Psych Symptoms And Health

RESEARCH REPORT

doi:10.1111/j.1360-0443.2006.01621.x

Psychological symptoms and physical health and health behaviours in adolescents: a prospective 2-year study in East London Charlotte Clark1, Mary M. Haines1, Jenny Head1, Emily Klineberg1, Muna Arephin1, Russell Viner2, Stephanie J. C. Taylor3, Robert Booy4, Kam Bhui1 & Stephen A. Stansfeld1 Centre for Psychiatry,Wolfson Institute of Preventive Medicine, Barts and the London, Queen Mary’s School of Medicine and Dentistry, University of London, UK,1 Department of Paediatrics, Royal Free and University College London Medical School, University College London, UK,2 Centre for Health Science, Barts and the London, Queen Mary’s School of Medicine and Dentistry, University of London, UK3 and Department of Child Health, Barts and the London, Queen Mary’s School of Medicine and Dentistry, University of London, UK4

ABSTRACT Aims To examine whether physical health and health-risk behaviours in young people are risk factors for psychological distress and depressive symptoms over a 2-year period. Design/setting A 2-year, prospective epidemiological cohort study in East London. Participants A total of 1615 adolescents from the Research with East London Adolescents: Community Health Survey (RELACHS)—a representative cohort of young people aged 11–12 and 13–14 years at baseline, followed-up after 2 years. Measurements Psychological distress and depressive symptoms identified by the self-report Strengths and Difficulties Questionnaire and the Short Moods and Feelings Questionnaire at baseline and follow-up. Data on overweight/obesity, general health, long-standing illness, physical activity, smoking, alcohol use and drug use were collected from questionnaires completed by the adolescents at baseline and follow-up. Findings At follow-up, 10.1% of males and 12.9% of females reported psychological distress; 20% of males and 33.7% of females reported depressive symptoms. Having tried drugs or engaged in two or more health-risk behaviours (smoking, alcohol use or drug use) at baseline predicted psychological distress and depressive symptoms at follow-up. Smoking on its own, long-standing illness, obesity/overweight and activity levels were not associated with later psychological health. Risk of poor psychological health at follow-up was associated strongly with psychological health at baseline. Conclusions Psychological health at baseline was the strongest predictor of psychological health at follow-up. Engaging in two or more health-risk behaviours moderately increased the risk of poor psychological health, suggesting that prevention strategies targeting co-occuring substance use may reduce burden of disease. Keywords

Adolescence, alcohol, drug use, psychological health, smoking.

Correspondence to: Charlotte Clark, Centre for Psychiatry, Wolfson Institute of Preventive Medicine, Barts and the London, Queen Mary’s School of Medicine and Dentistry, University of London, Old Anatomy Building, Charterhouse Square, London, EC1M 6BQ, UK. E-mail: [email protected] Submitted 7 December 2005; initial review completed 7 March 2006; final version accepted 23 June 2006

INTRODUCTION For many individuals mental disorders begin in early life [1] and there is increasing evidence of continuity between common mental disorders in adolescence and early adulthood [2–5]. Rates of psychological ill-health in adolescents are increasing [6] and identifying modifiable risk factors for adolescent psychological health becomes increasingly important, given the persistence of adolescent psychological health into adulthood.

Increasing rates of psychological ill health could be linked to increasing rates of health-risk behaviours such as smoking, alcohol and drug use [7]. Unlike innate risk factors, such as gender, health-risk behaviours are potentially modifiable through health promotion campaigns. Prospective studies have investigated the role of smoking, alcohol use and drug use in the aetiology of depression; however, it has been difficult to establish the pathways between substance use and psychological health, because of the considerable overlap in predictors [8]. Some studies have found that smoking leads to depression [9–12];

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135

Health behaviour predictors of psychological symptoms

however, there is stronger evidence that depression precedes smoking [9,12–19]. Conversely, cannabis and alcohol use may precede depression [8,20,21]. However, a recent study found no prospective association between alcohol use and depression after adjustment for background factors such as socio-economic status and previous psychological health [22]. It has also been suggested that the pathway for cannabis use may differ in adolescence and early adulthood, with poor psychological health preceding cannabis use in adolescence and cannabis use preceding poor psychological health in early adulthood [20]. Despite the predominant focus upon individual health-risk behaviours, we know that health-risk behaviours co-occur in adolescence [7,23,24]. This clustering of health-risk behaviours has been explained by gateway theories, where the use of one substance leads to experimentation and use of other substances [12,25] or by common risk factors relating to substance use [26,27]. Gateway effects are stronger in adolescence than in early adulthood [28]. It is not known whether greater risk for psychological health is related to individual health-risk behaviours or the co-occurrence of health-risk behaviours. Adolescent physical health also contributes to the aetiology of psychological health, yet fewer studies have investigated these relationships. While there is cross-sectional support for associations between physical activity levels, general health and psychological health [29–33], prospective studies of these associations are lacking [34]. Furthermore, the pathways between obesity and psychological health remain to be clarified: overall there is stronger evidence that depression leads to later obesity [35,36] than obesity leading to later depression [37]. Understanding the longitudinal associations between physical health and psychological health will enable the further identification of risk factors, which are potentially modifiable. Using longitudinal data from a community sample of UK adolescents, we examine the contribution of physical health and health-risk behaviours to the development of psychological distress and depressive symptoms. It was hypothesized that engaging in health-risk behaviours (smoking, alcohol use, drug use or the co-occurrence of these behaviours), general health, being overweight or obese, being physically inactive and long-standing illness would increase the risk of later psychological distress and depressive symptoms.

