An Olympic Legacy? Did the Urban Regeneration Associated With the London 2012 Olympic Games Influence Adolescent Mental Health?

An Olympic Legacy? Did the Urban Regeneration Associated With the London 2012 Olympic Games... Abstract Public expenditure on large events such as the London 2012 Olympic Games is often justified by the potential legacy of urban regeneration and its associated health and well-being benefits for local communities. In the Olympic Regeneration in East London Study, we examined whether there was an association between urban regeneration related to the 2012 Games and improved mental health in young people. Adolescents aged 11–12 years attending schools in the Olympic host borough of Newham in London or in 3 adjacent comparison London boroughs completed a survey before the 2012 Games and 6 and 18 months after the Games (in 2013 and 2014, respectively). Changes in depressive symptoms and well-being between baseline and each follow-up were examined. A total of 2,254 adolescents from 25 randomly selected schools participated. Adolescents from Newham were more likely to have remained depressed between baseline and the 6- and 18-month follow-up surveys (for 6-month follow-up, relative risk = 1.78, 95% confidence interval: 1.12, 2.83; for 18-month follow-up, relative risk = 1.93, 95% confidence interval: 1.01, 3.70) than adolescents from the comparison boroughs. No differences in well-being were observed. There was little evidence that urban regeneration had any positive influence on adolescent mental health and some suggestion that regeneration may have been associated with maintenance of depressive symptoms. Such programs may have limited short-term impact on the mental health of adolescents. adolescent, depressive symptoms, longitudinal, positive well-being, urban regeneration Public expenditure on large sporting events such as the London 2012 Olympic Games is often justified by the hypothesized positive impact of urban regeneration, including the provision of new sports and recreational facilities, related to hosting such events and the associated legacy benefits for local communities (1). Addressing deprivation by enhancing health and well-being through urban regeneration programs was among the legacy benefits identified for the London 2012 Olympic Games (2). However, public-health evaluations have thus far found little evidence of a positive association between large-scale investment in urban regeneration at the household, dwelling, community, or neighborhood levels with either mental health or social determinants of mental health (3, 4). In fact, urban regeneration may negatively impact mental health (3–6) via increased stress associated with the environmental nuisance of regeneration (6), gentrification (5, 7), changes to social networks (8), displacement (1), loss of amenities (1), lack of control over the planning process (1, 8), and an increase in relative deprivation among residents who do not benefit. Nevertheless, regeneration may impact positively on children and adolescents from these communities (7), via increasing feelings of safety, reducing exposure to stressors, and increasing access to amenities. However, there have been few studies in which the associations of this sort of urban regeneration with the mental health and well-being of children and adolescents have been evaluated (4). The Olympic Regeneration in East London (ORiEL) Study examined whether urban regeneration occurring around the 2012 London Olympic Games (hereafter, the Games) was associated with improved adolescent mental health and well-being among adolescents who lived in those areas compared with those who lived in areas with less or no Olympic-related urban regeneration. We hypothesized that adolescents living in an area receiving most of the Games-related urban regeneration would have greater positive change and better mental health 6 and 18 months after the regeneration than adolescents living in comparison areas receiving less or no Games-related urban regeneration. METHOD Study design and participants This longitudinal, quasi-experimental study followed for 3 years a cohort of adolescents recruited from randomly selected schools in the London boroughs of Barking and Dagenham, Hackney, Newham, and Tower Hamlets. The participants, ages 11–12 years in year 7 at baseline (January–June 2012), were first followed up in year 8 (ages 12–13 years; January–June 2013) 6 to 11 months after the completion of the Games, and again in year 9 (ages 13–14 years; January–June 2014) 18–23 months after the Games. Ethical approval was granted by the Queen Mary University of London Research Ethics Committee (QMREC2011/40), the Association of Directors of Children’s Services (RGE110927), and the London Boroughs Research Governance Framework (CERGF113). Head teachers gave written consent for the study in their school; parents gave passive informed consent and could opt their child out of the study; adolescents gave written informed assent for the study. Exposure to urban regeneration associated with the Games The intervention borough was the London borough of Newham, where the 2012 Games primarily were hosted and where the majority of regeneration occurred. Within Newham, regeneration associated with the Games was focused in 3 main areas: Stratford City development, the Olympic Park, and the Olympic Fringe. The main components of this regeneration program are outlined in Table 1. Although part of the retail complex (Westfield Stratford City) had opened in 2011–2012, before baseline data collection, large areas where regeneration was occurring were inaccessible to the local communities from 2008 to late 2012. For example, the Olympic Park development involved closing off the area from 2008 and opening for a limited time only for ticket holders for the Games at the end of July 2012; the surrounding communities did not gain regular access to the Olympic Park and its facilities until early 2013. Before 2008, much of the Olympic site was inaccessible, derelict industrial land. Table 1. Main Regeneration Components Associated With the London 2012 Olympic Games in the London Borough of Newham, United Kingdom, Olympic Regeneration in East London Study, 2011–2014 Datea  Area  Main Components  2011–2012  Stratford City Development  Retail and leisure center comprising 1.9 million ft2 (176,516 m2) of retail space (including Westfield Stratford City), 500,000 ft2 (46,452 m2) of office and business space, and new civic and public space  2012–2014  Olympic Park  Regenerated land (246 hectares) comprising new green spaces and parkland, public space, and play areas; world-class sports venues (i.e., main stadium, aquatics center, velodrome, bicycle motocross and mountain bike tracks, road cycle route), and associated facilities; improved physical connectivity and accessibility to the Olympic Park from surrounding areas (i.e., foot and cycle paths, bridges, waterways, road and rail links); new housing associated with the former Athletes village (East Village)  2012–2014  Olympic Fringe  Fringe surrounding the Olympic Park will receive 90 hectares of improved green/civic space and improved connectivity to the main Olympic Park  Datea  Area  Main Components  2011–2012  Stratford City Development  Retail and leisure center comprising 1.9 million ft2 (176,516 m2) of retail space (including Westfield Stratford City), 500,000 ft2 (46,452 m2) of office and business space, and new civic and public space  2012–2014  Olympic Park  Regenerated land (246 hectares) comprising new green spaces and parkland, public space, and play areas; world-class sports venues (i.e., main stadium, aquatics center, velodrome, bicycle motocross and mountain bike tracks, road cycle route), and associated facilities; improved physical connectivity and accessibility to the Olympic Park from surrounding areas (i.e., foot and cycle paths, bridges, waterways, road and rail links); new housing associated with the former Athletes village (East Village)  2012–2014  Olympic Fringe  Fringe surrounding the Olympic Park will receive 90 hectares of improved green/civic space and improved connectivity to the main Olympic Park  a Baseline was January–June 2012; first follow-up was January–June 2013; and second follow-up was January–June 2014. Three adjacent London boroughs (Barking and Dagenham, Hackney, Tower Hamlets) were selected as comparison areas. We hypothesized they would benefit less from the planned regeneration, because they were farther from the Olympic Park and were not direct recipients of the regeneration activities outlined in Table 1. Figure 1 shows the distribution of the schools across the 4 boroughs in East London. Figure 1. View largeDownload slide Map showing the intervention and comparison boroughs of the Olympic Regeneration in East London Study, 2011–2014. Adapted from Smith et al. (10), with permission under license CC-BY-NC 2.0 (https://creativecommons.org/licenses/by-nc/2.0/). Figure 1. View largeDownload slide Map showing the intervention and comparison boroughs of the Olympic Regeneration in East London Study, 2011–2014. Adapted from Smith et al. (10), with permission under license CC-BY-NC 2.0 (https://creativecommons.org/licenses/by-nc/2.0/). Although selection of the intervention and comparison boroughs underpins the epidemiologic design of the study, we also conducted analyses that used a different characterization of “exposure” to the urban regeneration associated with the Games. This measure was the Euclidean distance of the participant’s school to the Olympic Park, calculated using a geographical information system (9). These analyses examined the odds for changes in mental health and well-being for an interquartile increase in distance to the Olympic Park (interquartile range, 3,240.7 m). Procedure Of the 48 secondary schools in the 4 boroughs, we randomly selected 6 to 7 schools in each borough; when a selected school refused to participate, another school within the borough was randomly selected to participate. The study had 80% power to detect an 8% difference in well-being (5% significance level) with an 18-month follow-up sample of 1,766 adolescents from 24 schools (10). This calculation was informed by a study (11) that found well-being scores on a range of scales improved by 8%–25% for adults and children after a neighborhood intervention. Of 41 invited schools, 25 participated (school response rate, 60.9%), resulting in 6 intervention and 19 comparison schools. In 7 schools, all students in year 7 were invited, which led to the recruitment of more than 90 children per school. In the remaining schools, a mixed-ability sample was selected. The sociodemographic characteristics of the baseline sample were similar to those reported in the 2011 United Kingdom national census. Adolescents followed a standardized protocol to complete a questionnaire in their classroom; the questionnaire assessed well-being, mental health, physical activity, and sociodemographic factors. Outcomes Well-being and depressive symptoms were self-reported by the adolescents. The World Health Organization defines well-being as “a state in which every individual realizes his or her own potential, can cope with the normal stresses of life, can work productively, and is able to make a contribution to her or his community” (12). Well-being was assessed by the Warwick-Edinburgh Mental Well-Being Scale (WEMWBS), a 14-item (5 response categories) self-report measure of subjective, positive well-being (13) at baseline and at the 6- and 18-month follow-up study points. Before and after absolute changes in the continuous WEMWBS scores between baseline and the 6- and 18-month follow-up surveys were calculated by subtracting the baseline score from the 6-month or 18-month follow-up score, respectively. Depressive symptoms were assessed by the Short Moods and Feelings Questionnaire, a 13-item (3 response categories) self-report measure (14) valid for use with adolescents. The Short Moods and Feelings Questionnaire has been shown to discriminate a psychiatric sample from a pediatric control sample (14). Dichotomous scores (14) indicative of clinically relevant depressive symptoms were used to determine whether an adolescent had clinically relevant depressive symptoms at each time point (no depressive symptoms: score, 0–7; depressive symptoms: score, ≥8). These dichotomous variables were