviation or number (percentage).

CLBL, Child Behavior Checklist; WISC-R, Wechsler Intelligence Scale for Children-Revised.

Table 2. Distribution of green space.

Green space Urban Suburban Rural
25th percentile Median 75th percentile 25th percentile Median 75th percentile 25th percentile Median 75th percentile
5,000-m buffer, % 2.88 3.44 4.86 1.96 3.32 5.61 1.97 3.45 6.06
4,000-m buffer, % 2.56 3.78 4.91 1.63 3.39 5.99 1.45 3.09 6.13
3,000-m buffer, % 2.11 3.38 5.39 1.29 2.82 6.11 0.19 2.95 5.24
2,000-m buffer, % 0.25 3.01 5.96 0 3.2 7.33 0 1.13 4.09
1,000-m buffer, % 0 0 4.85 0 0 6.56 0 0 0.3
500-m buffer, % 0 0 0 0 0 0 0 0 0

Green space was defined as seminatural, forested, blue, and urban green areas based on CORINE Land Cover 1990 (European Environment Agency).

CORINE, Coordination of Information of the Environment.

Intelligence in association with green space in childhood

We noticed (Fig 1) a significant urbanicity-by-green space interaction on twin IQ (p for interaction <0 .001). We therefore stratified the analysis by degree of urbanicity. Higher residential surrounding green space is associated with intelligence in children living in an urban environment before (Fig 1) and after adjustment (Fig 2) for sex, age, parental education, neighborhood household income, year of assessment of intelligence, and zygosity and chorionicity. This was not observed among children living in a rural or suburban environment. In urban areas, residential green space within a 1,000- to 5,000-m buffer is significantly associated with TIQ and PIQ, whereas the association between urban green space and verbal IQ was significant in a 2,000- to 5,000-m buffer and not in a 500-m and 1,000-m buffer. Taking the percentage green within a 3,000-m buffer as a reference, we found that for an IQR (3.3%) increase in green space within a 3,000-m radius from the residential address is associated with a 2.6-point higher (95% CI 1.4–3.9; p < 0.001) TIQ, a 2.2-point higher (95% CI 0.9–3.4; p = 0.0008) VIQ, and a 2.4-point higher (95% CI 1.0–3.8; p = 0.0014) PIQ among children residing in an urban environment. The full results of the model with all the covariates are shown in supplement (S1 Table).

Fig 2.

Fig 2

Estimated change in IQ for an IQR increase in green space in a 500–5,000-m radius around the residence in twins living in (A) an urban (n = 234), (B) a suburban (n = 128), and (C) a rural area (n = 252) adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. We expressed the estimates for an IQR change in percentage green space for buffers from 1,000 to 5,000 m. Because of the skewedness of the data for the 500-m radius (and 1,000 m in rural), we expressed the effect size for a dichotomized variable representing the presence or absence of green space within this buffer. IQ, intelligence quotient; IQR, interquartile range.

Fig 3 shows the cumulative frequency distributions of total IQ scores in children residing in high green versus low green urban areas. Children in low green areas are more likely (4.2% versus 0%) to have a total IQ of 80 or lower. Furthermore, although 11.9% of those living in a green area had IQ scores in the superior range (>125), only 4.2% of the children living in a low green area had scored in this range. We observed no difference in cumulative frequency distribution between high green and low green in children living in either rural or suburban areas.

Fig 3.

Fig 3

This graph displays the cumulative frequency distribution of total IQ scores in children residing in high and low green space (based on the median) in twins living in (A) an urban, (B) a suburban, and (C) a rural area. IQ, intelligence quotient; TIQ, total intelligence quotient.

Besides green space at the residential address in childhood, we also observed a significant association between green space during pregnancy with intelligence in children living in an urban area (S3 Fig).

To study the robustness of our findings, we performed multiple sensitivity analysis. The results remain similar in a subpopulation of children (n = 144) living a minimum of 10 years at the current address (S4 Fig). Additional adjustment for distance to major roads and regional background levels of NO2 did not result in a considerable change in the aforementioned effect estimates (S5 Fig). We performed stratified analysis to study the potential effect modification by parental education. Patterns were similar according the parental educational level, but overall, we found a tendency of more pronounced associations between residential green space and IQ outcomes in urban children of parents with higher educational background (S6 Fig). Further, we investigated the association between intelligence and residential high green (vegetation height higher than 3 m) based on a high-resolution model. Our results (S7 Fig) show that high green is positively associated with VIQ and TIQ in large buffer sizes (> 500 m).

Child behavior in association with green space

Similarly, higher residential green space is associated with a reduction in behavioral problems in children living in an urban area, whereas no effect was noted in children living in a rural or suburban area (Fig 4). Percentage of residential green space within a 1,000-m to 3,000-m radius from the residential address in an urban environment was significantly associated with a reduction in total problem behavior, green space within a 500-m to 4,000-m radius is significantly associated with a reduction in externalizing problem behavior (including attention problems and aggressive behavior) and green space in a 3,000-m radius is associated with lower internalizing problem behavior (including anxiety and withdrawal). An IQR (3.6%) increase in green space within a 3,000-m radius from the residential address is associated with a 2.0 (95% CI −3.6 to −0.4; p = 0.015) lower CBCL total score and a 2.0 (95% CI −3.5 to −0.4; p = 0.017) lower CBCL externalizing score among urban-residing children.

Fig 4. Estimated change in CBCL score for an IQR increase in green space in a 500–5,000-m radius around the residence.

Fig 4

In twins living in (A) an urban (n = 168), (B) a suburban (n = 92), and (C) a rural area (n = 182). Adjusted for sex, age, parental education, neighborhood household income, month of CBCL, and zygosity and chorionicity. We expressed the estimates for an IQR change in percentage green space for buffers from 1,000 to 5,000 m. Because of the skewedness of the data for the 500-m radius (and 1,000 m in rural), we expressed the effect size for a dichotomized variable representing the presence or absence of green space within this buffer. CBCL, Child Behavior Checklist; IQR, interquartile range.

Discussion

The key finding of our analysis is that a higher percentage of residential green space is associated with higher intelligence and lower behavioral problems in 7–15-year-old children living in urban areas. More particularly, the level of urbanicity modified the association of green space exposure with intelligence and behavior in childhood.

To our knowledge, this is the first study investigating the association between residential green space and intelligence in children. Previous studies have already shown that urban green is important for cognitive development in children by improving working memory, attention, and school performance. A study in schoolchildren in Barcelona [29] shows that school and total surrounding green space were associated with enhanced progress in working memory and superior working memory and a greater reduction in inattentiveness. The same investigators found in a subcohort of 253 children, using high-resolution 3-D MRIs, several brain regions that had larger volumes in association with higher lifelong exposure to residential surrounding greenness [56]. In a prospective birth cohort study, Dadvand and colleagues established that higher lifelong residential surrounding green space was associated with better scores on tests of attention at 4–5 years and 7 years of age [57]. In addition, based on school-level student performance data from 905 public schools in Massachusetts (the second most tree-covered state) collected between 2006 and 2012 (n = 6,333), the effect of surrounding greenness on the school performance was investigated [34]. The results showed that children (age 8–9 years) with higher exposure to greenness show better performance in both English and math. However, this study was replicated in Chicago, in an area with low tree coverage, and found no convincing evidence for a positive relationship between residential greenness and academic performance [58]. The authors suggest that a main reason may be that greenness, the general degree of vegetation, based on NDVI (normalized difference vegetation index) fails to distinguish between vegetation types, such as grasses and trees. In connection to this, Kweon and colleagues found a positive association between tree cover in the school surrounding and academic performance, but grass and shrubs negatively correlate with performance [59].

