Morbidity is related to a green living environment

Transcription

1 1 EMGO Institute VU University Medial Centre, Amsterdam, The Netherlands; 2 NIVEL (Netherlands Institute for Health Servies Researh), Utreht, The Netherlands; 3 Alterra, Green World Researh, Wageningen, The Netherlands; 4 Utreht University, Department of Human Geography, Department of Soiology, Utreht, The Netherlands Correspondene to: Dr J Maas, EMGO Institute, VU University Medial Centre, Van der Boehorststraat 7, 1081 BT Amsterdam, The Netherlands; jolanda.maas@rivm.nl Aepted 10 July 2009 Published Online First 15 Otober 2009 Morbidity is related to a green living environment J Maas, 1 R A Verheij, 2 S de Vries, 3 P Spreeuwenberg, 2 F G Shellevis, 1,2 P P Groenewegen 2,4 ABSTRACT Bakground: As a result of inreasing urbanisation, people fae the prospet of living in environments with few green spaes. There is inreasing evidene for a positive relation between green spae in people s living environment and self-reported indiators of physial and mental health. This study investigates whether physiianassessed morbidity is also related to green spae in people s living environment. Methods: Morbidity data were derived from eletroni medial reords of 195 general pratitioners in 96 Duth praties, serving a population of people. Morbidity was lassified by the general pratitioners aording to the International Classifiation of Primary Care. The perentage of green spae within a 1 km and 3 km radius around the postal ode oordinates was derived from an existing database and was alulated for eah household. Multilevel logisti regression analyses were performed, ontrolling for demographi and soioeonomi harateristis. Results: The annual prevalene rate of 15 of the 24 disease lusters was lower in living environments with more green spae in a 1 km radius. The relation was strongest for anxiety disorder and depression. The relation was stronger for hildren and people with a lower soioeonomi status. Furthermore, the relation was strongest in slightly urban areas and not apparent in very strongly urban areas. Conlusion: This study indiates that the previously established relation between green spae and a number of self-reported general indiators of physial and mental health an also be found for lusters of speifi physiianassessed morbidity. The study stresses the importane of green spae lose to home for hildren and lower soioeonomi groups. INTRODUCTION As a result of inreasing urbanisation, ombined with a planning poliy of spatial densifiation, more people fae the prospet of living in residential environments with little green spae. At the same time, inreasing evidene shows that green spae has benefiial effets on people s health. Evidene has been found for a positive relation between green spae and self-pereived health, 1 4 longevity, 5 number of symptoms and the risk of psyhiatri morbidity. 1 Aess to a garden and shorter distanes to green areas from the dwelling were assoiated with less stress and a lower likelihood of obesity. 6 Experimental studies showed that there is a positive relation between green spae and restoration from stress and mental fatigue. More speifi, exposure to nature has been found to have a positive effet on mood, onentration, self-disipline and physiologial stress Researh report These studies indiate that there is a relation between green spae and self-reported general indiators of physial and mental health. Thus, people living in greener environments report better physial and mental health. The derease in green spae ould therefore have health onsequenes. However, it remains unknown whether living in residential environments with little green spae also has negative onsequenes for objetive health. In this explorative study we will go one step further than other studies and investigate whether the prevalene of several physiian-assessed morbidity lusters is also related to the amount of green spae in people s living environment. This is the first study to investigate the relation between green spae and prevalene of physiian-assessed morbidity. This study has an explorative harater and takes into aount a broad number of diseases highly prevalent in soiety. To gain more insight into the relation between green spae and physiian-assessed morbidity we analysed this relation separately for different age groups and different soioeonomi groups. We hypothesise that the relation is stronger for elderly people and hildren than adults beause, as a result of their lower mobility, they spend more time in the viinity of their home, resulting in higher exposure to green spae in their living environment. The same applies to people with a lower soioeonomi status (SES), whose ativities and soial ontats are situated lose to their homes. Therefore we also hypothesise that people with a lower SES are more exposed to the green spae in their living environment. Finally, the relation was analysed for different levels of urbaniity to investigate whether the relation varies between urban and rural areas. METHODS For this study data from two different datasets were ombined. Morbidity data were olleted within the framework of the seond Duth National Survey in General Pratie (DNSGP-2), whih inluded a nationwide, representative sample of 104 general praties with 195 GPs and a pratie population of approximately enlisted people, who were representative for the Duth population in terms of age, gender and type of health insurane. 