then used to assess change between baseline and the 6-month follow-up, and baseline and the 18-month follow-up. The resulting categorical outcomes were 1) no depressive symptoms at baseline or follow-up (reference group), 2) change from no depressive symptoms at baseline to depressive symptoms at follow-up (“became depressed”), 3) change from depressive symptoms at baseline to no depressive symptoms at follow-up (“no longer depressed”), and 4) depressive symptoms at baseline and follow-up (“remained depressed”). Covariates available at baseline and both follow-up surveys were identified a priori from existing literature, which demonstrated cross-sectional and longitudinal associations with depressive symptoms in our previous East London, school-based cohort study of adolescents (Research with East London Adolescents: Community Health Survey) conducted between 2001 and 2005 (15–18). The covariates were age (months); sex; ethnicity (assessed using an 11-category variable based on the 2011 United Kingdom Census); number of years lived in the United Kingdom; number of parents the participant lived with; parental income; bullied at school in the past 12 months; receiving free school meals (as a marker of social disadvantage); long-standing illness; number of life events; changed neighborhoods since last survey; and the number of days since the opening ceremony of the Games that the questionnaire was completed. Social support from family and friends was assessed using the Multidimensional Scale of Perceived Social Support (19), with scores divided into tertiles representing low, medium, and high support. Statistical analyses Weights were derived to take into account unequal probabilities of school and pupil selection. Prevalence and rates of missing data for the outcomes and covariates were examined; missing values ranged from 0.0% to 45.2%. We explored patterns and predictors of missing observations through logistic regression modelling. Analyses suggested that data were missing at random (MAR) (20). Data were imputed using multilevel multiple imputation in Realcom software (Realcom Applications, Coleorton, Leicestershire, UK) (21), which uses a joint, multivariate, normal modelling approach through the Markov Chain Monte Carlo method. We imputed with 2 levels (first, survey time (baseline, 6-month follow-up or 18-month follow-up); second, adolescent) with all the outcomes and covariates as fixed effects. Interaction terms between sex and the intervention, and free school meals and the intervention were also included. The imputation model was chosen to be congenial (22) with the most saturated model of interest; auxiliary variables were included to strengthen the MAR assumption. We used a “burn in” period of 25,000 iterations, followed by 50,000 iterations, producing a dataset every 1,000th iteration, which resulted in 50 imputed datasets. The Markov Chain Monte Carlo chains were examined to check for convergence. Analyses were carried out using STATA, version 14 (StataCorp LP, College Station, Texas) (23). Crude and adjusted linear, logistic, and multinomial regression models were run to assess the associations of urban regeneration (intervention vs comparison) with short-term (wave 1 to wave 2) and longer-term (wave 1 to wave 3) change in well-being and depressive symptoms. The models were adjusted for baseline demographic factors, household factors, family and friend social support, and psychological factors (Table 2). An additional adjustment for baseline well-being, as determined by the WEMWBS, was made to the adjusted model for well-being to test sensitivity of the findings for baseline well-being. Interactions between urban regeneration (i.e., borough or distance to the Olympic Park) with sex and free schools meals were tested; models were stratified and interactions were statistically significant when P ≤ 0.05. Inference sensitivity to departure from MAR was explored through tipping-point sensitivity analysis in which data were imputed under “missing not at random” with increasing departure from the MAR assumption. Inferences were robust to departures from the MAR assumption. Table 2. Descriptive Statistics for Key Measures at Each Survey, Olympic Regeneration in East London Study, London, United Kingdom 2011–2014 Parameter  Baseline Survey  6-Month Follow-Up Survey  18-Month Follow-Up Survey  No. of Participants  %  Missing %  No.  %  Missing %  No.  %  Missing %  Exposure  Intervention  2,254    0.0               Comparison boroughs  1,626  72.1                 Newham borough  628  27.9                Outcome Measures  Short Moods and Feelings Questionnaire  2,068    8.2  2,116    6.1  2,155    4.4   Not depressed  1,618  78.2    1,676  79.2    1,633  75.8     Depressed (≥8)  450  21.8    440  20.8    522  24.2      Scale reliabilitya  0.87    0.90    0.91    Warwick-Edinburgh Mental Wellbeing Scale  1,943    13.8  2,016    10.6  2,085    7.5    Scale scoreb  51.2 (9.0)    51.6 (9.7)    51.2 (10.1)      Scale reliabilitya  0.84    0.88    0.90    Baseline Demographic Measures  Sex  2,254    0.0               Male  1,271  56.4                 Female  983  43.6                Ethnicity  2,254    0.0               White: United Kingdom  380  16.9                 White: other  326  14.4                 White: mixed  190  8.4                 Asian: Indian  85  3.8                 Asian: Pakistani  86  3.8                 Asian: Bangladeshi  337  14.9                 Asian: other  72  3.2                 Black: Caribbean  111  4.9                 Black: African  249  11.0                 Black: other  263  11.7                 Other  155  6.9                Time in the United Kingdom, years  2,221    1.5               All my life  1,629  73.3                 >10  174  7.8                 6–10  190  8.6                 <6  228  10.3                Household Measures  No. of parents working  1,984    12.0  2,104    6.7  2,090    7.3   2  796  40.1    813  38.6    832  39.8     1  832  41.9    883  42.0    891  42.6     0  356  17.9    408  19.4    367  17.6    Household composition  2,221    1.5  2,234    0.9  2,238    0.7   Lives with both parents  1,503  67.7    1,544  69.1    1,510  67.5     Lives with 1 parent  673  30.3    643  28.8    688  30.7     Lives with no parents  45  2.0    47  2.1    40  1.8    Moved neighborhood in past year (baseline) or since previous survey  1,914    15.1  2,137    5.2  2,167    3.9   No  1,746  91.2    1,977  92.5    2,022  93.3     Yes  168  8.8    160  7.5    145  6.7    Receives free school meals  2,208    2.0  2,214    1.8  2,219    1.5   No  1,376  62.3    1,404  63.4    1,503  67.7     Yes  832  37.7    810  36.6    716  32.3    Social Support Measures  Tertile of multidimensional Scale of Perceived Social Support: family  1,236    45.2  1,663    26.2  1,962    13.0   Low  378  30.6    530  31.9    642  32.7     Medium  392  31.7    475  28.6    586  29.9     High  466  37.7    658  39.5    734  37.4    Multidimensional Scale of Perceived Social Support: friend; tertile  1,239    45.0  1,656    26.5  1,958    13.1   Low  349  28.2    495  29.9    649  33.1     Medium  470  37.9    592  35.7    634  32.4     High  420  33.9    569  34.4    675  34.5    Psychological Measures  Ever bullied  1,463    35.1               No  948  64.8                 Yes  515  35.2                No of negative life events  1,663    26.2  1,910    15.3  2,001    11.2   0  865  52.0    691  36.2    481  24.0     1  373  22.5    469  24.6    499  24.9     2  235  14.1    334  17.5    372  18.6     ≥3  190  11.4    416  21.8    649  32.4    Long-term illness  2,153    4.5               No  1,898  88.2                 Yes  255  11.8                Parameter  Baseline Survey  6-Month Follow-Up Survey  18-Month Follow-Up Survey  No. of Participants  %  Missing %  No.  %  Missing %  No.  %  Missing %  Exposure  Intervention  2,254    0.0               Comparison boroughs  1,626  72.1                 Newham borough  628  27.9                Outcome Measures  Short Moods and Feelings Questionnaire  2,068    8.2  2,116    6.1  2,155    4.4   Not depressed  1,618  78.2    1,676  79.2    1,633  75.8     Depressed (≥8)  450  21.8    440  20.8    522  24.2      Scale reliabilitya  0.87    0.90    0.91    Warwick-Edinburgh Mental Wellbeing Scale  1,943    13.8  2,016    10.6  2,085    7.5    Scale scoreb  51.2 (9.0)    51.6 (9.7)    51.2 (10.1)      Scale reliabilitya  0.84    0.88    0.90    Baseline Demographic Measures  Sex  2,254    0.0               Male  1,271  56.4                 Female  983  43.6                Ethnicity  2,254    0.0               White: United Kingdom  380  16.9                 White: other  326  14.4                 White: mixed  190  8.4                 Asian: Indian  85  3.8                 Asian: Pakistani  86  3.8                 Asian: Bangladeshi  337  14.9                 Asian: other  72  3.2                 Black: Caribbean  111  4.9                 Black: African  249  11.0                 Black: other  263  11.7                 Other  155  6.9                Time in the United Kingdom, years  2,221    1.5               All my life  1,629  73.3                 >10  174  7.8                 6–10  190  8.6                 <6  228  10.3                Household Measures  No. of parents working  1,984    12.0  2,104    6.7  2,090    7.3   2  796  40.1    813  38.6    832  39.8     1  832  41.9    883  42.0    891  42.6     0  356  17.9    408  19.4    367  17.6    Household composition  2,221    1.5  2,234    0.9  2,238    0.7   Lives with both parents  1,503  67.7    1,544  69.1    1,510  67.5     Lives with 1 parent  673  30.3    643  28.8    688  30.7     Lives with no parents  45  2.0    47  2.1    40  1.8    Moved neighborhood in past year (baseline) or since previous survey  1,914    15.1  2,137    5.2  2,167    3.9   No  1,746  91.2    1,977  92.5    2,022  93.3     Yes  168  8.8    160  7.5    145  6.7    Receives free school meals  2,208    2.0  2,214    1.8  2,219    1.5   No  1,376  62.3    1,404  63.4    1,503  67.7     Yes  832  37.7    810  36.6    716  32.3    Social Support Measures  Tertile of multidimensional Scale of Perceived Social Support: family  1,236    45.2  1,663    26.2  1,962    13.0   Low  378  30.6    530  31.9    642  32.7     Medium  392  31.7    475  28.6    586  29.9     High  466  37.7    658  39.5    734  37.4    Multidimensional Scale of Perceived Social Support: friend; tertile  1,239    45.0  1,656    26.5  1,958    13.1   Low  349  28.2    495  29.9    649  33.1     Medium  470  37.9    592  35.7    634  32.4     High  420  33.9    569  34.4    675  34.5    Psychological Measures  Ever bullied  1,463    35.1               No  948  64.8                 Yes  515  35.2                No of negative life events  1,663    26.2  1,910    15.3  2,001    11.2   0  865  52.0    691  36.2    481  24.0     1  373  22.5    469  24.6    499  24.9     2  235  14.1    334  17.5    372  18.6     ≥3  190  11.4    416  21.8    649  32.4    Long-term illness  2,153    4.5               No  1,898  88.2                 Yes  255  11.8                a Values are expressed as Cronbach α. b Values are expressed as mean (standard deviation). RESULTS Participants’ characteristics Table 2 lists the prevalence of the outcomes and covariates at each time point. Of the participants, 27.9% were from schools in the intervention borough and 72.1% from schools in the comparison boroughs. The distance from the participants’ schools to the Olympic Park ranged from 1,133 m to 12,589 m (interquartile range, 3,240.7 m). Participants were ethnically diverse; the largest groups described themselves as white from the United Kingdom (English/Welsh/Scottish/Northern Irish/British) (16.9%); Asian: Bangladeshi (14.9%); or white of any other background (excluding UK, Irish, and Gypsy or Irish Traveller) (14.4%). More than one-fifth of the sample reported depressive symptoms at each time point (baseline, 21.8%; 6-month follow-up, 20.8%; 18-month follow-up, 24.2%). The mean well-being score was 51 at each time point. Urban regeneration and depressive symptoms Depressive symptoms were higher in the adolescents in the intervention borough at baseline than in those in the comparison boroughs (27% vs 20%; unadjusted odds ratio (OR) = 1.49; P = 0.024). Table 3 lists the changes in depressive symptoms between baseline and the 6-month follow-up. In the fully adjusted models, at the 