Few studies on green space and academic performance have specifically tested for effect modification by urbanicity. A cross-sectional study in Maryland public schools tested the relationship between 5 greenspace measures and performance (n = 668) around schools and in children's neighborhoods [39]. Including interaction terms in the models showed no clear effect modification by urbanicity. We show that an IQR (3.3%) increase in green space within a 3,000-m radius from the residential address is associated with a 2.6-point higher (95% CI 1.4–3.9; p < 0.001) total IQ in childhood. The magnitude of our estimates show that our findings are relevant for public health. Indeed, although the contribution on the average population IQ might be moderate, children with low residential green space are more likely to gravitate to the lower tail of the IQ curve. Therefore, low residential green space in children living in an urban area might cause an inevitable “shift” in population IQ. This has been demonstrated by Needleman for lead exposure during childhood, showing that a greater proportion of children with high lead exposure gravitate to the lower tail of the normal IQ curve [60], and here, we show the same phenomena for low residential green space exposure. Indeed, 4.2% of the children had an IQ below 80 growing up in low urban green space areas versus 0% in high green space. This shift towards a higher incidence of a low IQ demonstrates the public health impact of our findings at an unprecedented level.

Besides green space at the residential address in childhood, we also observed a significant association between green space during pregnancy with intelligence in children living in an urban area. Until now, only 1 study in 1,312 women and their children investigated surrounding green spaces at the residential address at birth. It shows that exposure to higher levels of surrounding green spaces is associated with better early childhood cognition at the age of 2 [61].

Our findings show that residential green space is associated with a reduction in total and mostly externalizing problem behavior (including attention problems and aggressive behavior) in 7–15-year -old children living in a urban area. This in line with other studies on green space using the CBCL to assess behavior. School children 6–18 years of age living in the highest tertile of greenness had a lower total CBCL score, especially aggressive behavior and attention problems, compared with children living in the lowest tertile of greenness [40]. As in our study, the researchers found a slightly stronger association between greenness and externalizing behavior than with internalizing behavior [40]. In urban-dwelling adolescents, exposure to neighborhood greenspace is associated with reducing aggressive behavior as reported by parents on the aggression subscale of the CBCL [38]. A systematic review present evidence for an association between green space exposure and the total difficulties score of the Strengths and Difficulties Questionnaire (SDQ), a measure of general mental health in children [62]. Regarding the different subdomains of the SDQ, 5 out of 6 studies documented significant associations with hyperactivity and inattention problems in association with green space exposure [36,37,6365]. A study in 4–13-year-old children shows that green space quality and quantity were associated with internalizing and externalizing subscale scores for all age groups [66]. The participants in the study of Markevych and colleagues (2014) were spatially spread over the inner city of Munich and its surrounding areas [37]. The analyses stratified by whether or not a child lived in the inner city, showed that the associations between urban green spaces and lower behavioral problems was slightly stronger for children living in the inner city (n = 782) compared with those living in the surrounding suburban areas (n = 1,150); however, the interaction-term was not significant.

Several reasons can be postulated why no association is found between green space and IQ as well as behavioral problem scores in suburban and rural residents compared with urban residents. First, Markevych and colleagues (2014) suggest that there is a general lack of green spaces in urban areas, and thus, parks play a more important role for urban residents than for those living in suburban or rural areas, in which natural green spaces are more prevalent [37]. A previous study [67] shows that the association between green space and health varied according to the combination of urbanicity and area income deprivation. A possible explanation is that high-income residents in suburban and rural areas have their own domestic gardens, and municipal green space is thus less important to them [67]. Finally, rural areas might have fewer everyday destinations (e.g., grocery store, post office, and pharmacy) within walking distance, resulting in more car dependency and less active transportation [68]. Therefore, urban residents may have more benefits of green space in their neighborhood.

The underlying pathways responsible for the positive effect of residential green space exposure on intelligence is not fully understood, but several mechanisms are considered. First, green spaces provide environmental benefits (e.g., reducing exposure to air pollution, noise, and heat). Moreover, green spaces encourage health-promoting activities and facilitate social cohesion. A third potential domain linking green space to health is the ability of green space to restore capacities (e.g., attention restoration and physiological stress recovery) [68].

Our results were observed in twins. We do not expect that twins have a different exposure or react differently to green space than singletons. Studies show that twins score 4 points lower on an IQ test than singletons [69]. This difference is partially attributed to difference in intrauterine growth [69], nutrient deficiency, and shorter gestational length of twins [70]. In another study, however, no evidence on cognitive differences was found between adult twins and their nontwin siblings [71]. Besides, lower birth weight is a causal risk factor for child problem behavior [72]. However, a study comparing the developmental trajectories of behavioral problems in 6- to 12-year-old twins and singletons confirmed the generalizability with regard to externalizing problems in childhood of twin studies to singleton populations [73]. Therefore, we expect no large differences in regard to singletons.

We recognize limitations of the current study. First, surrounding green space was assessed based on residential location; however, no information on school location was available. However, children in Flanders (Belgium) usually attend schools near their home, on average, 2.8 km. Therefore, a buffer distance such as 3,000 m and more may reflect the full exposure to green space that is part of the residence and the school neighborhood where children likely spend their time [34]. Second, the twins were born from 1980 to 1991; however, to determine residential green space during pregnancy, we used data from 1990 as no earlier satellite data were available. However, strong correlations have been shown over time between periods of 1 decade [22]. The association between green space and the outcomes IQ and behavior was investigated in several radius distances around the residential address. We found that the positive impact of green in an urban setting for both IQ and behavioral scores depended on the percentage green within 2,000 and 3,000 m. It was unfeasible to choose buffers smaller than 500 m given the resolution of CORINE Land Cover data. However, our results were comparable when we use high-resolution maps (S7 Fig). Third, no information was available on time–activity patterns, such as time spend outdoors, and on possible mediators between green space and intelligence. An important strength of our study is that we excluded potential confounding by socioeconomic indicators and traffic-related air pollution. Indeed, we presented our associations independently of socioeconomic indicators both at the individual level, such as parental education, as well as by group level by including neighborhood household income. Furthermore, in a sensitivity analysis, we account for residential air pollution and residential distance to major road as a proxy for traffic-related noise and air pollution. The proxy of residential distance to a major road is currently still used in environmental epidemiology [7476] and is correlated with black carbon load in children’s urine [77].

The results of this study provide important policy and public health implications. Whereas in 1950, only 30% of the world’s population lived in urban areas; nowadays, this is already more than half of the global population, and it is expected to increase to 68% by 2050 [78]. Urban areas are characterized by a network of nonnatural built-up infrastructures where residents often have limited access to natural environments in their daily lives [56,79]. Understanding the health disparities that exist between urban and rural environments is essential for maintaining and improving human well-being in a rapidly urbanizing world [33].

Conclusions

Our results indicate that residential green space may be beneficial for intellectual and behavioral development of children living in an urban environment. An important contribution of this paper is that to our knowledge this is the first study on intelligence and residential green space. We showed a shift in the IQ distribution of urban children in association with residential green space exposure within 3,000-m radius. These findings are relevant for policy makers and urban planners designing urban environments resulting in a sustaining effective policy. Well-planned cities can offer unique opportunities to create an optimal environment for children to develop their full potential.