13 For this study data from 96 praties that reorded morbidity for a full period of 12 months or more were used. This seletion had no signifiant effet on the representativeness of the data, beause after the seletion the sample was still representative for the Duth population. 13 Only people who had been registered with their urrent GP for longer than 12 months prior to the study (n = ) were inluded, beause we J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. J Epidemiol Community Health 2009;63: doi: /jeh

2 Researh report assumed that people will have to live for at least 12 months in the same living environment before any effet of it would be notieable. Environmental data were derived from the National Land Cover Classifiation database (LGN4) in 2001, whih ontains the dominant type of land use of eah m grid ell in The Netherlands. 14 The two datasets were mathed on the basis of the x and y oordinates of the respondent s six-harater postal ode (on average about households have the same sixharater postal ode). The dataset inluded postal odes and on average 6.9 respondents resided in eah postal ode area. Morbidity data Morbidity data were derived from routine primary are eletroni medial reords. In The Netherlands morbidity presented in general pratie is a good indiator of morbidity in the population. 15 Basially all non-institutionalised people are registered with a GP. Furthermore, GPs have a gate-keeping role for seondary are and are usually the first point of ontat with the healthare system. The data have been validated for obtaining prevalene estimates. 16 During a 12-month period, data on all GP onsultations with patients were extrated from the eletroni medial reords. These data inluded ontat diagnoses and indiations (diagnoses) for mediation and referral to seondary are. Prevalene rates are based on ontats that were lassified by the GP aording to the International Classifiation of Primary Care and subsequently lustered into episodes of disease. 17 The most prevalent episodes were ombined into 24 disease lusters. These disease lusters have been used in several other studies and over the full range of the most prevalent diseases in general pratie (prevalene.10 per 1000) (see table 1). The 24 disease lusters have been distributed over seven disease ategories, namely ardiovasular diseases, musuloskeletal diseases, mental diseases, respiratory diseases, neurologial diseases, digestive diseases and misellaneous. The prinipal aim of this study was to explore possible assoiations between green spae and speifi diseases. Therefore, hosen disease lusters were not based on a presumed mehanism behind this assoiation beforehand. Instead, we hose a ommon basis ategorisation overing the full range of morbidity most frequently presented in general pratie, aording to the bodily system involved. Not all disease lusters were relevant for all age groups; therefore, the epidemiologial denominator varied (table 1). A prevalene rate for eah luster was alulated by dividing the number of patients with at least one disease episode in 2001 belonging to the luster by the population at risk. The population at risk was based on age groups in whih the diseases ourred. Some disease lusters, like for instane high blood pressure, were only present in the older age groups. Therefore, for high blood pressure the prevalene rate was alulated by dividing the number of patients with at least one high blood pressure episode in 2001 by the older age groups and not for the population as a whole. Charateristis of the respondents living environment The LGN4 database disriminates 39 land use lasses inluding rop types, forest types, water, various urban lasses and seminatural lasses and has been proven to be valid and aurate The total perentage of green spae in the respondents living environment was measured within a 1 km radius and within a 3 km radius around the postode entroid of a respondent s home, to see whether there is a stronger relation for green spae lose by than green spae further away. Only green spaes that dominate the land use in the m grid ell (more than 50% of the grid ell is green) have been lassified as green spae in the dataset. Small-sale green spaes, suh as street trees and roadside vegetation were only inluded as green spae if they were dominant in the grid ell. Urbaniity Another environmental harateristi is urbaniity. This variable onsists of five ategories, ranging from very strongly urban (1) to non-urban (5); it was measured at the muniipal level and was derived from Statistis Netherlands. The indiator is based on the number of households per square kilometre and is ommonly used in The Netherlands. 