6-month follow-up, adolescents who were no longer depressed (relative risk = 1.53, 95% confidence interval (CI): 1.07, 2.20) or who remained depressed (relative risk = 1.78, 95% CI: 1.12, 2.83) were more likely to be from the intervention borough than from the comparison boroughs. These data reflect, at least in part, the higher prevalence of depression in the intervention borough at baseline. Table 3. Associations of Urban Regeneration With Change in Depressive Symptoms (Baseline to 6-Month Follow-up and Baseline to 18-Month Follow-up), Olympic Regeneration in East London Study, London United Kingdom, 2011–2014 Interval and Intervention  Became Depresseda  No Longer Depresseda  Remain Depresseda  Unadjusted  Adjustedb  Unadjusted  Adjustedb  Unadjusted  Adjustedb  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  Baseline to 6-month follow-up                           Intervention boroughc  1.32  0.81, 2.12  1.44  0.95, 2.17  1.44d  1.03, 2.02  1.53d  1.07, 2.20  1.70d  1.04, 2.77  1.78d  1.12, 2.83   Distance to Olympic Parke  1.08  0.95, 1.23  0.96  0.81, 1.15  1.10  0.95, 1.27  1.04  0.89, 1.20  1.02  0.89, 1.16  0.88  0.72, 1.07  Baseline to 18-month follow-up                           Intervention boroughc  1.23  0.82, 1.83  1.30  0.97, 1.76  1.53d  1.07, 2.17  1.39  0.88, 2.18  1.57  0.83, 2.95  1.93d  1.01, 3.70   Distance to Olympic Parke  1.16d  1.02, 1.31  1.05  0.89, 1.23  1.03  0.89, 1.20  1.03  0.85, 1.25  1.12  0.96, 1.29  0.99  0.76, 1.29  Interval and Intervention  Became Depresseda  No Longer Depresseda  Remain Depresseda  Unadjusted  Adjustedb  Unadjusted  Adjustedb  Unadjusted  Adjustedb  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  Baseline to 6-month follow-up                           Intervention boroughc  1.32  0.81, 2.12  1.44  0.95, 2.17  1.44d  1.03, 2.02  1.53d  1.07, 2.20  1.70d  1.04, 2.77  1.78d  1.12, 2.83   Distance to Olympic Parke  1.08  0.95, 1.23  0.96  0.81, 1.15  1.10  0.95, 1.27  1.04  0.89, 1.20  1.02  0.89, 1.16  0.88  0.72, 1.07  Baseline to 18-month follow-up                           Intervention boroughc  1.23  0.82, 1.83  1.30  0.97, 1.76  1.53d  1.07, 2.17  1.39  0.88, 2.18  1.57  0.83, 2.95  1.93d  1.01, 3.70   Distance to Olympic Parke  1.16d  1.02, 1.31  1.05  0.89, 1.23  1.03  0.89, 1.20  1.03  0.85, 1.25  1.12  0.96, 1.29  0.99  0.76, 1.29  Abbreviations: CI, confidence interval; RR, relative risk. a Reference group for the outcome is “remain not depressed.” b Adjusted for sex, age, ethnicity, length of time lived in the United Kingdom, number of days since the Olympic Games, parental income, number of parents participant lives with, moved neighborhood since baseline, eligible for free school meals, family social support, friend social support, bullying, negative life events, and long-term illness. c Intervention London borough is Newham; comparison London boroughs are Hackney, Tower Hamlets, and Barking and Dagenham. dP ≤ 0.05. e Distance to Olympic Park is estimated per interquartile increase in distance (meters). Post hoc analysis restricted to those with baseline depression showed that adolescents from the intervention area were not more likely to have recovered at the 6-month follow-up compared with those from the comparison area (OR = 0.72, 95% CI: 0.44, 1.18). Indeed, among those with no baseline depression, adolescents from the intervention area were more likely to have become depressed by the 6-month follow-up than those from the comparison area (OR = 1.42, 95% CI: 1.09, 1.85). Distance to the Olympic Park was not associated with changes in depressive symptoms between baseline and the 6-month follow-up. There were no interactions between borough or distance to the Olympic Park and sex or eligibility for free school meals relative to change in depressive symptoms between baseline and the 6-month follow-up (P > 0.05). Table 3 also lists the changes in depressive symptoms between baseline and the 18-month follow-up. In the fully adjusted models, adolescents who remained depressed at the 18-month follow-up were more likely to be from the intervention borough than from the comparison boroughs (relative risk = 1.93, 95% CI: 1.01, 3.70). Distance to the Olympic Park was not associated with changes in depressive symptoms between baseline and the 18-month follow-up (Table 3). No interactions were observed between borough or distance to the Olympic Park and sex or eligibility for free school meals relative to change in depressive symptoms between baseline and the 18-month follow-up (P > 0.05). Post hoc analyses showed that for those with no baseline depression, the adolescents from the intervention area were more likely to have become depressed at the 18-month follow-up than those from the comparison area (OR = 1.38, 95% CI: 1.06, 1.80). For adolescents with baseline depression, those from the intervention area were not more likely to have recovered at the 18-month follow-up (OR = 0.67, 95% CI: 0.41, 1.10). Urban regeneration and change in well-being Well-being scores were lower in the intervention borough at baseline than in the comparison boroughs (mean score (standard error), 50.7 (0.6) vs 53.0 (0.2); P = 0.001). There was little change in well-being scores between the surveys (mean (standard error) baseline to 6-month follow-up, μ = −0.02 (0.03); baseline to 18-month follow-up, μ = −0.04 (0.04)). According to the data in Table 4, there was no association between borough and change in well-being between baseline and either follow-up survey. Distance to the Olympic Park was not associated with changes in well-being between baseline and either follow-up survey (Table 4). No interactions were observed between borough or distance to the Olympic Park and sex or eligibility for free school meals in relation to change in well-being between baseline and either follow-up survey. Table 4. Associations of Urban Regeneration With Change in Well-Being (Baseline to 6-Month Follow-up and Baseline to 18-Month Follow-up), Olympic Regeneration in East London Study, London United Kingdom, 2011–2014 Parameter  Baseline to 6-Month Follow-Up  Baseline to 18-Month Follow-Up  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Co-efficient  95% CI  Co-efficient  95% CI  Intervention boroughb  0.58  −1.21, 2.36  0.18  −1.45, 1.80  −0.50  −1.86, 0.85  1.23  −0.74, 3.20  0.52  −0.95, 1.99  −0.07  −1.59, 1.44  Distance to Olympic Parkc  −0.18  −0.93, 0.57  0.01  −0.76, 0.78  0.31  −0.20, 0.81  −0.60  −1.46, 0.26  −0.01  −0.72, 0.70  0.11  −0.39, 0.62  Parameter  Baseline to 6-Month Follow-Up  Baseline to 18-Month Follow-Up  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Co-efficient  95% CI  Co-efficient  95% CI  Intervention boroughb  0.58  −1.21, 2.36  0.18  −1.45, 1.80  −0.50  −1.86, 0.85  1.23  −0.74, 3.20  0.52  −0.95, 1.99  −0.07  −1.59, 1.44  Distance to Olympic Parkc  −0.18  −0.93, 0.57  0.01  −0.76, 0.78  0.31  −0.20, 0.81  −0.60  −1.46, 0.26  −0.01  −0.72, 0.70  0.11  −0.39, 0.62  Abbreviations: CI, confidence interval; WEMWBS, Warwick-Edinburgh Mental Well-Being Scale. a Adjusted for sex, age, ethnicity, length of time lived in the United Kingdom, number of days since the Olympic Games, parental income, number of parents participants lives with, moved neighborhood since baseline, eligible for free school meals, family social support, friend social support, bullying, negative life events, and long-term illness. b Intervention London borough is Newham; the comparison London boroughs are Hackney, Tower Hamlets, and Barking and Dagenham. c Distance to Olympic Park is estimated per interquartile increase in distance (meters). DISCUSSION The findings of this study indicate that as a population-health intervention (7, 24), the urban regeneration associated with the London 2012 Olympic Games had little or no detectable positive influence on changes in adolescent mental health in terms of depressive symptoms or well-being. Although those who were depressed at baseline were more likely to have improvement in their symptoms at the 6-month follow-up in the intervention borough compared with adolescents in the control boroughs, this was the only positive association observed and should be considered in the light of more adolescents having depressive symptoms at baseline in the intervention borough. This association was not maintained 18-months after regeneration nor was it replicated in other analyses. In fact, attending school in the intervention borough was associated with a greater chance of maintaining depressive symptoms. Our hypotheses that adolescents in an area of urban regeneration would experience greater positive change and better mental health after the regeneration, compared with adolescents in areas of less or no urban regeneration, were not supported by the study data. These conclusions were further supported by analyses using a different characterization of exposure to the intervention: distance of the participant’s school to the Olympic Park. Urban regeneration and adolescent depressive symptoms The higher levels of depressive symptoms, in terms of remaining depressed between baseline and the 6- and the 18-month follow-up surveys, in adolescents living in the Olympic host borough receiving urban regeneration replicate the findings of some previous studies of urban regeneration in adult populations (3, 4). There was evidence that regeneration was associated with becoming depressed for those with no baseline depression. Urban regeneration can be associated with increased feelings of social isolation (25), reduced social capital (25), increased exposure to stress (26), and relative deprivation, all of which may influence mental health. Regeneration may not address residents’ concerns (1, 5) and may not influence psychosocial, lifestyle, safety, or economic determinants of mental health (6, 27). Interviews with families in the ORiEL Study revealed that during the Games, residents in the intervention borough felt their environment was safer and more unified, but they appreciated this was temporary, with little impact on more immediate concerns such as poor housing and opportunities for employment (28). However, adolescents in the intervention borough had higher rates of depressive symptoms at baseline and at both follow-up visits than did adolescents in the comparison boroughs, which may be explained by other unmeasured differences among the boroughs such as social, economic, and environmental determinants of mental health (6, 27), including income, employment, physical activity, and diet. Whether these determinants have associations with urban regeneration is being explored in a study of the parents of adolescents in the ORiEL Study. Borough-level differences in depressive symptoms may partially account for the findings reported here. A study conducted between 2001 and 2005 of adolescents in 3 of the London boroughs included in the ORiEL Study (i.e., Newham, Tower Hamlets, and Hackney) reported the highest rates of depressive symptoms in Hackney and the lowest rates in Newham (29). This finding suggests differences between boroughs in rates of depressive symptoms are changeable over a relatively short time. The association between living in the Olympic host borough and a greater likelihood of remission of depressive symptoms at 6 months but not 18 months after the Games may reflect temporary positive associations with regeneration (28). However, this finding may also reflect selection into this outcome group by baseline depressive symptoms. We could not replicate the association of borough on remission from depressive symptoms in analyses restricted to those with baseline depression, which may indicate chance findings. In contrast to the findings characterizing exposure to the urban regeneration by borough, we found that distance of the participant’s school from the Olympic Park was not associated with change in depressive symptoms between baseline and either of the 2 follow-up surveys. This difference in findings may reflect the more robust methodology in the ORiEL Study for examining regeneration by borough than by distance to the Olympic Park, because the sample was selected on borough. Urban regeneration and adolescent well-being There were no differences in change in well-being between the intervention and comparison boroughs or by distance to the Olympic Park, which suggests no positive association between Games-related regeneration and