Supporting information

S1 STROBE Checklist. The checklist of STROBE guidelines.

STROBE, strengthening the reporting of observational studies in epidemiology.

(DOCX)

S1 Text. Questionnaire.

(PDF)

S2 Text. A priori statistical analysis plan.

(PDF)

S1 Table. Full model with all covariates of the association between green space in 3,000-m radius around the residence and intelligence.

CI, confidence interval; IQR, interquartile range; TIQ, total intelligence quotient; TPIQ, performance intelligence quotient; TVIQ, verbal intelligence quotient. *low versus high.

(DOCX)

S1 Fig. Statistical sectors in East Flanders classified as urban, suburban and rural areas depending on their population density, employment, location, and spatial planning.

Source: Flemish Government–Department Environment.

(TIF)

S2 Fig. Flowchart.

From the children with IQ data available, 599 mothers were invited to complete a questionnaire on residential history. A total of 324 mothers completed the questionnaire. From the twin pairs, we excluded 14 pairs because of missing data on residential location or parental education resulting in a final IQ study sample of 620 children. For a subset of 442 children, aged 6–15 years, data are available for studying behavioral characteristics. IQ, intelligence quotient.

(TIF)

S3 Fig. Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence during pregnancy and in childhood.

Dark blue: Residential green space in childhood at time of IQ measurement (n = 236) adjusted for sex, age, parental education, neighborhood household income, year of IQ test (linear and quadratic), and zygosity and chorionicity. Red: Residential green space during pregnancy (n = 216) adjusted for sex, gestational age, birth weight, birth year (linear and quadratic), zygosity and chorionicity, cord insertion, parental education, neighborhood household income, and smoking during pregnancy. IQ, intelligence quotient; IQR, interquartile range.

(TIF)

S4 Fig. Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence in children living in an urban (n = 236) area and in a subgroup of children living minimum 10 years (or their whole life if the age of the child <10 years old) at their current address (n = 144).

Adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. IQ, intelligence quotient; IQR, interquartile range.

(TIF)

S5 Fig. Main model additionally adjusted for distance to major road and NO2 exposure.

Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence in children living in an urban (n = 236) area. The main model is adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. We additionally adjust for distance to major road and regional background levels of NO2 exposure the year before IQ test. IQ, intelligence quotient. IQ, intelligence quotient; IQR, interquartile range; NO2, nitrogen dioxide.

(TIF)

S6 Fig. Effect modification socioeconomic status.

Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence in children living in an urban area. Based on parental education, we stratified the children residing in an urban area into groups low parental education (n = 82) and high parental education (n = 154). Adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. IQ, intelligence quotient; IQR, interquartile range.

(TIF)

S7 Fig. Estimated change in IQ for an IQR increase in green space (high-resolution model 2012) in a 50–2,000-m radius around the residence.

Green space is high green (vegetation height higher than 3m, including blue space). In twins living in (A) an urban (n = 236), (B) a suburban (n = 128), and (C) a rural area (n = 256). Adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. IQ, intelligence quotient; IQR, interquartile range.

(TIF)

Abbreviations

CBCL

Child Behavior Checklist

CORINE

Coordination of Information of the Environment

IQ

intelligence quotient

IQR

interquartile range

NDVI

normalized difference vegetation index

PAH

airborne polycyclic aromatic hydrocarbon

PIQ

performance intelligence quotient

SDQ

Strengths and Difficulties Questionnaire

TIQ

total intelligence quotient

VIQ

verbal intelligence quotient

WISC-R

Wechsler Intelligence Scale for Children-Revised

Data Availability

Green space data (CORINE land cover) used in this study can be obtained from the European Environment Agency (EEA) (https://land.copernicus.eu/pan-european/corine-land-cover). Air pollution data used in this study can be obtained from IRCEL (https://www.irceline.be/en). Access to and use of summarized health data of study participants requires submission of a research proposal to the East Flanders Prospective Twin Survey, Ghent University, contact email, twins@uzgent.be.

Funding Statement

Dr. Bijnens holds a fellowship from the Marguerite-Marie Delacroix foundation. Since its start, the East Flanders Prospective Twin Survey has been partly supported by grants from the Fund of Scientific Research Flanders and Twins, a nonprofit Association for Scientific Research in Multiple Births (Belgium). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