21 Demographi and soioeonomi harateristis Part of the relation between green spae and health may be the result of diret or indiret seletion. Diret seletion would take plae when people s health is related to their hanes of living in a green environment. Indiret seletion takes plae when people with ertain harateristis related to well-being (suh as inome) tend to live in a green environment. 22 As migration flows are related to suh soiodemographi harateristis as age, inome and eduation, 23 we deided to rule out indiret seletion effets as far as possible by ontrolling statistially for demographi and soioeonomi harateristis. The demographi harateristis taken into aount were gender (female = 1) and age (whih was taken into aount as a polynomial until the third order beause there was no linear relation between the disease lusters and age), and were derived from the patient lists of the partiipating praties. To find out whether the relation between green spae and morbidity differed between age groups, age was divided into six ategories: hildren (,12 years), adolesents (13 17 years), youths (18 25 years), young adults (26 45 years), older adults (46 65 years) and elderly people (65+ years). Soioeonomi harateristis were olleted by a registration form that was sent by mail to all people listed in the partiipating praties in the DNSGP-2 (n = , response 76.5%) 13 and inluded eduation, work status and healthare insurane type. For a number of people these soioeonomi harateristis were unknown. To redue the number of missing entries we inluded a ategory unknown in the analyses. Eduation was measured as the highest level of ompleted eduation (unknown, no eduation ompleted, primary eduation, seondary eduation, higher eduation). Work situation was ategorised as work situation unknown, paid job, attending shool/studying, housewife/houseman, retired, disability pension, unemployed. Soioeonomi status was additionally impliitly measured by type of healthare insurane (unknown, publi or private). The type of healthare insurane an be regarded as an indiator of SES in the Duth ontext in 2001, as only people with a higher inome had private health insurane, whereas people with a lower inome had obligatory publi health insurane. When testing the relation between green spae and the annual prevalene of disease lusters for different SES groups, SES was operationalised as the level of eduation divided into three ategories: higher eduation (university or higher voational eduation), seondary eduation and primary or no eduation. Charateristis of the study population are displayed in table 2. J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. 968 J Epidemiol Community Health 2009;63: doi: /jeh

3 Table 1 Annual prevalene rates of lusters of diseases presented in general pratie (ases per 1000) (n = unless stated otherwise) Cluster ICPC odes N (abs) Per 1000 Cardiovasular High blood pressure (n = ) K85 K86 K Cardia disease K71 K73 K74 K77 K78 K79 K80 K81 K82 K K84 Coronary heart disease (n = ) K74 K75 K Stroke, brain haemorrhage (n = ) K89 K Musuloskeletal Nek and bak omplaints L01 L02 L03 L84 L Severe bak omplaints L02 L03 L85 L Severe nek and shoulder omplaints L01 L08 L83 L Severe elbow, wrist and hand omplaints L10 L11 L12 L72 L Osteoarthritis (n = ) L89 L90 L Arthritis (n = ) L88 T Mental Depression P03 P Anxiety disorder P01 P Respiratory Upper respiratory trat infetion A77 R72 R74 R75 R76 R Bronhi(oli)tis/pneumonia R78 R Asthma, COPD R91 R95 R Neurologial Migraine/severe headahe N01 N02 N03 N89 N90 N Vertigo N Digestive Severe intestinal omplaints D81 D85 D86 D92 D93 D Infetious disease of the intestinal anal D70 D Misellaneous MUPS A01 A04 D01 D08 D09 D12 D18 D21 D K01 K02 K04 L01 L02 L03 L08 L09 L14 L20 N01 N02 N17 P06 P20 R02 R21 T03 T07 T08 Chroni ezema S86 S87 S Aute urinary trat infetion U70 U71 U Diabetes (n = ) T88 T Caner A79 B72 B73 B74 D74 D75 D76 D77 F74 H75 K72 L71 N74 R84 R85 S77 S80 T71 T73 U75 U76 U77 U79 W72 X75 X76 X77 X81 Y77 Y COPD, hroni obstrutive pulmonary disease; ICPC, International Classifiation of Primary Care; MUPS, medially unexplained physial symptoms. Statistial analysis The relation between perentage of green spae in people s living environment and morbidity was assessed using multilevel logisti regression analyses, ontrolling for demographi and soioeonomi harateristis and urbaniity. We inluded two