well-being. In our analyses, few factors predicted positive or negative change in well-being, which may be because we observed little change in well-being over time. Negative change was most strongly associated with baseline well-being. Authors of a recent systematic review of interventions using the WEMWBS found large variability in change scores for different interventions (30), but there were few studies in which associations with neighborhood interventions had been examined. Our study findings, along with those of 1 study of adults (31), suggest the WEMWBS may not be particularly sensitive to change for evaluating neighborhood population-health interventions. In our population, well-being appeared to be more of a stable trait than a changing state (32). Limitations and strengths Exposure to Games-associated regeneration was determined by school location, which may have resulted in some exposure misclassification. It is difficult to characterize the exact dose, or when regeneration begins and ends, for regeneration associated with large events like the Olympic Games; a small degree of regeneration may have started before our baseline survey and continued beyond our 18-month follow-up. There may be contamination between the intervention and comparison boroughs due to geographic proximity. Deprivation varied little between schools; however, area-level confounding by deprivation may remain. The regeneration was highly localized, resulting in much of Newham remaining unchanged. Furthermore, demographic changes in East London in the past decade, such as increases in population, ethnic diversity, and the private renting of flats could also have influenced the findings. Positive associations of urban regeneration with mental health may be demonstrated over a longer time or may have been observed immediately after the Games, but this was not tested in this study. This study reports results of analyses that characterize urban regeneration as a holistic event. Additional ongoing analyses will examine specific neighborhood environmental changes associated with the regeneration activities, such as access to greenness and walkability, and associations with health (33). Strengths of this study include the large representative sample indicative of the ethnic superdiversity of East London. To our knowledge, this is one of the first studies to examine the longitudinal associations of urban regeneration on the mental health of an adolescent population. The study had high response rates, thus overcoming a key limitation of many previous studies (4). Few participants relocated during the study, attrition was low, and multiple imputation was used to take into account missing data. Conclusion Complex public-health interventions like urban regeneration are very challenging to evaluate (7, 33). This large-scale, quasi-experimental study provides little evidence that the urban regeneration due the London 2012 Games was associated with adolescent mental health in terms of reduced depressive symptoms or increased well-being. In fact, urban regeneration may have maintained depressive symptoms in the studied group of adolescents, but this may be explained by differences between boroughs in the social and economic determinants of mental health. The predicted legacy benefit of the Games on enhancing well-being (2) was not observed here. Regeneration may have maintained and contributed to the onset of depressive symptoms. Any beneficial effects of regeneration may be elusive or may take longer to appear. ACKNOWLEDGMENTS Author affiliations: Centre for Psychiatry, Barts and the London School of Medicine, Queen Mary University of London, London, United Kingdom (Charlotte Clark, Melanie Smuk, Amanda Fahy, Neil Smith, Stephen A. Stansfeld); Department of Social and Environmental Health Research, London School of Hygiene and Tropical Medicine, London, United Kingdom (Steven Cummins, Daniel Lewis); Centre for Primary Care and Public Health, Barts and the London School of Medicine, Queen Mary University of London, London, United Kingdom (Sandra Eldridge, Stephanie J. C. Taylor); and School of Psychology, University of Surrey, Guildford, United Kingdom (Derek G. Moore). All authors contributed equally to this work. This work was funded by the National Institute for Health Research's Public Health Research Programme (grant 09/3005/09). We thank members of the Olympic Regeneration in East London Study research team, including Dr. Ellen Flint, Dr. Mark Petticrew, Dr. Adrian Renton, Vanathi Tharmaratnam, Dr. Claire Thompson, Danielle House, Shaneka Foster, Bukola Thompson, Rebecca Evans, Lianne Austin, Janat Hossain, and all other team members who helped with the study. The funders of the study had no role in the study design; data collection, analysis, or interpretation; or writing of the report. Conflict of interest: none declared. Abbreviations CI confidence interval MAR missing at random OR odds ratio ORiEL Olympic Regeneration in East London WEMWBS Warwick Edinburgh Mental Well-Being Scale REFERENCES 1 Clark J, Kearns A, Cleland C. Spatial scale, time and process in mega-events: the complexity of host community perspectives on neighbourhood change. Cities . 2016; 53: 87– 97. Google Scholar CrossRef Search ADS   2 Host Boroughs Joint Committee. Strategic Regeneration Framework: an Olympic Legacy for the Host Boroughs. London, UK: London Boroughs of Greenwich, Hackney, Newham, Tower Hamlets & Waltham Forest; 2009. http://www.gamesmonitor.org.uk/files/strategic-regeneration-framework-report.pdf. Accessed September 1, 2016. 3 Thomson H, Atkinson R, Petticrew M, et al.  . Do urban regeneration programmes improve public health and reduce health inequalities? A synthesis of the evidence from UK policy and practice (1980–2004). J Epidemiol Community Health . 2006; 60( 2): 108– 115. Google Scholar CrossRef Search ADS PubMed  4 Clark C, Myron R, Stansfeld S, et al.  . A systematic review of the evidence on the effect of the built and physical environment on mental health. J Public Ment Health . 2007; 6( 2): 14– 27. Google Scholar CrossRef Search ADS   5 Huxley P, Evans S, Leese M, et al.  . Urban regeneration and mental health. Soc Psychiatry Psychiatr Epidemiol . 2004; 39( 4): 280– 285. Google Scholar CrossRef Search ADS PubMed  6 Thomas R, Evans S, Huxley P, et al.  . Housing improvement and self-reported mental distress among council estate residents. Soc Sci Med . 2005; 60( 12): 2773– 2783. Google Scholar CrossRef Search ADS PubMed  7 Bond L, Egan M, Kearns A, et al.  . GoWell: the challenges of evaluating regeneration as a population health intervention. Prev Med . 2013; 57( 6): 941– 974. Google Scholar CrossRef Search ADS PubMed  8 Halpern D. Mental Health and the Built Environment . London, UK: Taylor & Francis; 1995. 9 Environmental Systems Research Institute. ArcGIS 10.3 . Redlands, CA: Environmental Systems Research Institute; 2014. 10 Smith NR, Clark C, Fahy AE, et al.  . The Olympic Regeneration in East London (ORiEL) study: protocol for a prospective controlled quasi-experiment to evaluate the impact of urban regeneration on young people and their families. BMJ Open . 2012; 2( 4): e001840. Google Scholar CrossRef Search ADS PubMed  11 Leventhal T, Brooks-Gunn J. Moving to opportunity: an experimental study of neighbourhood effects on mental health. Am J Public Health . 2003; 93( 9): 1576– 1582. Google Scholar CrossRef Search ADS PubMed  12 World Health Organization. Promoting Mental Health: Concepts, Emerging Evidence, and Practice. Summary Report. Geneva, Switzerland: World Health Organization, 2004. 13 Stewart-Brown S, Tennant A, Tennant R, et al.  . Internal construct validity of the Warwick-Edinburgh Mental Well-being Scale (WEMWBS): a Rasch analysis using data from the Scottish Health Education Population Survey. Health Qual Life Outcomes . 2009; 7: 15. Google Scholar CrossRef Search ADS PubMed  14 Angold A, Costello EJ, Messer SC, et al.  . Development of a short questionnaire for use in epidemiological studies of depression in children and adolescents. Int J Meth Psychiatr Res . 1995; 5: 237– 249. 15 Stansfeld SA, Haines MM, Head JA, et al.  . Ethnicity, social deprivation and psychological distress in adolescents: school-based epidemiological study in East London. Br J Psychiatry . 2004; 185( 3): 233– 238. Google Scholar CrossRef Search ADS PubMed  16 Klineberg E, Clark C, Bhui KS, et al.  . Social support, ethnicity and mental health in adolescents. Soc Psychiatry Psychiatr Epidemiol . 2006; 41( 9): 755– 760. Google Scholar CrossRef Search ADS PubMed  17 Fagg J, Curtis S, Clark C, et al.  . Neighbourhood perceptions among inner-city adolescents: relationships with their individual characteristics and with independently assessed neighbourhood conditions. J Environ Psychol . 2008; 28( 2): 128– 142. Google Scholar CrossRef Search ADS   18 Rothon C, Head J, Klineberg E, et al.  . Can social support protect bullied adolescents from adverse outcomes? A prospective study on the effects of bullying on the educational achievement and mental health of adolescents at secondary schools in East London. J Adolesc . 2011; 34( 3): 579– 588. Google Scholar CrossRef Search ADS PubMed  19 Zimet GD, Dahlem NW, Zimet SG, et al.  . The multidimensional scale of perceived social support. J Pers Assess . 1988; 52( 1): 30– 41. Google Scholar CrossRef Search ADS   20 Rubin DB. Multiple Imputation for Nonresponse in Surveys . New York, NY: John Wiley and Sons; 1987. Google Scholar CrossRef Search ADS   21 University of Bristol. REALCOM-Impute Software. Bristol, UK: University of Bristol; 2015. 22 Meng XL. Multiple-imputation inferences with uncongenial sources of input (with discussion). Stat Sci . 1994; 9( 4): 538– 573. Google Scholar CrossRef Search ADS   23 StataCorp. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP; 2015. 24 Craig P, Cooper C, Gunnell D, et al.  . Using natural experiments to evaluate population health interventions: new Medical Research Council guidance. J Epidemiol Community Health . 2012; 66( 12): 1182– 1186. Google Scholar CrossRef Search ADS PubMed  25 Rogers A, Huxley P, Evans S, et al.  . More than jobs and houses: mental health, quality of life and the perceptions of locality in an area undergoing urban regeneration. Soc Psychiatry Psychiatr Epidemiol . 2008; 43( 5): 364– 372. Google Scholar CrossRef Search ADS PubMed  26 Thomas R, Evans S, Gately C, et al.  . State-event relations among indicators of susceptibility to mental distress in Wythenshawe in the UK. Soc Sci Med . 2002; 55( 6): 921– 935. Google Scholar CrossRef Search ADS PubMed  27 Kearns A, Mason P. Regeneration, relocation and health behaviours in deprived communities. Health Place . 2015; 32: 43– 58. Google Scholar CrossRef Search ADS PubMed  28 Thompson C, Lewis DJ, Greenhalgh T, et al.  . “Everyone was looking at you smiling”: East London residents’ experiences of the 2012 Olympics and its legacy on the social determinants of health. Health Place . 2015; 36: 18– 24. Google Scholar CrossRef Search ADS PubMed  29 Institute of Community Health Sciences. Health of Young People in East London. The RELACHS Study . Norwich, UK: The Stationary Office; 2003. 30 Maheswaran H, Weich S, Powell J, et al.  . Evaluating the responsiveness of the Warwick Edinburgh Mental Well-Being Scale (WEMWBS): group and individual level analysis. Health Qual Life Outcomes . 2012; 10: 156. Google Scholar CrossRef Search ADS PubMed  31 Phillips G, Bottomley C, Schmidt E, et al.  . Well London Phase-1: results among adults of a cluster-randomised trial of a community engagement approach to improving health behaviours and mental well-being in deprived inner-city neighbourhoods. J Epidemiol Community Health . 2014; 68( 7): 606– 614. Google Scholar CrossRef Search ADS PubMed  32 Ormel J. Neuroticism and well-being inventories: measuring traits or states?. Psychol Med . 1983; 13( 1): 165– 176. Google Scholar CrossRef Search ADS PubMed  33 Humphreys DK, Panter J, Sahlqvist S, et al.  . Changing the environment to improve population health: a framework for considering exposure in natural experimental studies. J Epidemiol Community Health . 2016; 70( 9): 941– 946. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Epidemiology Oxford University Press

An Olympic Legacy? Did the Urban Regeneration Associated With the London 2012 Olympic Games Influence Adolescent Mental Health?

Loading next page...