  • 1.Grandjean P, Landrigan PJ. Neurobehavioural effects of developmental toxicity. Lancet Neurol. 2014;13(3):330–8. Epub 2014/02/22. 10.1016/S1474-4422(13)70278-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Peterson BS, Rauh VA, Bansal R, Hao X, Toth Z, Nati G, et al. Effects of prenatal exposure to air pollutants (polycyclic aromatic hydrocarbons) on the development of brain white matter, cognition, and behavior in later childhood. JAMA psychiatry. 2015;72(6):531–40. Epub 2015/03/26. 10.1001/jamapsychiatry.2015.57 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Calderon-Garciduenas L, Mora-Tiscareno A, Ontiveros E, Gomez-Garza G, Barragan-Mejia G, Broadway J, et al. Air pollution, cognitive deficits and brain abnormalities: a pilot study with children and dogs. Brain Cogn. 2008;68(2):117–27. Epub 2008/06/14. 10.1016/j.bandc.2008.04.008 [DOI] [PubMed] [Google Scholar]
  • 4.Ullman H, Almeida R, Klingberg T. Structural maturation and brain activity predict future working memory capacity during childhood development. J Neurosci. 2014;34(5):1592–8. Epub 2014/01/31. 10.1523/JNEUROSCI.0842-13.2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Edwards SC, Jedrychowski W, Butscher M, Camann D, Kieltyka A, Mroz E, et al. Prenatal exposure to airborne polycyclic aromatic hydrocarbons and children's intelligence at 5 years of age in a prospective cohort study in Poland. Environ Health Perspect. 2010;118(9):1326–31. Epub 2010/04/22. 10.1289/ehp.0901070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Perera F, Li TY, Lin C, Tang D. Effects of prenatal polycyclic aromatic hydrocarbon exposure and environmental tobacco smoke on child IQ in a Chinese cohort. Environ Res. 2012;114:40–6. Epub 2012/03/06. 10.1016/j.envres.2011.12.011 [DOI] [PubMed] [Google Scholar]
  • 7.Perera FP, Li Z, Whyatt R, Hoepner L, Wang S, Camann D, et al. Prenatal airborne polycyclic aromatic hydrocarbon exposure and child IQ at age 5 years. Pediatrics. 2009;124(2):e195–202. Epub 2009/07/22. 10.1542/peds.2008-3506 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Vienneau D, de Hoogh K, Faeh D, Kaufmann M, Wunderli JM, Roosli M. More than clean air and tranquillity: Residential green is independently associated with decreasing mortality. Environ Int. 2017;108:176–84. Epub 2017/09/02. 10.1016/j.envint.2017.08.012 [DOI] [PubMed] [Google Scholar]
  • 9.Taylor L, Hochuli. Defining greenspace: Multiple uses across multiple disciplines. Landsc Urban Plan. 2017;158:25–38. 10.1016/j.landurbplan.2016.09.024 [DOI] [Google Scholar]
  • 10.Dadvand P, de NA, Figueras F, Basagana X, Su J, Amoly E, et al. Green space, health inequality and pregnancy. Environ Int. 2012;40:110–5. 10.1016/j.envint.2011.07.004 [DOI] [PubMed] [Google Scholar]
  • 11.Gidlöf-Gunnarsson A ÖE. Noise and well-being in urban residential environments: The potential role of perceived availability to nearby green areas. Landsc Urban Plan. 2017;83(2–3):115–26. 10.1016/j.landurbplan.2007.03.003 [DOI] [Google Scholar]
  • 12.Calderon-Garciduenas L, Leray E, Heydarpour P, Torres-Jardon R, Reis J. Air pollution, a rising environmental risk factor for cognition, neuroinflammation and neurodegeneration: The clinical impact on children and beyond. Rev Neurol (Paris). 2016;172(1):69–80. Epub 2016/01/01. 10.1016/j.neurol.2015.10.008 [DOI] [PubMed] [Google Scholar]
  • 13.Klatte M, Bergstrom K, Lachmann T. Does noise affect learning? A short review on noise effects on cognitive performance in children. Front Psychol. 2013;4:578 Epub 2013/09/07. 10.3389/fpsyg.2013.00578 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sunyer J, Esnaola M, Alvarez-Pedrerol M, Forns J, Rivas I, Lopez-Vicente M, et al. Association between Traffic-Related Air Pollution in Schools and Cognitive Development in Primary School Children: A Prospective Cohort Study. PLoS Med. 2015;12(3):e1001792 Epub 2015/03/04. 10.1371/journal.pmed.1001792 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Saenen ND, Provost EB, Viaene MK, Vanpoucke C, Lefebvre W, Vrijens K, et al. Recent versus chronic exposure to particulate matter air pollution in association with neurobehavioral performance in a panel study of primary schoolchildren. Environ Int. 2016;95:112–9. Epub 2016/08/31. 10.1016/j.envint.2016.07.014 [DOI] [PubMed] [Google Scholar]
  • 16.Almanza E, Jerrett M, Dunton G, Seto E, Pentz MA. A study of community design, greenness, and physical activity in children using satellite, GPS and accelerometer data. Health Place. 2012;18(1):46–54. Epub 2012/01/17. 10.1016/j.healthplace.2011.09.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ward JS, Duncan JS, Jarden A, Stewart T. The impact of children's exposure to greenspace on physical activity, cognitive development, emotional wellbeing, and ability to appraise risk. Health Place. 2016;40:44–50. Epub 2016/05/15. 10.1016/j.healthplace.2016.04.015 [DOI] [PubMed] [Google Scholar]
  • 18.Lee AC, Jordan HC, Horsley J. Value of urban green spaces in promoting healthy living and wellbeing: prospects for planning. Risk Manag Healthc Policy. 2015;8:131–7. Epub 2015/09/09. 10.2147/RMHP.S61654 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Maas J, van Dillen SME, Verheij RA, Groenewegen PP. Social contacts as a possible mechanism behind the relation between green space and health. Health Place. 2009;15(2):586–95. Epub 2008/11/22. 10.1016/j.healthplace.2008.09.006 [DOI] [PubMed] [Google Scholar]
  • 20.Beyer KM, Kaltenbach A, Szabo A, Bogar S, Nieto FJ, Malecki KM. Exposure to neighborhood green space and mental health: evidence from the survey of the health of Wisconsin. Int J Environ Res Public Health. 2014;11(3):3453–72. Epub 2014/03/26. 10.3390/ijerph110303453 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Ward Thompson C, Roe J, Aspinall P, Mitchell R, Clow A, Miller D. More green space is linked to less stress in deprived communities: Evidence from salivary cortisol patterns. Landsc Urban Plan. 2012;105(3):221–9. 10.1016/j.landurbplan.2011.12.015 [DOI] [Google Scholar]
  • 22.Bijnens E, Zeegers MP, Gielen M, Kicinski M, Hageman GJ, Pachen D, et al. Lower placental telomere length may be attributed to maternal residential traffic exposure; a twin study. Environ Int. 2015;79:1–7. Epub 2015/03/11. 10.1016/j.envint.2015.02.008 [DOI] [PubMed] [Google Scholar]
  • 23.Haahtela T, Holgate S, Pawankar R, Akdis CA, Benjaponpitak S, Caraballo L, et al. The biodiversity hypothesis and allergic disease: world allergy organization position statement. World Allergy Organ J. 2013;6(1):3 Epub 2013/05/15. 10.1186/1939-4551-6-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Rook GA. Regulation of the immune system by biodiversity from the natural environment: an ecosystem service essential to health. Proc Natl Acad Sci U S A. 2013;110(46):18360–7. Epub 2013/10/25. 10.1073/pnas.1313731110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Sudo N, Chida Y, Aiba Y, Sonoda J, Oyama N, Yu XN, et al. Postnatal microbial colonization programs the hypothalamic-pituitary-adrenal system for stress response in mice. J Physiol. 2004;558(Pt 1):263–75. Epub 2004/05/11. 10.1113/jphysiol.2004.063388 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Diaz Heijtz R, Wang S, Anuar F, Qian Y, Bjorkholm B, Samuelsson A, et al. Normal gut microbiota modulates brain development and behavior. Proc Natl Acad Sci U S A. 2011;108(7):3047–52. Epub 2011/02/02. 10.1073/pnas.1010529108 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Clarke G, Grenham S, Scully P, Fitzgerald P, Moloney RD, Shanahan F, et al. The microbiome-gut-brain axis during early life regulates the hippocampal serotonergic system in a sex-dependent manner. Mol Psychiatry. 2013;18(6):666–73. Epub 2012/06/13. 10.1038/mp.2012.77 [DOI] [PubMed] [Google Scholar]