levels, individuals and praties, beause of the hierarhial struture of the data within DNSGP-2. The multilevel logisti regression analyses were performed with MLwiN. The independent variables, inluding the perentage of green spae, were entred around their average. The results thus represent morbidity of the average population living in an area with an average amount of green spae. We used interation effets between respetive age groups, SES groups and urbaniity and the green spae indiator to investigate the relation for different age groups, SES groups and in different levels of urbaniity. Beause of the large dataset we adopted a strit type 1 error riterion of a = RESULTS On average there is 42.4% of green spae in a 1 km radius and 60.8% of green spae in a 3 km radius around people s homes. Table 3 presents the ORs for the annual prevalene rate of the Researh report 24 disease lusters for people who have 10% more green spae than average. In general, a signifiant relation between the perentage of green spae and the annual prevalene rate was only present for green spae in a 1 km radius. Only for anxiety disorders, infetious diseases of the digestive system and medially unexplained physial symptoms (MUPS) the annual prevalene rate was lower in environments with more green spae in a 3 km radius. For 15 of the 24 disease lusters the annual prevalene rate was lower in living environments with a higher perentage of green spae in a 1 km radius. This relation is apparent for diseases in all seven disease ategories. It is strongest for anxiety disorders and depression. The relationship is negative for none of the disease lusters. Strength of the relation An indiation of the strength of the relation is given in table 4, whih shows the annual prevalene per 1000 for people with average harateristis on the ontrol variables with respetively 10% and 90% green spae in a 1 km radius around their home. For anxiety disorders, the annual prevalene for people with average harateristis with 10% green spae in a 1 km radius J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. J Epidemiol Community Health 2009;63: doi: /jeh

4 Researh report Table 2 Charateristis of the study population Charateristis of the study population (%) (n = ) Demographi harateristis Gender Male 49.5 Age (12 years years years years years year 12.8 Soioeonomi harateristis Highest level of eduation Unknown 25.2 No eduation ompleted 11.7 Primary eduation 14.2 Seondary eduation 36.8 Higher eduation 12.1 Health insurane Unknown 23.9 Publi 50.9 Private 25.3 Work situation Work situation unknown 27.9 Paid job 31.5 Attending shool/studying 16.4 Housewife/houseman 11.1 Retired 9 Disability pension 3 Unemployed 1.1 Urbaniity Very strongly urban 13.9 Strongly urban 22.2 Moderately urban 22.6 Slightly urban 31.7 Non urban 9.7 was 26 per 1000 people and for those with 90% green spae in a 1 km radius 18 per 1000 people. For depression these figures are respetively 32 and 24 per In general, the found relation between green spae and physiian-assessed morbidity is omparable with the relation between age and morbidity. An inrease in 1 perentage point of green spae on physiianassessed morbidity equals the effet of 1-year lower age. Relation in different age groups Further analysis showed that the relation was strongest for hildren younger than 12 and people between 46 and 65 (not in table). For hildren the relation was not only apparent for the perentage of green spae in a 1 km radius, but also for the perentage of green spae in a 3 km radius. For a few disease lusters the relation for hildren was espeially strong, for example for vertigo (1 km OR 0.81, 95% CI 0.74 to 0.90; 3 km OR 0.85, 95% CI 0.77 to 0.94) and severe intestinal omplaints (1 km OR 0.85, 95% CI 0.80 to 0.90; 3 km OR 0.89, 95% CI 0.84 to 0.94). The strongest relation for hildren was found for depression (1 km OR 0.79, 95% CI 0.72 to 0.88; 3 km OR 0.84, 95% CI 0.78 to 0.91). The relations for the other age groups were similar to the overall relations shown in table 3. Table 3 the relation between having 10% more green spae than average in one s living environment and the prevalene of disease lusters (n = unless stated otherwise) Cluster Perentage of green spae in 1 km radius OR (95% CI) Perentage of green spae in 3 km radius OR (95% CI) Cardiovasular High blood pressure 0.99 (0.98 to 1.00) 1.00 (0.98 to 1.02) (n = ) Cardia disease 0.98 (0.97 to 0.99) 1.00 (0.96 to 1.04) Coronary heart disease 0.97 (0.95 to 0.99) 0.97 (0.93 to 1.01) (n = ) Stroke, brain haemorrhage 0.98 (0.95 to 1.00) 0.98 (0.92 to 1.04) Musuloskeletal Nek and bak omplaints 0.98 (0.97 to 0.99) 0.99 (0.97 to 1.00) Severe bak omplaints 0.98 (0.97 to 0.99) 1.00 (0.98 to 1.01) Severe nek and shoulder 0.98 (0.97 to 0.99) 1.00 (0.98 to 1.01) omplaints Severe elbow, wrist and hand 0.97 (0.96 to 0.98) 1.01 (0.99 to 1.03) omplaints Osteoarthritis (n = ) 0.97 (0.93 to 1.01) 0.97 (0.92 to 1.03) Arthritis (n = ) 0.99 (0.97 to 1.01) 1.00 (0.96 to 1.04) Mental Depression 0.96 (0.95 to 0.98) 0.98 (0.96 to 1.00) Anxiety disorder 0.95 (0.94 to 0.97) 0.96 (0.93 to 0.99) Respiratory Upper respiratory trat infetion 0.97 (0.96 to 0.98) 0.99 (0.97 to 1.01) Bronhi(oli)tis/pneumonia 0.99 (0.97 to 1.00) 1.02 (0.99 to 1.04) Asthma, COPD 0.97 (0.96 to 0.98) 1.01 (0.99 to 1.03) Neurologial Migraine/severe headahe 0.98 (0.97 to 0.99) 0.98 (0.96 to 1.00) Vertigo 0.97 (0.95 to 0.99) 0.98 (0.94 to 1.02) Digestive Severe intestinal omplaints 0.98 (0.96 to 1.00) 0.99 (0.95 to 1.03) Infetious disease of the 0.97 (0.95 to 0.99) 0.95 (0.91 to 0.99) intestinal anal Misellaneous MUPS 0.97 (0.96 to 0.98) 0.98 (0.97 to 0.99) Chroni ezema 0.99 (0.97 to 1.00) 0.99 (0.95 to 1.03) Aute urinary trat infetion 0.97 (0.96 to 0.98) 0.98 (0.95 to 1.01) Diabetes mellitus (n = ) 0.98 (0.97 to 0.99) 0.98 (0.97 to 1.00) Caner 1.00 (0.98 to 1.02) 0.99 (0.95 to 1.03) ORs are derived from multilevel logisti regression analysis, ontrolling for demographi and soioeonomi harateristi and urbaniity. COPD, hroni obstrutive pulmonary disease; MUPS, medially unexplained physial symptoms. Numbers in bold signify p,0.01. Relation for different soioeonomi groups Espeially the lower eduated groups had a lower annual prevalene rate when they had more green spae in a 1 km radius around their home. For example, the odds for hroni obstrutive pulmonary disease (COPD) were smaller for the lower eduated (1 km OR 0.97; 95% CI to 0.95 to 0.99) than for higher eduated (OR 0.98; 95% CI 0.96 to 1.00). Relation for different levels of urbaniity Conerning the level of urbaniity our analyses show that urbaniity influenes the relation between green spae and the annual prevalene of disease lusters (not in table). There is often no relation between green spae and the annual prevalene of disease lusters in the very strongly urban areas. At all other levels of urbaniity people with more green spae in a 1 km radius around their home had a lower annual prevalene rate. The relations between green spae and annual prevalene rates were strongest in slightly urban areas. J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. 970 J Epidemiol Community Health 2009;63: doi: /jeh

5 Table 4 Prevalene rates per 1000 in living environments with 10% and 90% green spae for different disease lusters Prevalene per 1000 Cluster 10% green spae 90% green spae Cardiovasular High blood pressure Cardia disease Coronary heart disease Stroke, brain haemorrhage Musuloskeletal Nek and bak omplaints Severe bak omplaints Severe nek and shoulder omplaints Severe elbow, wrist and hand omplaints Osteoarthritis Arthritis Mental Depression Anxiety disorder Respiratory Upper respiratory trat infetion Bronhi(oli)tis/pneumonia Asthma, COPD Neurologial Migraine/severe headahe Vertigo Digestive Severe intestinal omplaints Infetious disease of the intestinal anal Misellaneous MUPS Chroni ezema Aute urinary trat infetion Diabetes Mellitus 10 8 Caner This table is based on results from multilevel logisti regression analysis ontrolling for demographi and soioeonomi harateristi and urbaniity that were entred around the average. COPD, hroni obstrutive pulmonary disease; MUPS, medially unexplained physial symptoms. DISCUSSION Prinipal findings This explorative study shows that the previously established relation between green spae and a number of self-reported general indiators of physial and mental health an also be found for speifi, dotor-assessed disease ategories. The annual prevalene rates for 15 of the 24 investigated disease lusters is lower in living environments with more green spae in a 1 km radius. Green spae lose to home appeared to be more important than green spae further away. This is in ontrast with our previous studies 1 2 whih found the relation between self-pereived health and the amount of green spae in a 1 km and a 3 km radius around people s homes to be equally strong. It appears that for the prevalene of these more speifi diseases green spae lose to home is more important. This study differs from other studies, whih mainly foused on the relation between green spae and self-pereived measures of physial and mental health This is the first study to assess the relation between green spae and speifi diseases, derived from eletroni medial reords of GPs. This dataset helps better establish the relation between green spae and health, beause it Researh report used physiian-assessed morbidity as outome, beause there was no single soure bias in the data, and beause we used a large dataset that was representative for The Netherlands. In line with our hypothesis, the relation was strongest for people who were expeted to spend more time in the viinity of their homes, namely