 
/lp/ou_press/an-olympic-legacy-did-the-urban-regeneration-associated-with-the-Mt86CFhLBi
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0002-9262
eISSN
1476-6256
D.O.I.
10.1093/aje/kwx205
Publisher site
See Article on Publisher Site

Abstract

Abstract Public expenditure on large events such as the London 2012 Olympic Games is often justified by the potential legacy of urban regeneration and its associated health and well-being benefits for local communities. In the Olympic Regeneration in East London Study, we examined whether there was an association between urban regeneration related to the 2012 Games and improved mental health in young people. Adolescents aged 11–12 years attending schools in the Olympic host borough of Newham in London or in 3 adjacent comparison London boroughs completed a survey before the 2012 Games and 6 and 18 months after the Games (in 2013 and 2014, respectively). Changes in depressive symptoms and well-being between baseline and each follow-up were examined. A total of 2,254 adolescents from 25 randomly selected schools participated. Adolescents from Newham were more likely to have remained depressed between baseline and the 6- and 18-month follow-up surveys (for 6-month follow-up, relative risk = 1.78, 95% confidence interval: 1.12, 2.83; for 18-month follow-up, relative risk = 1.93, 95% confidence interval: 1.01, 3.70) than adolescents from the comparison boroughs. No differences in well-being were observed. There was little evidence that urban regeneration had any positive influence on adolescent mental health and some suggestion that regeneration may have been associated with maintenance of depressive symptoms. Such programs may have limited short-term impact on the mental health of adolescents. adolescent, depressive symptoms, longitudinal, positive well-being, urban regeneration Public expenditure on large sporting events such as the London 2012 Olympic Games is often justified by the hypothesized positive impact of urban regeneration, including the provision of new sports and recreational facilities, related to hosting such events and the associated legacy benefits for local communities (1). Addressing deprivation by enhancing health and well-being through urban regeneration programs was among the legacy benefits identified for the London 2012 Olympic Games (2). However, public-health evaluations have thus far found little evidence of a positive association between large-scale investment in urban regeneration at the household, dwelling, community, or neighborhood levels with either mental health or social determinants of mental health (3, 4). In fact, urban regeneration may negatively impact mental health (3–6) via increased stress associated with the environmental nuisance of regeneration (6), gentrification (5, 7), changes to social networks (8), displacement (1), loss of amenities (1), lack of control over the planning process (1, 8), and an increase in relative deprivation among residents who do not benefit. Nevertheless, regeneration may impact positively on children and adolescents from these communities (7), via increasing feelings of safety, reducing exposure to stressors, and increasing access to amenities. However, there have been few studies in which the associations of this sort of urban regeneration with the mental health and well-being of children and adolescents have been evaluated (4). The Olympic Regeneration in East London (ORiEL) Study examined whether urban regeneration occurring around the 2012 London Olympic Games (hereafter, the Games) was associated with improved adolescent mental health and well-being among adolescents who lived in those areas compared with those who lived in areas with less or no Olympic-related urban regeneration. We hypothesized that adolescents living in an area receiving most of the Games-related urban regeneration would have greater positive change and better mental health 6 and 18 months after the regeneration than adolescents living in comparison areas receiving less or no Games-related urban regeneration. METHOD Study design and participants This longitudinal, quasi-experimental study followed for 3 years a cohort of adolescents recruited from randomly selected schools in the London boroughs of Barking and Dagenham, Hackney, Newham, and Tower Hamlets. The participants, ages 11–12 years in year 7 at baseline (January–June 2012), were first followed up in year 8 (ages 12–13 years; January–June 2013) 6 to 11 months after the completion of the Games, and again in year 9 (ages 13–14 years; January–June 2014) 18–23 months after the Games. Ethical approval was granted by the Queen Mary University of London Research Ethics Committee (QMREC2011/40), the Association of Directors of Children’s Services (RGE110927), and the London Boroughs Research Governance Framework (CERGF113). Head teachers gave written consent for the study in their school; parents gave passive informed consent and could opt their child out of the study; adolescents gave written informed assent for the study. Exposure to urban regeneration associated with the Games The intervention borough was the London borough of Newham, where the 2012 Games primarily were hosted and where the majority of regeneration occurred. Within Newham, regeneration associated with the Games was focused in 3 main areas: Stratford City development, the Olympic Park, and the Olympic Fringe. The main components of this regeneration program are outlined in Table 1. Although part of the retail complex (Westfield Stratford City) had opened in 2011–2012, before baseline data collection, large areas where regeneration was occurring were inaccessible to the local communities from 2008 to late 2012. For example, the Olympic Park development involved closing off the area from 2008 and opening for a limited time only for ticket holders for the Games at the end of July 2012; the surrounding communities did not gain regular access to the Olympic Park and its facilities until early 2013. Before 2008, much of the Olympic site was inaccessible, derelict industrial land. Table 1. Main Regeneration Components Associated With the London 2012 Olympic Games in the London Borough of Newham, United Kingdom, Olympic Regeneration in East London Study, 2011–2014 Datea  Area  Main Components  2011–2012  Stratford City Development  Retail and leisure center comprising 1.9 million ft2 (176,516 m2) of retail space (including Westfield Stratford City), 500,000 ft2 (46,452 m2) of office and business space, and new civic and public space  2012–2014  Olympic Park  Regenerated land (246 hectares) comprising new green spaces and parkland, public space, and play areas; world-class sports venues (i.e., main stadium, aquatics center, velodrome, bicycle motocross and mountain bike tracks, road cycle route), and associated facilities; improved physical connectivity and accessibility to the Olympic Park from surrounding areas (i.e., foot and cycle paths, bridges, waterways, road and rail links); new housing associated with the former Athletes village (East Village)  2012–2014  Olympic Fringe  Fringe surrounding the Olympic Park will receive 90 hectares of improved green/civic space and improved connectivity to the main Olympic Park  Datea  Area  Main Components  2011–2012  Stratford City Development  Retail and leisure center comprising 1.9 million ft2 (176,516 m2) of retail space (including Westfield Stratford City), 500,000 ft2 (46,452 m2) of office and business space, and new civic and public space  2012–2014  Olympic Park  Regenerated land (246 hectares) comprising new green spaces and parkland, public space, and play areas; world-class sports venues (i.e., main stadium, aquatics center, velodrome, bicycle motocross and mountain bike tracks, road cycle route), and associated facilities; improved physical connectivity and accessibility to the Olympic Park from surrounding areas (i.e., foot and cycle paths, bridges, waterways, road and rail links); new housing associated with the former Athletes village (East Village)  2012–2014  Olympic Fringe  Fringe surrounding the Olympic Park will receive 90 hectares of improved green/civic space and improved connectivity to the main Olympic Park  a Baseline was January–June 2012; first follow-up was January–June 2013; and second follow-up was January–June 2014. Three adjacent London boroughs (Barking and Dagenham, Hackney, Tower Hamlets) were selected as comparison areas. We hypothesized they would benefit less from the planned regeneration, because they were farther from the Olympic Park and were not direct recipients of the regeneration activities outlined in Table 1. Figure 1 shows the distribution of the schools across the 4 boroughs in East London. Figure 1. View largeDownload slide Map showing the intervention and comparison boroughs of the Olympic Regeneration in East London Study, 2011–2014. Adapted from Smith et al. (10), with permission under license CC-BY-NC 2.0 (https://creativecommons.org/licenses/by-nc/2.0/). Figure 1. View largeDownload slide Map showing the intervention and comparison boroughs of the Olympic Regeneration in East London Study, 2011–2014. Adapted from Smith et al. (10), with permission under license CC-BY-NC 2.0 (https://creativecommons.org/licenses/by-nc/2.0/). Although selection of the intervention and comparison boroughs underpins the epidemiologic design of the study, we also conducted analyses that used a different characterization of “exposure” to the urban regeneration associated with the Games. This measure was the Euclidean distance of the participant’s school to the Olympic Park, calculated using a geographical information system (9). These analyses examined the odds for changes in mental health and well-being for an interquartile increase in distance to the Olympic Park (interquartile range, 3,240.7 m). Procedure Of the 48 secondary schools in the 4 boroughs, we randomly selected 6 to 7 schools in each borough; when a selected school refused to participate, another school within the borough was randomly selected to participate. The study had 80% power to detect an 8% difference in well-being (5% significance level) with an 18-month follow-up sample of 1,766 adolescents from 24 schools (10). This calculation was informed by a study (11) that found well-being scores on a range of scales improved by 8%–25% for adults and children after a neighborhood intervention. Of 41 invited schools, 25 participated (school response rate, 60.9%), resulting in 6 intervention and 19 comparison schools. In 7 schools, all students in year 7 were invited, which led to the recruitment of more than 90 children per school. In the remaining schools, a mixed-ability sample was selected. The sociodemographic characteristics of the baseline sample were similar to those reported in the 2011 United Kingdom national census. Adolescents followed a standardized protocol to complete a questionnaire in their classroom; the questionnaire assessed well-being, mental health, physical activity, and sociodemographic factors. Outcomes Well-being and depressive symptoms were self-reported by the adolescents. The World Health Organization defines well-being as “a state in which every individual realizes his or her own potential, can cope with the normal stresses of life, can work productively, and is able to make a contribution to her or his community” (12). Well-being was assessed by the Warwick-Edinburgh Mental Well-Being Scale (WEMWBS), a 14-item (5 response categories) self-report measure of subjective, positive well-being (13) at baseline and at the 6- and 18-month follow-up study points. Before and after absolute changes in the continuous WEMWBS scores between baseline and the 6- and 18-month follow-up surveys were calculated by subtracting the baseline score from the 6-month or 18-month follow-up score, respectively. Depressive symptoms were assessed by the Short Moods and Feelings Questionnaire, a 13-item (3 response categories) self-report measure (14) valid for use with adolescents. The Short Moods and Feelings Questionnaire has been shown to discriminate a psychiatric sample from a pediatric control sample (14). Dichotomous scores (14) indicative of clinically relevant depressive symptoms were used to determine whether an adolescent had clinically relevant depressive symptoms at each time point (no depressive symptoms: score, 0–7; depressive symptoms: score, ≥8). These dichotomous