  • 28.Sandifer PA, Sutton-Grier AE, Ward BP. Exploring connections among nature, biodiversity, ecosystem services, and human health and well-being: Opportunities to enhance health and biodiversity conservation. Ecosyst Serv. 2015;12:1–15. 10.1016/j.ecoser.2014.12.007 [DOI] [Google Scholar]
  • 29.Dadvand P, Nieuwenhuijsen MJ, Esnaola M, Forns J, Basagana X, Alvarez-Pedrerol M, et al. Green spaces and cognitive development in primary schoolchildren. Proc Natl Acad Sci U S A. 2015;112(26):7937–42. Epub 2015/06/17. 10.1073/pnas.1503402112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Browning M, Rigolon A. School Green Space and Its Impact on Academic Performance: A Systematic Literature Review. Int J Environ Res Public Health. 2019;16(3). Epub 2019/02/06. 10.3390/ijerph16030429 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Evans GW. Child development and the physical environment. Annu Rev Psychol. 2006;57:423–51. Epub 2005/12/02. 10.1146/annurev.psych.57.102904.190057 [DOI] [PubMed] [Google Scholar]
  • 32.Bell AC, Dyment JE. Grounds for health: the intersection of green school grounds and health-promoting schools. Environ Edu Res. 2008;14(1):77–90. 10.1080/13504620701843426 [DOI] [PubMed] [Google Scholar]
  • 33.Keniger LE, Gaston KJ, Irvine KN, Fuller RA. What are the benefits of interacting with nature? Int J Environ Res Public Health. 2013;10(3):913–35. Epub 2013/03/08. 10.3390/ijerph10030913 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Wu CD, McNeely E, Cedeno-Laurent JG, Pan WC, Adamkiewicz G, Dominici F, et al. Linking student performance in Massachusetts elementary schools with the "greenness" of school surroundings using remote sensing. PLoS ONE. 2014;9(10):e108548 Epub 2014/10/14. 10.1371/journal.pone.0108548 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Van Aart CJC, Michels N, Sioen I, De Decker A, Bijnens EM, Janssen BG, et al. Residential landscape as a predictor of psychosocial stress in the life course from childhood to adolescence. Environ Int. 2018;120:456–63. Epub 2018/08/27. 10.1016/j.envint.2018.08.028 [DOI] [PubMed] [Google Scholar]
  • 36.Amoly E, Dadvand P, Forns J, Lopez-Vicente M, Basagana X, Julvez J, et al. Green and blue spaces and behavioral development in Barcelona schoolchildren: the BREATHE project. Environ Health Perspect. 2014;122(12):1351–8. Epub 2014/09/10. 10.1289/ehp.1408215 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Markevych I, Tiesler CM, Fuertes E, Romanos M, Dadvand P, Nieuwenhuijsen MJ, et al. Access to urban green spaces and behavioural problems in children: Results from the GINIplus and LISAplus studies. Environ Int. 2014;71:29–35. Epub 2014/06/24. 10.1016/j.envint.2014.06.002 [DOI] [PubMed] [Google Scholar]
  • 38.Younan D, Tuvblad C, Li L, Wu J, Lurmann F, Franklin M, et al. Environmental Determinants of Aggression in Adolescents: Role of Urban Neighborhood Greenspace. J Am Acad Child Adolesc Psychiatry. 2016;55(7):591–601. Epub 2016/06/28. 10.1016/j.jaac.2016.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Browning HEM, Locke DH. The greenspace-academic performance link varies by remote sensing measure and urbanicity around Maryland public schools. Landsc Urban Plan. 2020;195 10.1016/j.landurbplan.2019.103706 [DOI] [Google Scholar]
  • 40.Lee M, Kim S, Ha M. Community greenness and neurobehavioral health in children and adolescents. Sci Total Environ. 2019;672:381–8. Epub 2019/04/09. 10.1016/j.scitotenv.2019.03.454 [DOI] [PubMed] [Google Scholar]
  • 41.Derom C, Thiery E, Peeters H, Vlietinck R, Defoort P, Frijns JP. The East Flanders Prospective Twin Survey (EFPTS): an actual perception. Twin Res Hum Genet. 2013;16(1):58–63. Epub 2012/10/30. 10.1017/thg.2012.75 [DOI] [PubMed] [Google Scholar]
  • 42.Derom C, Thiery E, Rutten BPF, Peeters H, Gielen M, Bijnens E, et al. The East Flanders Prospective Twin Survey (EFPTS): 55 Years Later. Twin Res Hum Genet. 2019:1–6. Epub 2019/09/10. 10.1017/thg.2019.64 [DOI] [PubMed] [Google Scholar]
  • 43.Van Hecke L, Van Cauwenberg J, Clarys P, Van Dyck D, Veitch J, Deforche B. Active Use of Parks in Flanders (Belgium): An Exploratory Observational Study. Int J Environ Res Public Health. 2016;14(1). Epub 2017/01/04. 10.3390/ijerph14010035 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Van Dyck D, Sallis JF, Cardon G, Deforche B, Adams MA, Geremia C, et al. Associations of neighborhood characteristics with active park use: an observational study in two cities in the USA and Belgium. Int J Health Geogr. 2013;12:26 Epub 2013/05/08. 10.1186/1476-072X-12-26 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Derom R, Derom C, Vlietinck R. Placentation. In: Keith L, Papiernik E, Keith D, Luke B, editors. Multiple pregnancy: Epidemiology, Gestation & Perinatal outcome New York: The Parthenon Publishing Group; 1995. p. 113–28. [Google Scholar]
  • 46.Vlietinck R. Determination of the zygosity of twins [thesis]. Leuven: KU Leuven; 1986. [Google Scholar]
  • 47.Loos RJ, Ridgway CL, Ong K. Theoretical underpinning of the use of twin studies in life course epidemiology Family matters: Designing, analysing and understanding family based studies in life course epidemiology. Oxford: Oxford University Press; 2009. p. 57–84. [Google Scholar]
  • 48.Janssen S, Dumont G, Fierens F, Mensink C. Spatial interpolation of air pollution measurements using CORINE land cover data. Atmos Environ. 2008;42(20):4884–903. 10.1016/j.atmosenv.2008.02.043 [DOI] [Google Scholar]
  • 49.Jacobs L, Emmerechts J, Mathieu C, Hoylaerts MF, Fierens F, Hoet PH, et al. Air pollution related prothrombotic changes in persons with diabetes. Environ Health Perspect. 2010;118(2):191–6. 10.1289/ehp.0900942 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Jacobs N, Rijsdijk F, Derom C, Danckaerts M, Thiery E, Derom R, et al. Child psychopathology and lower cognitive ability: a general population twin study of the causes of association. Mol Psychiatry. 2002;7(4):368–74. Epub 2002/05/03. 10.1038/sj.mp.4000971 [DOI] [PubMed] [Google Scholar]
  • 51.Bijnens EM, Derom C, Weyers S, Janssen BG, Thiery E, Nawrot TS. Placental mitochondrial DNA content is associated with childhood intelligence. J Transl Med. 2019;17(1):361 Epub 2019/11/11. 10.1186/s12967-019-2105-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Wechsler D. Wechsler Intelligence Scale for Children-R (Dutch version): Swets and Zeitlinger B. V., Lisse, the Netherlands; 1986. [Google Scholar]