hildren and people with lower SES. However, ontrary to our expetations the relation appeared to be stronger for people aged between 46 and 65 than for elderly people. Conerning urbaniity, the relation appeared to be strongest in slightly urban areas. In very strongly urban areas there was no relation with the annual prevalene of disease lusters. This may be related to the fat that green spaes in highly urban areas are more often found to evoke feelings of inseurity, 24 thereby inhibiting their use. This study only gives some indiations for the relation between green spae and morbidity for different subgroups. Further researh should fous speifially on one of the subgroups to investigate the relation for subgroups more thoroughly. This study shows that the role of green spae in the living environment for health should not be underestimated. Most of the diseases whih were found to be related to the perentage of green spae in the living environment are highly prevalent in soiety and in many ountries they are subjet of large-sale prevention programmes. Furthermore, in many ountries, diseases of the irulatory system and mental disorders are among the most expensive diseases with respet to healthare osts. 25 Our study ontributes to the evidene that green spae an help fight some major publi health threats in Western soieties. Healthy spatial planning should take the amount of green spae in the living environment into aount when endeavouring to improve the health situation, partiularly of hildren and lower soioeonomi groups. Underlying mehanisms The results of this study give some indiations for the possible mehanism behind the relation between green spae and health. Several mehanisms ould be responsible for the relation between green spae and health, of whih the following are most ommonly mentioned: reovery from stress and attention fatigue, enouragement of physial ativity, failitation of soial ontat and better air quality What do the results tell us about the mehanism at work? The strong relation we found, partiularly for anxiety disorder and depression, suggests that mental health in partiular might be affeted by the amount of loal green spae. Reovery from stress and attention fatigue then seems a likely mehanism behind the relation between green spae and health; also, failitation of soial ontats might ontribute. However, there is no reason to disard any of the other possible mehanisms. For example, physial ativity is also known to have mental health benefits. 27 Furthermore, in living environments with more green spae the prevalene of most respiratory illnesses was lower, indiating that air quality ould also be a possible mehanism behind the relation between green spae and health. For diseases related to physial ativity (diabetes, oronary heart disease, musuloskeletal diseases) somewhat less strong relations were found. But as assoiations were present, physial ativity ould also be a possible mehanism. Strength and limitations This is the first large epidemiologial study investigating the relation between the amount of green spae in the living environment of people and the prevalene of physiian-assessed morbidity. J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. J Epidemiol Community Health 2009;63: doi: /jeh

6 Researh report Morbidity data were derived from a different database than the data on green spae; onsequently, there is no single soure or method bias. On the other hand, we don t have information on exposure time. The morbidity data are aurate beause they were extrated from routine eletroni medial reords of general praties, and the interobserver reliability of grouping ontats into episodes was high. 15 The registration overed a 12-month period for eah pratie in order to eliminate seasonal influenes. Considering the representativeness of the partiipating GPs and their patients and the high validity of the data the results of the present study an be assumed to validly represent morbidity in Duth general pratie. Furthermore, beause general pratie in The Netherlands is usually the first point of ontat with the healthare system, and beause the GP has a gate-keeping role for speialist are, and beause there are no large geographi 28 or soial differenes in aess to general pratie, morbidity presented in general pratie an be regarded as a very lose approximation of morbidity present in the general population. The data used for this study also have some shortomings. First, our data on green spae, although assessed on a small sale, does not take small green spaes in the living environment into aount. A m grid ell was only regarded as green spae when green spae dominates in the grid ell. Gardens and small-sale green spaes, suh as street trees and green verges, whih ould also influene people s health, are not regarded as green spae in our study. Consequently, the