variables were then used to assess change between baseline and the 6-month follow-up, and baseline and the 18-month follow-up. The resulting categorical outcomes were 1) no depressive symptoms at baseline or follow-up (reference group), 2) change from no depressive symptoms at baseline to depressive symptoms at follow-up (“became depressed”), 3) change from depressive symptoms at baseline to no depressive symptoms at follow-up (“no longer depressed”), and 4) depressive symptoms at baseline and follow-up (“remained depressed”). Covariates available at baseline and both follow-up surveys were identified a priori from existing literature, which demonstrated cross-sectional and longitudinal associations with depressive symptoms in our previous East London, school-based cohort study of adolescents (Research with East London Adolescents: Community Health Survey) conducted between 2001 and 2005 (15–18). The covariates were age (months); sex; ethnicity (assessed using an 11-category variable based on the 2011 United Kingdom Census); number of years lived in the United Kingdom; number of parents the participant lived with; parental income; bullied at school in the past 12 months; receiving free school meals (as a marker of social disadvantage); long-standing illness; number of life events; changed neighborhoods since last survey; and the number of days since the opening ceremony of the Games that the questionnaire was completed. Social support from family and friends was assessed using the Multidimensional Scale of Perceived Social Support (19), with scores divided into tertiles representing low, medium, and high support. Statistical analyses Weights were derived to take into account unequal probabilities of school and pupil selection. Prevalence and rates of missing data for the outcomes and covariates were examined; missing values ranged from 0.0% to 45.2%. We explored patterns and predictors of missing observations through logistic regression modelling. Analyses suggested that data were missing at random (MAR) (20). Data were imputed using multilevel multiple imputation in Realcom software (Realcom Applications, Coleorton, Leicestershire, UK) (21), which uses a joint, multivariate, normal modelling approach through the Markov Chain Monte Carlo method. We imputed with 2 levels (first, survey time (baseline, 6-month follow-up or 18-month follow-up); second, adolescent) with all the outcomes and covariates as fixed effects. Interaction terms between sex and the intervention, and free school meals and the intervention were also included. The imputation model was chosen to be congenial (22) with the most saturated model of interest; auxiliary variables were included to strengthen the MAR assumption. We used a “burn in” period of 25,000 iterations, followed by 50,000 iterations, producing a dataset every 1,000th iteration, which resulted in 50 imputed datasets. The Markov Chain Monte Carlo chains were examined to check for convergence. Analyses were carried out using STATA, version 14 (StataCorp LP, College Station, Texas) (23). Crude and adjusted linear, logistic, and multinomial regression models were run to assess the associations of urban regeneration (intervention vs comparison) with short-term (wave 1 to wave 2) and longer-term (wave 1 to wave 3) change in well-being and depressive symptoms. The models were adjusted for baseline demographic factors, household factors, family and friend social support, and psychological factors (Table 2). An additional adjustment for baseline well-being, as determined by the WEMWBS, was made to the adjusted model for well-being to test sensitivity of the findings for baseline well-being. Interactions between urban regeneration (i.e., borough or distance to the Olympic Park) with sex and free schools meals were tested; models were stratified and interactions were statistically significant when P ≤ 0.05. Inference sensitivity to departure from MAR was explored through tipping-point sensitivity analysis in which data were imputed under “missing not at random” with increasing departure from the MAR assumption. Inferences were robust to departures from the MAR assumption. Table 2. Descriptive Statistics for Key Measures at Each Survey, Olympic Regeneration in East London Study, London, United Kingdom 2011–2014 Parameter  Baseline Survey  6-Month Follow-Up Survey  18-Month Follow-Up Survey  No. of Participants  %  Missing %  No.  %  Missing %  No.  %  Missing %  Exposure  Intervention  2,254    0.0               Comparison boroughs  1,626  72.1                 Newham borough  628  27.9                Outcome Measures  Short Moods and Feelings Questionnaire  2,068    8.2  2,116    6.1  2,155    4.4   Not depressed  1,618  78.2    1,676  79.2    1,633  75.8     Depressed (≥8)  450  21.8    440  20.8    522  24.2      Scale reliabilitya  0.87    0.90    0.91    Warwick-Edinburgh Mental Wellbeing Scale  1,943    13.8  2,016    10.6  2,085    7.5    Scale scoreb  51.2 (9.0)    51.6 (9.7)    51.2 (10.1)      Scale reliabilitya  0.84    0.88    0.90    Baseline Demographic Measures  Sex  2,254    0.0               Male  1,271  56.4                 Female  983  43.6                Ethnicity  2,254    0.0               White: United Kingdom  380  16.9                 White: other  326  14.4                 White: mixed  190  8.4                 Asian: Indian  85  3.8                 Asian: Pakistani  86  3.8                 Asian: Bangladeshi  337  14.9                 Asian: other  72  3.2                 Black: Caribbean  111  4.9                 Black: African  249  11.0                 Black: other  263  11.7                 Other  155  6.9                Time in the United Kingdom, years  2,221    1.5               All my life  1,629  73.3                 >10  174  7.8                 6–10  190  8.6                 <6  228  10.3                Household Measures  No. of parents working  1,984    12.0  2,104    6.7  2,090    7.3   2  796  40.1    813  38.6    832  39.8     1  832  41.9    883  42.0    891  42.6     0  356  17.9    408  19.4    367  17.6    Household composition  2,221    1.5  2,234    0.9  2,238    0.7   Lives with both parents  1,503  67.7    1,544  69.1    1,510  67.5     Lives with 1 parent  673  30.3    643  28.8    688  30.7     Lives with no parents  45  2.0    47  2.1    40  1.8    Moved neighborhood in past year (baseline) or since previous survey  1,914    15.1  2,137    5.2  2,167    3.9   No  1,746  91.2    1,977  92.5    2,022  93.3     Yes  168  8.8    160  7.5    145  6.7    Receives free school meals  2,208    2.0  2,214    1.8  2,219    1.5   No  1,376  62.3    1,404  63.4    1,503  67.7     Yes  832  37.7    810  36.6    716  32.3    Social Support Measures  Tertile of multidimensional Scale of Perceived Social Support: family  1,236    45.2  1,663    26.2  1,962    13.0   Low  378  30.6    530  31.9    642  32.7     Medium  392  31.7    475  28.6    586  29.9     High  466  37.7    658  39.5    734  37.4    Multidimensional Scale of Perceived Social Support: friend; tertile  1,239    45.0  1,656    26.5  1,958    13.1   Low  349  28.2    495  29.9    649  33.1     Medium  470  37.9    592  35.7    634  32.4     High  420  33.9    569  34.4    675  34.5    Psychological Measures  Ever bullied  1,463    35.1               No  948  64.8                 Yes  515  35.2                No of negative life events  1,663    26.2  1,910    15.3  2,001    11.2   0  865  52.0    691  36.2    481  24.0     1  373  22.5    469  24.6    499  24.9     2  235  14.1    334  17.5    372  18.6     ≥3  190  11.4    416  21.8    649  32.4    Long-term illness  2,153    4.5               No  1,898  88.2                 Yes  255  11.8                Parameter  Baseline Survey  6-Month Follow-Up Survey  18-Month Follow-Up Survey  No. of Participants  %  Missing %  No.  %  Missing %  No.  %  Missing %  Exposure  Intervention  2,254    0.0               Comparison boroughs  1,626  72.1                 Newham borough  628  27.9                Outcome Measures  Short Moods and Feelings Questionnaire  2,068    8.2  2,116    6.1  2,155    4.4   Not depressed  1,618  78.2    1,676  79.2    1,633  75.8     Depressed (≥8)  450  21.8    440  20.8    522  24.2      Scale reliabilitya  0.87    0.90    0.91    Warwick-Edinburgh Mental Wellbeing Scale  1,943    13.8  2,016    10.6  2,085    7.5    Scale scoreb  51.2 (9.0)    51.6 (9.7)    51.2 (10.1)      Scale reliabilitya  0.84    0.88    0.90    Baseline Demographic Measures  Sex  2,254    0.0               Male  1,271  56.4                 Female  983  43.6                Ethnicity  2,254    0.0               White: United Kingdom  380  16.9                 White: other  326  14.4                 White: mixed  190  8.4                 Asian: Indian  85  3.8                 Asian: Pakistani  86  3.8                 Asian: Bangladeshi  337  14.9                 Asian: other  72  3.2                 Black: Caribbean  111  4.9                 Black: African  249  11.0                 Black: other  263  11.7                 Other  155  6.9                Time in the United Kingdom, years  2,221    1.5               All my life  1,629  73.3                 >10  174  7.8                 6–10  190  8.6                 <6  228  10.3                Household Measures  No. of parents working  1,984    12.0  2,104    6.7  2,090    7.3   2  796  40.1    813  38.6    832  39.8     1  832  41.9    883  42.0    891  42.6     0  356  17.9    408  19.4    367  17.6    Household composition  2,221    1.5  2,234    0.9  2,238    0.7   Lives with both parents  1,503  67.7    1,544  69.1    1,510  67.5     Lives with 1 parent  673  30.3    643  28.8    688  30.7     Lives with no parents  45  2.0    47  2.1    40  1.8    Moved neighborhood in past year (baseline) or since previous survey  1,914    15.1  2,137    5.2  2,167    3.9   No  1,746  91.2    1,977  92.5    2,022  93.3     Yes  168  8.8    160  7.5    145  6.7    Receives free school meals  2,208    2.0  2,214    1.8  2,219    1.5   No  1,376  62.3    1,404  63.4    1,503  67.7     Yes  832  37.7    810  36.6    716  32.3    Social Support Measures  Tertile of multidimensional Scale of Perceived Social Support: family  1,236    45.2  1,663    26.2  1,962    13.0   Low  378  30.6    530  31.9    642  32.7     Medium  392  31.7    475  28.6    586  29.9     High  466  37.7    658  39.5    734  37.4    Multidimensional Scale of Perceived Social Support: friend; tertile  1,239    45.0  1,656    26.5  1,958    13.1   Low  349  28.2    495  29.9    649  33.1     Medium  470  37.9    592  35.7    634  32.4     High  420  33.9    569  34.4    675  34.5    Psychological Measures  Ever bullied  1,463    35.1               No  948  64.8                 Yes  515  35.2                No of negative life events  1,663    26.2  1,910    15.3  2,001    11.2   0  865  52.0    691  36.2    481  24.0     1  373  22.5    469  24.6    499  24.9     2  235  14.1    334  17.5    372  18.6     ≥3  190  11.4    416  21.8    649  32.4    Long-term illness  2,153    4.5               No  1,898  88.2                 Yes  255  11.8                a Values are expressed as Cronbach α. b Values are expressed as mean (standard deviation). RESULTS Participants’ characteristics Table 2 lists the prevalence of the outcomes and covariates at each time point. Of the participants, 27.9% were from schools in the intervention borough and 72.1% from schools in the comparison boroughs. The distance from the participants’ schools to the Olympic Park ranged from 1,133 m to 12,589 m (interquartile range, 3,240.7 m). Participants were ethnically diverse; the largest groups described themselves as white from the United Kingdom (English/Welsh/Scottish/Northern Irish/British) (16.9%); Asian: Bangladeshi (14.9%); or white of any other background (excluding UK, Irish, and Gypsy or Irish Traveller) (14.4%). More than one-fifth of the sample reported depressive symptoms at each time point (baseline, 21.8%; 6-month follow-up, 20.8%; 18-month follow-up, 24.2%). The mean well-being score was 51 at each time point. Urban regeneration and depressive symptoms Depressive symptoms were higher in the adolescents in the intervention borough at baseline than in those in the comparison boroughs (27% vs 20%; unadjusted odds ratio (OR) = 1.49; P = 0.024). Table 3 lists the changes in depressive symptoms between baseline and the 6-month follow-up. In the fully adjusted models, at the 6-month