  • 53.Achenbach T. Integrative guide for the 1991 CBCL/4–18,YSR, and TRF profiles. Burlington (VT): Department of Psychiatry, University of Vermont; 1991. [Google Scholar]
  • 54.Verhulst FC, van der Ende J, Koot H. Dutch Manual for the CBCL/4-18. Rotterdam: Afdeling Kinder- en Jeugdpsychiatrie, Sophia Kinderziekenhuis Academisch Ziekenhuis Rotterdam, Erasmus Universiteit Rotterdam; 1996. [Google Scholar]
  • 55.Achenbach T, Rescorla L. Manual for the ASEBA School-Age Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families; 2001. [Google Scholar]
  • 56.Dadvand P, Pujol J, Macia D, Martinez-Vilavella G, Blanco-Hinojo L, Mortamais M, et al. The Association between Lifelong Greenspace Exposure and 3-Dimensional Brain Magnetic Resonance Imaging in Barcelona Schoolchildren. Environ Health Perspect. 2018;126(2):027012 Epub 2018/03/06. 10.1289/EHP1876 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Dadvand P, Tischer C, Estarlich M, Llop S, Dalmau-Bueno A, Lopez-Vicente M, et al. Lifelong Residential Exposure to Green Space and Attention: A Population-based Prospective Study. Environ Health Perspect. 2017;125(9):097016 Epub 2017/09/22. 10.1289/EHP694 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Browning M, Kuo F, Sachdeva S, Lee K, Westphal L. Greenness and school-wide test scores are not always positively associated–A replication of “linking student performance in Massachusetts elementary schools with the ‘greenness’ of school surroundings using remote sensing”. Landsc Urban Plan. 2018;178:69–72. 10.1016/j.landurbplan.2018.05.007 [DOI] [Google Scholar]
  • 59.Kweon B-S, Ellis CD, Lee J, Jacobs K. The link between school environments and student academic performance. Urban For Urban Green. 2017;23:35–43. 10.1016/j.ufug.2017.02.002 [DOI] [Google Scholar]
  • 60.Needleman HL, Leviton A, Bellinger D. Lead-associated intellectual deficit. N Engl J Med. 1982;306(6):367 Epub 1982/02/11. 10.1056/nejm198202113060619 [DOI] [PubMed] [Google Scholar]
  • 61.Liao J, Zhang B, Xia W, Cao Z, Zhang Y, Liang S, et al. Residential exposure to green space and early childhood neurodevelopment. Environ Int. 2019;128:70–6. Epub 2019/04/29. 10.1016/j.envint.2019.03.070 [DOI] [PubMed] [Google Scholar]
  • 62.Vanaken GJ, Danckaerts M. Impact of Green Space Exposure on Children's and Adolescents' Mental Health: A Systematic Review. Int J Environ Res Public Health. 2018;15(12):2668 Epub 2018/11/30. 10.3390/ijerph15122668 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Zach A, Meyer N, Hendrowarsito L, Kolb S, Bolte G, Nennstiel-Ratzel U, et al. Association of sociodemographic and environmental factors with the mental health status among preschool children-Results from a cross-sectional study in Bavaria, Germany. Int J Hyg Environ Health. 2016;219(4–5):458–67. Epub 2016/05/18. 10.1016/j.ijheh.2016.04.012 [DOI] [PubMed] [Google Scholar]
  • 64.Balseviciene B, Sinkariova L, Grazuleviciene R, Andrusaityte S, Uzdanaviciute I, Dedele A, et al. Impact of residential greenness on preschool children's emotional and behavioral problems. Int J Environ Res Public Health. 2014;11(7):6757–70. Epub 2014/07/01. 10.3390/ijerph110706757 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Richardson EA, Pearce J, Shortt NK, Mitchell R. The role of public and private natural space in children's social, emotional and behavioural development in Scotland: A longitudinal study. Environ Res. 2017;158:729–36. Epub 2017/07/28. 10.1016/j.envres.2017.07.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Feng X, Astell-Burt T. Residential Green Space Quantity and Quality and Child Well-being: A Longitudinal Study. Am J Prev Med. 2017;53(5):616–24. Epub 2017/09/03. 10.1016/j.amepre.2017.06.035 [DOI] [PubMed] [Google Scholar]
  • 67.Mitchell R, Popham F. Greenspace, urbanity and health: relationships in England. J Epidemiol Community Health. 2007;61(8):681–3. Epub 2007/07/17. 10.1136/jech.2006.053553 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Markevych I, Schoierer J, Hartig T, Chudnovsky A, Hystad P, Dzhambov AM, et al. Exploring pathways linking greenspace to health: Theoretical and methodological guidance. Environ Res. 2017;158:301–17. Epub 2017/07/04. 10.1016/j.envres.2017.06.028 [DOI] [PubMed] [Google Scholar]
  • 69.Voracek M, Haubner T. Twin-singleton differences in intelligence: a meta-analysis. Psychol Rep. 2008;102(3):951–62. Epub 2008/09/04. 10.2466/pr0.102.3.951-962 [DOI] [PubMed] [Google Scholar]
  • 70.Eriksen W, Sundet JM, Tambs K. Twin-singleton differences in intelligence: a register-based birth cohort study of Norwegian males. Twin Res Hum Genet. 2012;15(5):649–55. Epub 2012/08/11. 10.1017/thg.2012.40 [DOI] [PubMed] [Google Scholar]
  • 71.Posthuma D, de Geus EJ, Bleichrodt N, Boomsma DI. Twin-singleton differences in intelligence? Twin Res. 2000;3(2):83–7. 10.1375/136905200320565535 [DOI] [PubMed] [Google Scholar]
  • 72.van Os J, Wichers M, Danckaerts M, Van Gestel S, Derom C, Vlietinck R. A prospective twin study of birth weight discordance and child problem behavior. Biological psychiatry. 2001;50(8):593–9. Epub 2001/11/03. 10.1016/s0006-3223(01)01085-x [DOI] [PubMed] [Google Scholar]
  • 73.Robbers SC, Bartels M, van Oort FV, van Beijsterveldt CE, van der Ende J, Verhulst FC, et al. A twin-singleton comparison of developmental trajectories of externalizing and internalizing problems in 6- to 12-year-old children. Twin Res Hum Genet. 2010;13(1):79–87. Epub 2010/02/18. 10.1375/twin.13.1.79 [DOI] [PubMed] [Google Scholar]
  • 74.Nawrot TS, Perez L, Kunzli N, Munters E, Nemery B. Public health importance of triggers of myocardial infarction: a comparative risk assessment. Lancet. 2011;377(9767):732–40. 10.1016/S0140-6736(10)62296-9 [DOI] [PubMed] [Google Scholar]
  • 75.Dijkema MB, Mallant SF, Gehring U, van den Hurk K, Alssema M, van Strien RT, et al. Long-term exposure to traffic-related air pollution and type 2 diabetes prevalence in a cross-sectional screening-study in the Netherlands. Environ Health. 2011;10:76 Epub 2011/09/06. 10.1186/1476-069X-10-76 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Baccarelli A, Martinelli I, Pegoraro V, Melly S, Grillo P, Zanobetti A, et al. Living near major traffic roads and risk of deep vein thrombosis. Circulation. 2009;119(24):3118–24. Epub 2009/06/10. 10.1161/CIRCULATIONAHA.108.836163 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Saenen ND, Bove H, Steuwe C, Roeffaers MBJ, Provost EB, Lefebvre W, et al. Children's Urinary Environmental Carbon Load. A Novel Marker Reflecting Residential Ambient Air Pollution Exposure? Am J Respir Crit Care Med. 2017;196(7):873–81. Epub 2017/07/08. 10.1164/rccm.201704-0797OC [DOI] [PubMed] [Google Scholar]
  • 78.Population Division of the United Nations Department of Economic and Social Affairs (UN DESA). New York: World Urbanization Prospects; 2018. [Google Scholar]
  • 79.de Keijzer C, Gascon M, Nieuwenhuijsen MJ, Dadvand P. Long-Term Green Space Exposure and Cognition Across the Life Course: a Systematic Review. Curr Envrion Health Rep. 2016;3(4):468–77. Epub 2016/10/13. 10.1007/s40572-016-0116-x [DOI] [PubMed] [Google Scholar]