relation might be slightly underestimated. Seond, beause of the ross-setional design of the study, it is not possible to make strong inferenes about the ausality of the relations that were found. The observed relations between green spae and health may partly be aused by seletion. We tried to rule out this possibility by taking soioeonomi and demographi harateristis into aount, but the effets of seletion annot be ruled out ompletely. The results from the subgroup analyses by SES groups, however, make it rather unlikely that seletion is the mehanism responsible. The relationship observed between green spae and morbidity was stronger for the less well-eduated group and this is exatly the subgroup that has fewer options in their hoie of neighbourhood of residene. Our results may be influened by seletive migration based on people s health (diret seletion). However, longitudinal studies on health-related migration show that diret seletion annot be held responsible for geographial differenes that remain if soioeonomi and demographi fators are taken into aount. Third, we tried to ontrol as muh as possible for individual SES. However, we did not have any information on the inome of the respondents, whih is a relevant indiator for SES. Furthermore, we did not ontrol for other onfounders at the neighbourhood level, although different studies have shown that, for example, neighbourhood SES ould also influene health Beause this was an explorative study we hose to keep the design somewhat simple. Further researh should try to find out whether a relation an also be found when neighbourhood SES is ontrolled for. The aim of this study was to explore possible assoiations between speifi diseases and green spae. The disease lusters used in this study were therefore based on bodily systems, and overing the full range of morbidity most frequently presented in general pratie. However, this does not mean that we onsider all relationships equally plausible (in a ausal sense). For some disease lusters it seems more diffiult to understand What is already known on this subjet There is inreasing evidene for a positive relation between green spae in the living environment and a number of selfreported indiators of physial and mental health. Small-sale psyhologial researh showed that exposure to green spae has a positive effet on stress redution and attention restoration. Several epidemiologial studies have shown that green spae is positively orrelated with self-pereived health, number of symptoms experiened and mortality. What this study adds This study uses large-sale representative medial reord data to investigate whether the prevalene of a number of disease lusters is related to the amount of green spae in people s living environment. The annual prevalene rates for 15 of the 24 investigated disease lusters is lower in living environments with more green spae in a 1 km radius around people s homes. The study stresses the importane of green spae lose to people s homes. The relationship is partiularly strong for hildren and lower soioeonomi groups. why their prevalene would be positively related to the loal amount of green spae, for example infetious diseases of the intestinal anal. Furthermore, given the many signifiant relationships the absene of others relationships is also worthwhile noting. For example, high blood pressure ould be hypothesised to be linked to hronially high stress levels as well as lak of physial ativity, but was not related to the amount of green spae. Further researh will have to shed more light on the mehanisms behind the relation between green spae and health, and to what extent green spae indeed plays a ausal role in the observed relationships. Funding: This study was supported by a grant from The Netherlands Organisation for Sientifi Researh. Competing interests: None. Provenane and peer review: Not ommissioned; externally peer reviewed. REFERENCES 1. de Vries S, Verheij RA, Groenewegen PP, et al. Natural environments healthy environments? An exploratory analysis of the relationship between greenspae and health. Environ Plan A 2003;35: Maas J, Verheij RA, Groenewegen PP, et al. Green spae, urbanity and health. how strong is the relation? J Epidemiol Community Health 2006;60: Mithell R, Popham F. Greenspae, urbanity and health: relationships in England. J Epidemiol Community Health 2007;61: Sugiyama T, Leslie E, Giles-Corti B, et al. Assoiations of neighbourhood greenness with physial and mental health: do walking, soial oherene and loal soial interation explain the relationships? J Epidemiol Community Health 2008;62:e9; doi: /jeh Takano T, Nakamura K, Watanabe M. Urban residential environments and senior itizens longevity in megaity areas. The importane of walkable green spaes. J Epidemiol Community Health 2002;56: Nielsen TS, Hansen KB. Do green areas affet health? Results from a Danish survey on the use of green areas and health indiators. Health Plae 2007;13: Health Counil of The Netherlands, Duth Advisory Counil for researh on Spatial Planning Nature and the Environment. Nature and health. The influene of nature on soial, psyhologial and physial well-being. The Hague: Health Counil of The Netherlands; RMNO, J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. 972 J Epidemiol Community Health 2009;63: doi: /jeh

