Supplementary figure 1.

Associations between the distance to the nearest urban green space

and behavioural problems assessed by the Strengths and Difficulties Questionnaire (SDQ),
stratified by whether the residence is in the inner city of Munich or the surrounding suburban
areas. Results are presented per 500 m increase in distance to the nearest urban green space.
All models are adjusted for study, sex, age, parental level of education, age of mother at time
of birth, single parent status at the 10-year follow-up, time spent in front of a screen and time
spent outdoors. Inverse triangles indicate the inner city of Munich; squares indicate the
surrounding areas; circles indicate results for the total population. Black color indicates
proportional odds ratios; white indicates odds ratios (a - abnormal / borderline vs. normal, b -
abnormal vs. borderline / normal). Asterixes indicate statistical significance at the 5 % level.

1
Supplementary table 1
Associations between the distance to the nearest urban green space and behavioural problems
assessed by the Strengths and Difficulties Questionnaire (SDQ), stratified by sex. Results are
presented as proportional odds ratios (pOR) or odds ratios (OR) with corresponding 95 %
confidence intervals (CI).
Adjusteda pOR or ORb (95 % CI)c
Males (n = 994) Females (n = 938)
Total difficulties score 1.21 (1.02 – 1.44) 0.79 (0.58 – 1.07)

Emotional symptoms 1.06 (0.87 – 1.27)

Abnormal / borderline vs. normal 0.84 (0.66 – 1.07)
Abnormal vs. borderline / normal 0.88 (0.66 – 1.19)

Conduct problems 1.01 (0.84 – 1.23) 1.05 (0.82 - 1.35)

Hyperactivity / inattention 1.28 (1.08 – 1.50) 0.75 (0.52 – 1.09)

Peer relationship problems 1.20 (0.98 -1.47) 1.17 (0.89 – 1.54)

a
Adjusted for study, age, parental level of education, age of mother at time of birth, single
parent status at the 10-year follow-up, time spent in front of a screen and time spent outdoors.
Bold text indicates statistical significance at the 5 % level.
b
Proportional odds ratios are homogenous across the cutoffs of the ordinal SDQ variables. If
the crude or adjusted models violated this assumption, cutpoint-specific odds ratios are
reported.
c
Effects reported per 500 m increase in the distance to the nearest urban green space.