follow-up, adolescents who were no longer depressed (relative risk = 1.53, 95% confidence interval (CI): 1.07, 2.20) or who remained depressed (relative risk = 1.78, 95% CI: 1.12, 2.83) were more likely to be from the intervention borough than from the comparison boroughs. These data reflect, at least in part, the higher prevalence of depression in the intervention borough at baseline. Table 3. Associations of Urban Regeneration With Change in Depressive Symptoms (Baseline to 6-Month Follow-up and Baseline to 18-Month Follow-up), Olympic Regeneration in East London Study, London United Kingdom, 2011–2014 Interval and Intervention  Became Depresseda  No Longer Depresseda  Remain Depresseda  Unadjusted  Adjustedb  Unadjusted  Adjustedb  Unadjusted  Adjustedb  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  Baseline to 6-month follow-up                           Intervention boroughc  1.32  0.81, 2.12  1.44  0.95, 2.17  1.44d  1.03, 2.02  1.53d  1.07, 2.20  1.70d  1.04, 2.77  1.78d  1.12, 2.83   Distance to Olympic Parke  1.08  0.95, 1.23  0.96  0.81, 1.15  1.10  0.95, 1.27  1.04  0.89, 1.20  1.02  0.89, 1.16  0.88  0.72, 1.07  Baseline to 18-month follow-up                           Intervention boroughc  1.23  0.82, 1.83  1.30  0.97, 1.76  1.53d  1.07, 2.17  1.39  0.88, 2.18  1.57  0.83, 2.95  1.93d  1.01, 3.70   Distance to Olympic Parke  1.16d  1.02, 1.31  1.05  0.89, 1.23  1.03  0.89, 1.20  1.03  0.85, 1.25  1.12  0.96, 1.29  0.99  0.76, 1.29  Interval and Intervention  Became Depresseda  No Longer Depresseda  Remain Depresseda  Unadjusted  Adjustedb  Unadjusted  Adjustedb  Unadjusted  Adjustedb  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  RR  95% CI  Baseline to 6-month follow-up                           Intervention boroughc  1.32  0.81, 2.12  1.44  0.95, 2.17  1.44d  1.03, 2.02  1.53d  1.07, 2.20  1.70d  1.04, 2.77  1.78d  1.12, 2.83   Distance to Olympic Parke  1.08  0.95, 1.23  0.96  0.81, 1.15  1.10  0.95, 1.27  1.04  0.89, 1.20  1.02  0.89, 1.16  0.88  0.72, 1.07  Baseline to 18-month follow-up                           Intervention boroughc  1.23  0.82, 1.83  1.30  0.97, 1.76  1.53d  1.07, 2.17  1.39  0.88, 2.18  1.57  0.83, 2.95  1.93d  1.01, 3.70   Distance to Olympic Parke  1.16d  1.02, 1.31  1.05  0.89, 1.23  1.03  0.89, 1.20  1.03  0.85, 1.25  1.12  0.96, 1.29  0.99  0.76, 1.29  Abbreviations: CI, confidence interval; RR, relative risk. a Reference group for the outcome is “remain not depressed.” b Adjusted for sex, age, ethnicity, length of time lived in the United Kingdom, number of days since the Olympic Games, parental income, number of parents participant lives with, moved neighborhood since baseline, eligible for free school meals, family social support, friend social support, bullying, negative life events, and long-term illness. c Intervention London borough is Newham; comparison London boroughs are Hackney, Tower Hamlets, and Barking and Dagenham. dP ≤ 0.05. e Distance to Olympic Park is estimated per interquartile increase in distance (meters). Post hoc analysis restricted to those with baseline depression showed that adolescents from the intervention area were not more likely to have recovered at the 6-month follow-up compared with those from the comparison area (OR = 0.72, 95% CI: 0.44, 1.18). Indeed, among those with no baseline depression, adolescents from the intervention area were more likely to have become depressed by the 6-month follow-up than those from the comparison area (OR = 1.42, 95% CI: 1.09, 1.85). Distance to the Olympic Park was not associated with changes in depressive symptoms between baseline and the 6-month follow-up. There were no interactions between borough or distance to the Olympic Park and sex or eligibility for free school meals relative to change in depressive symptoms between baseline and the 6-month follow-up (P > 0.05). Table 3 also lists the changes in depressive symptoms between baseline and the 18-month follow-up. In the fully adjusted models, adolescents who remained depressed at the 18-month follow-up were more likely to be from the intervention borough than from the comparison boroughs (relative risk = 1.93, 95% CI: 1.01, 3.70). Distance to the Olympic Park was not associated with changes in depressive symptoms between baseline and the 18-month follow-up (Table 3). No interactions were observed between borough or distance to the Olympic Park and sex or eligibility for free school meals relative to change in depressive symptoms between baseline and the 18-month follow-up (P > 0.05). Post hoc analyses showed that for those with no baseline depression, the adolescents from the intervention area were more likely to have become depressed at the 18-month follow-up than those from the comparison area (OR = 1.38, 95% CI: 1.06, 1.80). For adolescents with baseline depression, those from the intervention area were not more likely to have recovered at the 18-month follow-up (OR = 0.67, 95% CI: 0.41, 1.10). Urban regeneration and change in well-being Well-being scores were lower in the intervention borough at baseline than in the comparison boroughs (mean score (standard error), 50.7 (0.6) vs 53.0 (0.2); P = 0.001). There was little change in well-being scores between the surveys (mean (standard error) baseline to 6-month follow-up, μ = −0.02 (0.03); baseline to 18-month follow-up, μ = −0.04 (0.04)). According to the data in Table 4, there was no association between borough and change in well-being between baseline and either follow-up survey. Distance to the Olympic Park was not associated with changes in well-being between baseline and either follow-up survey (Table 4). No interactions were observed between borough or distance to the Olympic Park and sex or eligibility for free school meals in relation to change in well-being between baseline and either follow-up survey. Table 4. Associations of Urban Regeneration With Change in Well-Being (Baseline to 6-Month Follow-up and Baseline to 18-Month Follow-up), Olympic Regeneration in East London Study, London United Kingdom, 2011–2014 Parameter  Baseline to 6-Month Follow-Up  Baseline to 18-Month Follow-Up  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Co-efficient  95% CI  Co-efficient  95% CI  Intervention boroughb  0.58  −1.21, 2.36  0.18  −1.45, 1.80  −0.50  −1.86, 0.85  1.23  −0.74, 3.20  0.52  −0.95, 1.99  −0.07  −1.59, 1.44  Distance to Olympic Parkc  −0.18  −0.93, 0.57  0.01  −0.76, 0.78  0.31  −0.20, 0.81  −0.60  −1.46, 0.26  −0.01  −0.72, 0.70  0.11  −0.39, 0.62  Parameter  Baseline to 6-Month Follow-Up  Baseline to 18-Month Follow-Up  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Unadjusted  Adjusteda  Adjusteda + Baseline WEMWBS  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Coefficient  95% CI  Co-efficient  95% CI  Co-efficient  95% CI  Intervention boroughb  0.58  −1.21, 2.36  0.18  −1.45, 1.80  −0.50  −1.86, 0.85  1.23  −0.74, 3.20  0.52  −0.95, 1.99  −0.07  −1.59, 1.44  Distance to Olympic Parkc  −0.18  −0.93, 0.57  0.01  −0.76, 0.78  0.31  −0.20, 0.81  −0.60  −1.46, 0.26  −0.01  −0.72, 0.70  0.11  −0.39, 0.62  Abbreviations: CI, confidence interval; WEMWBS, Warwick-Edinburgh Mental Well-Being Scale. a Adjusted for sex, age, ethnicity, length of time lived in the United Kingdom, number of days since the Olympic Games, parental income, number of parents participants lives with, moved neighborhood since baseline, eligible for free school meals, family social support, friend social support, bullying, negative life events, and long-term illness. b Intervention London borough is Newham; the comparison London boroughs are Hackney, Tower Hamlets, and Barking and Dagenham. c Distance to Olympic Park is estimated per interquartile increase in distance (meters). DISCUSSION The findings of this study indicate that as a population-health intervention (7, 24), the urban regeneration associated with the London 2012 Olympic Games had little or no detectable positive influence on changes in adolescent mental health in terms of depressive symptoms or well-being. Although those who were depressed at baseline were more likely to have improvement in their symptoms at the 6-month follow-up in the intervention borough compared with adolescents in the control boroughs, this was the only positive association observed and should be considered in the light of more adolescents having depressive symptoms at baseline in the intervention borough. This association was not maintained 18-months after regeneration nor was it replicated in other analyses. In fact, attending school in the intervention borough was associated with a greater chance of maintaining depressive symptoms. Our hypotheses that adolescents in an area of urban regeneration would experience greater positive change and better mental health after the regeneration, compared with adolescents in areas of less or no urban regeneration, were not supported by the study data. These conclusions were further supported by analyses using a different characterization of exposure to the intervention: distance of the participant’s school to the Olympic Park. Urban regeneration and adolescent depressive symptoms The higher levels of depressive symptoms, in terms of remaining depressed between baseline and the 6- and the 18-month follow-up surveys, in adolescents living in the Olympic host borough receiving urban regeneration replicate the findings of some previous studies of urban regeneration in adult populations (3, 4). There was evidence that regeneration was associated with becoming depressed for those with no baseline depression. Urban regeneration can be associated with increased feelings of social isolation (25), reduced social capital (25), increased exposure to stress (26), and relative deprivation, all of which may influence mental health. Regeneration may not address residents’ concerns (1, 5) and may not influence psychosocial, lifestyle, safety, or economic determinants of mental health (6, 27). Interviews with families in the ORiEL Study revealed that during the Games, residents in the intervention borough felt their environment was safer and more unified, but they appreciated this was temporary, with little impact on more immediate concerns such as poor housing and opportunities for employment (28). However, adolescents in the intervention borough had higher rates of depressive symptoms at baseline and at both follow-up visits than did adolescents in the comparison boroughs, which may be explained by other unmeasured differences among the boroughs such as social, economic, and environmental determinants of mental health (6, 27), including income, employment, physical activity, and diet. Whether these determinants have associations with urban regeneration is being explored in a study of the parents of adolescents in the ORiEL Study. Borough-level differences in depressive symptoms may partially account for the findings reported here. A study conducted between 2001 and 2005 of adolescents in 3 of the London boroughs included in the ORiEL Study (i.e., Newham, Tower Hamlets, and Hackney) reported the highest rates of depressive symptoms in Hackney and the lowest rates in Newham (29). This finding suggests differences between boroughs in rates of depressive symptoms are changeable over a relatively short time. The association between living in the Olympic host borough and a greater likelihood of remission of depressive symptoms at 6 months but not 18 months after the Games may reflect temporary positive associations with regeneration (28). However, this finding may also reflect selection into this outcome group by baseline depressive symptoms. We could not replicate the association of borough on remission from depressive symptoms in analyses restricted to those with baseline depression, which may indicate chance findings. In contrast to the findings characterizing exposure to the urban regeneration by borough, we found that distance of the participant’s school from the Olympic Park was not associated with change in depressive symptoms between baseline and either of the 2 follow-up surveys. This difference in findings may reflect the more robust methodology in the ORiEL Study for examining regeneration by borough than by distance to the Olympic Park, because the sample was selected on borough. Urban regeneration and adolescent well-being There were no differences in change in well-being between the intervention and comparison boroughs or