Decision Letter 0

Clare Stone

13 Mar 2020

Dear Dr. Bijnens,

Thank you very much for submitting your manuscript "The importance of residential greenness on intellectual and behavioural development in children living in urban versus rural environments" (PMEDICINE-D-19-03156) for consideration at PLOS Medicine. Please accept my sincere apologies for the unusual delay in getting back to you about it.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Apr 03 2020 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see http://journals.plos.org/plosmedicine/s/submission-guidelines#loc-methods.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Clare Stone, PhD

Managing Editor

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

Please revise your title according to PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible. "Effect of" should be used only if causality can be inferred, i.e., for an RCT. Please place the study design ("A randomized controlled trial," "A retrospective study," "A modelling study," etc.) in the subtitle (ie, after a colon).In addition, as the clarity around the use of ‘greenness’ has been raised by referees, please also consider this and please add the country setting.

Abstract – Please start the ‘Methods and Findings’ section with more general background about the study rather than the intelligence scale used (focus should be more on how selection took place, numbers, dates, locations and some brief demographic information); please add a sentence on the limitations of the study as the final sentence or two of the Methods and Findings section.

Data - PLOS Medicine requires that the de-identified data underlying the specific results in a published article be made available, without restrictions on access, in a public repository or as Supporting Information at the time of article publication, provided it is legal and ethical to do so. Note Corresponding authors cannot be points of contact for data requests. Please see the policy at

http://journals.plos.org/plosmedicine/s/data-availability

and FAQs at

http://journals.plos.org/plosmedicine/s/data-availability#loc-faqs-for-data-policy

Line 109 – please ensure all questionnaires are provided (and translated to English, as needed) as Supp Files and provide a callout in the methods.

Line 113 – what is the ‘Child Behaviour Checklist’

Line 228 – remove ‘till now’ and instead replace with ‘to our knowledge this is the first…’

Be cautious around overstating findings (eg in conclusion…’ is especially’ ….may be would be better as this isn’t a trial)

Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section.

a) If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study, and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b) If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c) In either case, changes in the analysis-- including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

Please ensure that the study is reported according to the STROBE guideline, and include the completed STROBE checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)."

The STROBE guideline can be found here: http://www.equator-network.org/reporting-guidelines/strobe/

When completing the checklist, please use section and paragraph numbers, rather than page numbers.

Comments from the reviewers:

Reviewer #1: See attachment

Michael Dewey

Reviewer #2: This study is among the first studies on IQ and greenspace, and is among the few existing studies on behavioural problems and greenspace. After revision, it may be considered for publication. My specific comments and recommendations are outlined below.

L67-69: The Authors never explain what is greenness, and instead, go to green spaces right away. These are two different concepts, which shuld be at least briefly mentioned. Greenness referes to general degree of vegetation while by green spaces, structured vegetation in parks, forests, meadows, etc. is meant. Greenness is typically satellite-derived and green spaces are extracted from land use/land cover datasets (if these data are not self-reported, of course).

L73-80: Talking about greenspace and academic performance, Authors ignore negative findings, as summarized by https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6388261/. Moreover, the analysis of Wu et al. had methodological shortcomings, judging by the re-analysis of the data (https://www.sciencedirect.com/science/article/pii/S0169204618303086).

L69-71: Pathways explaining effects of green spaces/greenness are outlined here https://pubmed.ncbi.nlm.nih.gov/28672128-exploring-pathways-linking-greenspace-to-health-theoretical-and-methodological-guidance/ and here https://pubmed.ncbi.nlm.nih.gov/24387090-nature-and-health/.

L106: Please correct the reference. It is Supplement Figure 1, not Figure 1.

L104: Just to make it clear: children born from 1964 were included into the analysis? It does not seem so from the further text.

L121: A very brief explanation that monochorionic twins share the same placenta would be nice for readers without medical background.

L133-134: Was complete residential history unknown? At least, time lived at the current address? How about moving during pregnancy? All this may have caused exposure misclassification.

L138-141:

(1) How did this correspond to the years of birth and IQ assessment? Please provide the information on when the children were born and tested. Land cover did not remain absolutely the same over decades.

(2) Furthermore, why not smaller buffers were considered? At least 500-m buffer would be informative. If distribution is too skewed (seems so), I would recommend to dichotomize the variabe into presence vs absence of green spaces in the given buffer. It is very hard for me to believe that green space in such big buffers would be more important for children than green space close to their home/school.

(3) Also, CORINE's minimum mapping unit is 25 ha which means that smaller parks are not present in those dataset. Some Authors even say that CORINE is not exactly applicable if urban areas (e.g., https://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-12-337), which I agree with. Have Authors tried using Landsat-derived vegetation indexes (NDVI, SAVI, EVI, etc.) as a sesnsitivity analysis? I would be curious to see whther results are similar.

(4) Finally, since Authors dealt with CORINE-derived green spaces, I would rename the article and use „green spaces" instead of „greenness" throughout the manuscript.

L144-159: What are minimum and maximum scores of the both WISC-R and CBCL tests, in total and od the subscales? This information would be of interest for a reader.

L171, L173: Please briefly mention why quadratic year was added to the models.

Have the Authors checked non-linearity of the studied associations somehow? This might be a reason for not seeing some associations. At least, I would recommend to categorize green space variables in tertiles in a sensitivity analysis.

L172: Why models were adjusted for distance to major road (and how this major road was defined?)? As a proxy for air pollution and/or noise? In this case, I would suggest to at least look at the effect estimates without this variables (if the Authors believe air pollution and noise are confounders) or even to do mediation analysis (if the Authors think this is mediation).

Table 1: How exactly parental education was defined - as the highest education of both parents baed on the available information? Please clarify in the text.

L203 and Figure 2: The strongest association was actually with 2000m buffer.

To all figures: Please consider using different symbols rather than colours, as many people have problems with differentiating colours. Figure 4 caused me eye pain, for instance.

L228-233: Actually, there is no associations overall, only in urban children. The three sentences should be re-formulated accordingly.

L237: This paper can also be cited: https://ehp.niehs.nih.gov/doi/full/10.1289/EHP1876?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed

L237-243: This is pretty much a repetition of L71-77.

L261: This study should also be cited: https://www.ajpmonline.org/action/showPdf?pii=S0749-3797%2817%2930377-X

L266-267: Markevych et al. (ref. 17 in your list) also had urban and suburban-rural composition of the study area, speaking of urbanicity. There are many more studies on other outcomes that investigated effect modification by urbanicity, just from what I can easily remember: https://www.sciencedirect.com/science/article/pii/S1353829218311158 and https://www.mdpi.com/1660-4601/13/3/311/htm.

L270-272: In this context, I would especially encourage the Authors to consider some satellite-derived vegetation index, which would capture all vegetation, not only green spaces min 25 ha in area.

L273-300: This section is too long considering that the Authors did not attempt to investigate any of the mechanisms by mediation analyses. I would suggest condensing this part.

L305-307: This is not true. Landsat 4-5 TM satellite images are available since early 1980s, and they could have been used in this study in addition to CORINE 1990.

Among the limitations, which are clearly more than two: poor quality land cover data to derive green spaces data, as specified before, unavailability of full residential history and time-activity pattern, absence of mediation analysis. Even effect modification by SES was not checked.

L309: Additional adjustment for distance to major road does not exclude confounding by air pollution, as distance to major road is just a mere proxy. I wonder if real air pollution data were not available. For example, there are ozone and NO2 maps from the ELAPSE project for the year 2000, and the estimates can be back-extrapolated. In any case, I would at least soften the language in this part.