7 8. van den Berg AE, Hartig T, Staats H. Preferene for nature in urbanized soieties: stress, restoration, and the pursuit of sustainability. J So Issues 2007;63: Hartig T, Mang M, Evans GW. Restorative effets of natural environment experienes. Environ Behav 1991;23: Kaplan R, Kaplan S. The experiene of nature. A psyhologial perspetive. Cambridge: Cambridge University Press, Shwanen T, Dijst M, Dieleman FM. A mirolevel analysis of residential ontext and travel time. Environ Plan A 2002;34: Harms L. Op weg in de vrije tijd: ontext, kenmerken en dynamiek van vrijetijdsmobiliteit. Den Haag: Soiaal Cultureel Planbureau, Westert GP, Shellevis FG, de Bakker DH, et al. Monitoring health inequalities through general pratie: the Seond Duth National Survey of General Pratie. Eur J Publi Health 2005;15: Thunissen HAM, De Wit AJW. The national land over database of The Netherlands. ISPRS Journal of Photogrammetry & Remote Sensing 2000; Van der Linden MW, Westert GP, De Bakker DH, et al. Klahten en aandoeningen in de bevolking en in de huisartspraktijk. Utreht, Bilthoven: NIVEL, Rijksinstituut voor Volksgezondheid en Milieu, Shland M, Berenner MH, Hoopman M, et al. Approahes to the denominator in pratie-based epidemiology: a ritial overview. J Epidemiol Community Health 1998;52(Suppl 1): Lamberts H, Wood M. ICPC. International lassifiation of primary are. Oxford: Oxford University Press, van Lindert H, Droomers M, Westert GP. Een kwestie van vershil: vershillen in zelf-gerapporteerde leefstijl, gezondheid en zorggebruik. Utreht/Bilthoven: NIVEL, RIVM, Nielen M, Verheij RA, De Bakker DH, et al. Vooronderzoek verbetering kwaliteit huisartsenzorg in ahterstandsgebieden grote steden. Utreht: NIVEL, Darwinism 20. De Wit AJW, Clevers JGPW. Effiieny and auray of per-field lassifiation for operational rop mapping. Int J Remote Sens 2004;25: Den Dulk CJ, Van de Stadt H, Vliegen JM. Een nieuwe maatstaf voor stedelijkheid: de omgevingsadressendihtheid [A new measure for degree of urbanisation: the address density of the surrounding area]. Mndstat Bevolk 1992;7: Verheij RA. Explaining urban-rural variations in health: a review of interations between individual and environment. So Si Med 1996;42: Heins S. Rurale woonmilieus in stad en land. Plattelandsbeelden, vraag naar en aanbod van rurale woonmilieus. Delft: Eburon, Jorgensen A, Hithmough J, Calvert T. Woodland spaes and edges: their impat on pereption of safety and preferene. Lands Urban Plan 2002;60: Groenewegen PP, van den Berg AE, de Vries S, et al. effets of green spae on health, well-being, and soial safety. BMC Publi Health 2006;6: Maas J, Van Dillen SMJ, Verheij RA, et al. Soial ontats as a possible mehanism behind the relation between green spae and health. Health Plae 2008;15: US Department of Health and Human Servies. Physial ativity and health: a report of the Surgeon General. Atlanta: US Department of HHS, Heijink R, Koopmanshap MA, Polder JJ. International omparison of ost of illness. Bilthoven: RIVM, Verheij RA, van de Mheen HD, Groenewegen PP, et al. Urban-rural variations in health. Does seletive migration play a part? J Epidemiol Community Health 1998;52: Van Lenthe FJ, Martikainen P, Makenbah JP. Neighbourhood inequalities in health and health-related behaviour: results of seletive migration? Health Plae 2007;13: Kawahi I, Berkman LF. Neighborhood and health. New York: Oxford University Press, Ross CE, Mirowksi J. Neighborhood soioeonomi status and health: ontext or omposition? City and Community 2008;7: Barker DJP. Glossary: Developmental origins of adult health and disease. J Epidemiol Community Health 2004;58: This term is often used to desribe the natural seletion of genes that optimises the fitness of a speies in a partiular environment. Darwin was also aware, however, that the environment to whih individuals are exposed during development produes variation within one generation. When a variation is of the slightest use to a being, hewrote, we annot tell how muh of it to attribute to the aumulative ation of natural seletion, and how muh to the onditions of life. J Epidemiol Community Health 2009;63:973. doi: /jeh Darwin anniversary Researh report J Epidemiol Community Health: first published as /jeh on 15 Otober Downloaded from on 21 January 2019 by guest. Proteted by opyright. J Epidemiol Community Health Deember 2009 Vol 63 No