by distance to the Olympic Park, which suggests no positive association between Games-related regeneration and well-being. In our analyses, few factors predicted positive or negative change in well-being, which may be because we observed little change in well-being over time. Negative change was most strongly associated with baseline well-being. Authors of a recent systematic review of interventions using the WEMWBS found large variability in change scores for different interventions (30), but there were few studies in which associations with neighborhood interventions had been examined. Our study findings, along with those of 1 study of adults (31), suggest the WEMWBS may not be particularly sensitive to change for evaluating neighborhood population-health interventions. In our population, well-being appeared to be more of a stable trait than a changing state (32). Limitations and strengths Exposure to Games-associated regeneration was determined by school location, which may have resulted in some exposure misclassification. It is difficult to characterize the exact dose, or when regeneration begins and ends, for regeneration associated with large events like the Olympic Games; a small degree of regeneration may have started before our baseline survey and continued beyond our 18-month follow-up. There may be contamination between the intervention and comparison boroughs due to geographic proximity. Deprivation varied little between schools; however, area-level confounding by deprivation may remain. The regeneration was highly localized, resulting in much of Newham remaining unchanged. Furthermore, demographic changes in East London in the past decade, such as increases in population, ethnic diversity, and the private renting of flats could also have influenced the findings. Positive associations of urban regeneration with mental health may be demonstrated over a longer time or may have been observed immediately after the Games, but this was not tested in this study. This study reports results of analyses that characterize urban regeneration as a holistic event. Additional ongoing analyses will examine specific neighborhood environmental changes associated with the regeneration activities, such as access to greenness and walkability, and associations with health (33). Strengths of this study include the large representative sample indicative of the ethnic superdiversity of East London. To our knowledge, this is one of the first studies to examine the longitudinal associations of urban regeneration on the mental health of an adolescent population. The study had high response rates, thus overcoming a key limitation of many previous studies (4). Few participants relocated during the study, attrition was low, and multiple imputation was used to take into account missing data. Conclusion Complex public-health interventions like urban regeneration are very challenging to evaluate (7, 33). This large-scale, quasi-experimental study provides little evidence that the urban regeneration due the London 2012 Games was associated with adolescent mental health in terms of reduced depressive symptoms or increased well-being. In fact, urban regeneration may have maintained depressive symptoms in the studied group of adolescents, but this may be explained by differences between boroughs in the social and economic determinants of mental health. The predicted legacy benefit of the Games on enhancing well-being (2) was not observed here. Regeneration may have maintained and contributed to the onset of depressive symptoms. Any beneficial effects of regeneration may be elusive or may take longer to appear. ACKNOWLEDGMENTS Author affiliations: Centre for Psychiatry, Barts and the London School of Medicine, Queen Mary University of London, London, United Kingdom (Charlotte Clark, Melanie Smuk, Amanda Fahy, Neil Smith, Stephen A. Stansfeld); Department of Social and Environmental Health Research, London School of Hygiene and Tropical Medicine, London, United Kingdom (Steven Cummins, Daniel Lewis); Centre for Primary Care and Public Health, Barts and the London School of Medicine, Queen Mary University of London, London, United Kingdom (Sandra Eldridge, Stephanie J. C. Taylor); and School of Psychology, University of Surrey, Guildford, United Kingdom (Derek G. Moore). All authors contributed equally to this work. This work was funded by the National Institute for Health Research's Public Health Research Programme (grant 09/3005/09). We thank members of the Olympic Regeneration in East London Study research team, including Dr. Ellen Flint, Dr. Mark Petticrew, Dr. Adrian Renton, Vanathi Tharmaratnam, Dr. Claire Thompson, Danielle House, Shaneka Foster, Bukola Thompson, Rebecca Evans, Lianne Austin, Janat Hossain, and all other team members who helped with the study. The funders of the study had no role in the study design; data collection, analysis, or interpretation; or writing of the report. Conflict of interest: none declared. Abbreviations CI confidence interval MAR missing at random OR odds ratio ORiEL Olympic Regeneration in East London WEMWBS Warwick Edinburgh Mental Well-Being Scale REFERENCES 1 Clark J, Kearns A, Cleland C. Spatial scale, time and process in mega-events: the complexity of host community perspectives on neighbourhood change. Cities . 2016; 53: 87– 97. Google Scholar CrossRef Search ADS   2 Host Boroughs Joint Committee. Strategic Regeneration Framework: an Olympic Legacy for the Host Boroughs. London, UK: London Boroughs of Greenwich, Hackney, Newham, Tower Hamlets & Waltham Forest; 2009. http://www.gamesmonitor.org.uk/files/strategic-regeneration-framework-report.pdf. Accessed September 1, 2016. 3 Thomson H, Atkinson R, Petticrew M, et al.  . Do urban regeneration programmes improve public health and reduce health inequalities? A synthesis of the evidence from UK policy and practice (1980–2004). J Epidemiol Community Health . 2006; 60( 2): 108– 115. Google Scholar CrossRef Search ADS PubMed  4 Clark C, Myron R, Stansfeld S, et al.  . A systematic review of the evidence on the effect of the built and physical environment on mental health. J Public Ment Health . 2007; 6( 2): 14– 27. Google Scholar CrossRef Search ADS   5 Huxley P, Evans S, Leese M, et al.  . Urban regeneration and mental health. Soc Psychiatry Psychiatr Epidemiol . 2004; 39( 4): 280– 285. Google Scholar CrossRef Search ADS PubMed  6 Thomas R, Evans S, Huxley P, et al.  . Housing improvement and self-reported mental distress among council estate residents. Soc Sci Med . 2005; 60( 12): 2773– 2783. Google Scholar CrossRef Search ADS PubMed  7 Bond L, Egan M, Kearns A, et al.  . GoWell: the challenges of evaluating regeneration as a population health intervention. Prev Med . 2013; 57( 6): 941– 974. Google Scholar CrossRef Search ADS PubMed  8 Halpern D. Mental Health and the Built Environment . London, UK: Taylor & Francis; 1995. 9 Environmental Systems Research Institute. ArcGIS 10.3 . Redlands, CA: Environmental Systems Research Institute; 2014. 10 Smith NR, Clark C, Fahy AE, et al.  . The Olympic Regeneration in East London (ORiEL) study: protocol for a prospective controlled quasi-experiment to evaluate the impact of urban regeneration on young people and their families. BMJ Open . 2012; 2( 4): e001840. Google Scholar CrossRef Search ADS PubMed  11 Leventhal T, Brooks-Gunn J. Moving to opportunity: an experimental study of neighbourhood effects on mental health. Am J Public Health . 2003; 93( 9): 1576– 1582. Google Scholar CrossRef Search ADS PubMed  12 World Health Organization. Promoting Mental Health: Concepts, Emerging Evidence, and Practice. Summary Report. Geneva, Switzerland: World Health Organization, 2004. 13 Stewart-Brown S, Tennant A, Tennant R, et al.  . Internal construct validity of the Warwick-Edinburgh Mental Well-being Scale (WEMWBS): a Rasch analysis using data from the Scottish Health Education Population Survey. Health Qual Life Outcomes . 2009; 7: 15. Google Scholar CrossRef Search ADS PubMed  14 Angold A, Costello EJ, Messer SC, et al.  . Development of a short questionnaire for use in epidemiological studies of depression in children and adolescents. Int J Meth Psychiatr Res . 1995; 5: 237– 249. 15 Stansfeld SA, Haines MM, Head JA, et al.  . Ethnicity, social deprivation and psychological distress in adolescents: school-based epidemiological study in East London. Br J Psychiatry . 2004; 185( 3): 233– 238. Google Scholar CrossRef Search ADS PubMed  16 Klineberg E, Clark C, Bhui KS, et al.  . Social support, ethnicity and mental health in adolescents. Soc Psychiatry Psychiatr Epidemiol . 2006; 41( 9): 755– 760. Google Scholar CrossRef Search ADS PubMed  17 Fagg J, Curtis S, Clark C, et al.  . Neighbourhood perceptions among inner-city adolescents: relationships with their individual characteristics and with independently assessed neighbourhood conditions. J Environ Psychol . 2008; 28( 2): 128– 142. Google Scholar CrossRef Search ADS   18 Rothon C, Head J, Klineberg E, et al.  . Can social support protect bullied adolescents from adverse outcomes? A prospective study on the effects of bullying on the educational achievement and mental health of adolescents at secondary schools in East London. J Adolesc . 2011; 34( 3): 579– 588. Google Scholar CrossRef Search ADS PubMed  19 Zimet GD, Dahlem NW, Zimet SG, et al.  . The multidimensional scale of perceived social support. J Pers Assess . 1988; 52( 1): 30– 41. Google Scholar CrossRef Search ADS   20 Rubin DB. Multiple Imputation for Nonresponse in Surveys . New York, NY: John Wiley and Sons; 1987. Google Scholar CrossRef Search ADS   21 University of Bristol. REALCOM-Impute Software. Bristol, UK: University of Bristol; 2015. 22 Meng XL. Multiple-imputation inferences with uncongenial sources of input (with discussion). Stat Sci . 1994; 9( 4): 538– 573. Google Scholar CrossRef Search ADS   23 StataCorp. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP; 2015. 24 Craig P, Cooper C, Gunnell D, et al.  . Using natural experiments to evaluate population health interventions: new Medical Research Council guidance. J Epidemiol Community Health . 2012; 66( 12): 1182– 1186. Google Scholar CrossRef Search ADS PubMed  25 Rogers A, Huxley P, Evans S, et al.  . More than jobs and houses: mental health, quality of life and the perceptions of locality in an area undergoing urban regeneration. Soc Psychiatry Psychiatr Epidemiol . 2008; 43( 5): 364– 372. Google Scholar CrossRef Search ADS PubMed  26 Thomas R, Evans S, Gately C, et al.  . State-event relations among indicators of susceptibility to mental distress in Wythenshawe in the UK. Soc Sci Med . 2002; 55( 6): 921– 935. Google Scholar CrossRef Search ADS PubMed  27 Kearns A, Mason P. Regeneration, relocation and health behaviours in deprived communities. Health Place . 2015; 32: 43– 58. Google Scholar CrossRef Search ADS PubMed  28 Thompson C, Lewis DJ, Greenhalgh T, et al.  . “Everyone was looking at you smiling”: East London residents’ experiences of the 2012 Olympics and its legacy on the social determinants of health. Health Place . 2015; 36: 18– 24. Google Scholar CrossRef Search ADS PubMed  29 Institute of Community Health Sciences. Health of Young People in East London. The RELACHS Study . Norwich, UK: The Stationary Office; 2003. 30 Maheswaran H, Weich S, Powell J, et al.  . Evaluating the responsiveness of the Warwick Edinburgh Mental Well-Being Scale (WEMWBS): group and individual level analysis. Health Qual Life Outcomes . 2012; 10: 156. Google Scholar CrossRef Search ADS PubMed  31 Phillips G, Bottomley C, Schmidt E, et al.  . Well London Phase-1: results among adults of a cluster-randomised trial of a community engagement approach to improving health behaviours and mental well-being in deprived inner-city neighbourhoods. J Epidemiol Community Health . 2014; 68( 7): 606– 614. Google Scholar CrossRef Search ADS PubMed  32 Ormel J. Neuroticism and well-being inventories: measuring traits or states?. Psychol Med . 1983; 13( 1): 165– 176. Google Scholar CrossRef Search ADS PubMed  33 Humphreys DK, Panter J, Sahlqvist S, et al.  . Changing the environment to improve population health: a framework for considering exposure in natural experimental studies. J Epidemiol Community Health . 2016; 70( 9): 941– 946. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

American Journal of EpidemiologyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off