More effort should be put into explaining why green space was not important for suburban and rural residents compared to urban residents. Also, why no associations were observed for internalizing problems? Why not for all buffers associations were present? In my view, Discussion needs to be shorten in some parts and elaborated in others.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Attachment

Submitted filename: bijnens.pdf

Attachment

Submitted filename: ReviewPLOS Medecine.docx

Decision Letter 1

Clare Stone

3 Jun 2020

Dear Dr. Bijnens,

Thank you very much for re-submitting your manuscript "The importance of residential green space on intellectual and behavioural development in children living in urban versus rural environments in Belgium; a longitudinal birth cohort study" (PMEDICINE-D-19-03156R1) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by the previous reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

Our publications team (plosmedicine@plos.org) will be in touch shortly about the production requirements for your paper, and the link and deadline for resubmission. DO NOT RESUBMIT BEFORE YOU'VE RECEIVED THE PRODUCTION REQUIREMENTS.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jun 10 2020 11:59PM.

Sincerely,

Clare Stone, PhD

Managing Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

We would ask that the title is revised to: "Residential green space and child intelligence and behaviour across urban, suburban and rural areas in Belgium: a longitudinal birth cohort study of twins"

Please list the adjustment factors in the abstract

Around line 50, please quote at least one further limitation, e.g. unmeasured confounding (it seems likely that confounding would be an issue in this study)

At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

Please make changes to statistical values: p<0.0001->p<0.001 in the results, and elsewhere

Please ensure the STROBE checklist is in sections and paragraphs instead of page numbers as these can change during formatting etc.

Please confirm that the email address for data (twins@uzgent.be.) does not direct to one of the authors on this manuscript, as this is not permitted. If so, please find another point of contact or URL as for the other data points in the study.

Please remove ‘Statements’ (lines 450-474) from the Word doc as these will be pulled in automatically from information in EM.

Comments from Reviewers:

Reviewer #1: The authors have addressed all my points thoroughly.

Michael Dewey

Reviewer #3:

after reviewing all the answers of the comments, I consider authors made all the modifications required and therefore, I would agree continuing with the proccess.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Decision Letter 2

Clare Stone

8 Jul 2020

Dear Dr. Bijnens,

On behalf of my colleagues and the academic editor, Dr. Iana Markevych, I am delighted to inform you that your manuscript entitled "Residential green space and child intelligence and behaviour across urban, suburban and rural areas in Belgium; a longitudinal birth cohort study of twins" (PMEDICINE-D-19-03156R2) has been accepted for publication in PLOS Medicine.

PRODUCTION PROCESS

Before publication you will see the copyedited word document (in around 1-2 weeks from now) and a PDF galley proof shortly after that. The copyeditor will be in touch shortly before sending you the copyedited Word document. We will make some revisions at the copyediting stage to conform to our general style, and for clarification. When you receive this version you should check and revise it very carefully, including figures, tables, references, and supporting information, because corrections at the next stage (proofs) will be strictly limited to (1) errors in author names or affiliations, (2) errors of scientific fact that would cause misunderstandings to readers, and (3) printer's (introduced) errors.

If you are likely to be away when either this document or the proof is sent, please ensure we have contact information of a second person, as we will need you to respond quickly at each point.

PRESS

A selection of our articles each week are press released by the journal. You will be contacted nearer the time if we are press releasing your article in order to approve the content and check the contact information for journalists is correct. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact.

PROFILE INFORMATION

Now that your manuscript has been accepted, please log into EM and update your profile. Go to https://www.editorialmanager.com/pmedicine, log in, and click on the "Update My Information" link at the top of the page. Please update your user information to ensure an efficient production and billing process.

Thank you again for submitting the manuscript to PLOS Medicine. We look forward to publishing it.

Best wishes,

Clare Stone, PhD

Managing Editor

PLOS Medicine

plosmedicine.org

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 STROBE Checklist. The checklist of STROBE guidelines.

    STROBE, strengthening the reporting of observational studies in epidemiology.

    (DOCX)

    S1 Text. Questionnaire.

    (PDF)

    S2 Text. A priori statistical analysis plan.

    (PDF)

    S1 Table. Full model with all covariates of the association between green space in 3,000-m radius around the residence and intelligence.

    CI, confidence interval; IQR, interquartile range; TIQ, total intelligence quotient; TPIQ, performance intelligence quotient; TVIQ, verbal intelligence quotient. *low versus high.

    (DOCX)

    S1 Fig. Statistical sectors in East Flanders classified as urban, suburban and rural areas depending on their population density, employment, location, and spatial planning.

    Source: Flemish Government–Department Environment.

    (TIF)

    S2 Fig. Flowchart.

    From the children with IQ data available, 599 mothers were invited to complete a questionnaire on residential history. A total of 324 mothers completed the questionnaire. From the twin pairs, we excluded 14 pairs because of missing data on residential location or parental education resulting in a final IQ study sample of 620 children. For a subset of 442 children, aged 6–15 years, data are available for studying behavioral characteristics. IQ, intelligence quotient.

    (TIF)

    S3 Fig. Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence during pregnancy and in childhood.

    Dark blue: Residential green space in childhood at time of IQ measurement (n = 236) adjusted for sex, age, parental education, neighborhood household income, year of IQ test (linear and quadratic), and zygosity and chorionicity. Red: Residential green space during pregnancy (n = 216) adjusted for sex, gestational age, birth weight, birth year (linear and quadratic), zygosity and chorionicity, cord insertion, parental education, neighborhood household income, and smoking during pregnancy. IQ, intelligence quotient; IQR, interquartile range.

    (TIF)

    S4 Fig. Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence in children living in an urban (n = 236) area and in a subgroup of children living minimum 10 years (or their whole life if the age of the child <10 years old) at their current address (n = 144).

    Adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. IQ, intelligence quotient; IQR, interquartile range.

    (TIF)

    S5 Fig. Main model additionally adjusted for distance to major road and NO2 exposure.

    Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence in children living in an urban (n = 236) area. The main model is adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. We additionally adjust for distance to major road and regional background levels of NO2 exposure the year before IQ test. IQ, intelligence quotient. IQ, intelligence quotient; IQR, interquartile range; NO2, nitrogen dioxide.

    (TIF)

    S6 Fig. Effect modification socioeconomic status.

    Estimated change in IQ for an IQR increase in green space in a 1,000–5,000-m radius around the residence in children living in an urban area. Based on parental education, we stratified the children residing in an urban area into groups low parental education (n = 82) and high parental education (n = 154). Adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. IQ, intelligence quotient; IQR, interquartile range.

    (TIF)

    S7 Fig. Estimated change in IQ for an IQR increase in green space (high-resolution model 2012) in a 50–2,000-m radius around the residence.

    Green space is high green (vegetation height higher than 3m, including blue space). In twins living in (A) an urban (n = 236), (B) a suburban (n = 128), and (C) a rural area (n = 256). Adjusted for sex, age, parental education, neighborhood household income, year of IQ test, and zygosity and chorionicity. IQ, intelligence quotient; IQR, interquartile range.

    (TIF)

    Attachment

    Submitted filename: bijnens.pdf

    Attachment

    Submitted filename: ReviewPLOS Medecine.docx

    Attachment

    Submitted filename: 200501_response_letter_PlosMed2.docx

    Data Availability Statement

    Green space data (CORINE land cover) used in this study can be obtained from the European Environment Agency (EEA) (https://land.copernicus.eu/pan-european/corine-land-cover). Air pollution data used in this study can be obtained from IRCEL (https://www.irceline.be/en). Access to and use of summarized health data of study participants requires submission of a research proposal to the East Flanders Prospective Twin Survey, Ghent University, contact email, twins@uzgent.be.


    Articles from PLoS Medicine are provided here courtesy of PLOS

    RESOURCES