METHODS The RELACHS study (Research with East London Adolescents: Community Health Survey) is a longitudinal school-based questionnaire study of a representative sample of adolescents attending 28 schools in East

127

London. The schools were selected randomly and balanced in terms of single-gender and mixed-gender schools. Two classes from each year group were selected from each school, which gives unequal probabilities of selection. At baseline, 2790 adolescents from year 7 (11– 12 years) and year 9 (12–14 years) took part (84% response rate, 73% from non-white UK ethnicities). The participants were followed-up 2 years later when the adolescents were in year 9 (13–14 years) and year 11 (15– 16 years), resampling 75% of the baseline participants. Reasons for loss to follow-up included having left school between phases (7.8%), being absent from school with no reason (6.5%), parent opt-out (2.4%), being absent for other school activities (1.5%) and pupil opt-out (1.1%). Measures The questionnaire contained the same questions relating to psychological health, physical health and health-risk behaviours at baseline and follow-up. Psychological health Two measures of psychological health were used. Psychological distress was measured by the self-report Strengths and Difficulties Questionnaire (SDQ) [38,39], which is a psychometrically valid instrument for assessing psychological morbidity in British adolescents. An overall score on the SDQ is derived by summing four subscales relating to conduct problems, emotional difficulties, hyperactivity and peer relationship problems. A dichotomous measure of psychological distress was derived, with a score equal or greater than 18 on the total difficulties scale indicating caseness [40]. Depressive symptoms were measured using the Short Moods and Feeling Questionnaire (SMFQ) [41]. A dichotomous measure of depressive symptoms was derived, with a score of 8 or more symptoms indicating a high score. Physical health and health-risk behaviours Physical health and health-risk behaviour outcomes in the baseline questionnaire included general health status (very good, good versus fair, bad, very bad), presence of a long-standing illness (asthma, eczema, epilepsy, diabetes, hearing problems, eyesight problems, hayfever, chronic fatigue syndrome/ME or other chronic illness), physical activity (active versus inactive, exercising more or less than twice a week for an hour [42], smoking behaviour (never smoked versus ever tried or regular smoker) [43], alcohol use (never tried, tried alcohol, consumed alcohol in the past week) [43] and drug use (never tried versus tried drugs—cannabis, glue/solvents/gas, ecstasy, crack, heroin, amphetamines, LSD, cocaine, khat) [43]. A composite measure of health-risk behaviour was created from

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135

128

Charlotte Clark et al.

the smoking behaviour, alcohol use and drug use questions, which measured whether the participant had engaged in none, one, or two or more of these behaviours, as published previously [24]. The composite measure of health-risk behaviour was substance-specific and did not include any additional health-risk behaviours such as sexual activity. In addition, at baseline all participants had their height (Leicester portable stadiometer, CMS Camden Ltd, London) and weight (Tanita Body Fact 300 electronic scales, Tanita UK, Yiewsley, Middlesex) measured by trained researchers working to a strict protocol [44]. These data were used to derive obesity and overweight outcomes for the sample, using the 1990 growth reference (UK 90) definitions [45]. Obesity was defined as a body mass index (BMI) above the 95th percentile and overweight (which included obesity) being above the 85th percentile. Socio-economic status and ethnicity Socio-economic status was measured using data on each pupil’s eligibility for free schools meals, which was provided by schools and Local Education Authorities. Free school meals are known to be a reliable indicator of socio-economic status in UK studies of young people [46]. Ethnicity was assessed using an adaptation of the Census 2001 questions; eight groups were use in the analysis (White UK, White Other, Bangladeshi, Pakistani, Asian, Indian, Black, Mixed, Other). Procedure Parents were informed by letter about the study and passive parental consent was sought (with active optout). The participants were informed verbally about the study and gave written consent to take part. The questionnaire was group administered and completed by the participants in classrooms under the supervision of the research team, who answered any questions and checked questionnaires for missing data. The research team also took physical measurements in the classroom during this session. Ethical approval was obtained from the East London and City Local Research Ethics Committee. Statistical analysis All analyses were weighted to take account of unequal probabilities of selection and differential sample attrition. Logistic regression analyses were conducted to compare the risk for psychological distress or depressive symptoms at follow-up by physical health and health behaviours at baseline (model 1). This model was then additionally adjusted for psychological distress or depressive symptoms at baseline (model 2). Model 1 was additionally conducted

on a subsample with no psychological health problems at baseline (model 3). All models were adjusted for age, gender and their interaction, eligibility for free school meals and ethnicity. In addition, two versions of each model were conducted; the first adjusted for the individual smoking, alcohol and drug health-risk behaviours at baseline, the second adjusted for the composite measure of these health-risk behaviours at baseline. Odds ratios and corresponding 95% confidence intervals were calculated using the logistic regression command in STATA (version 8), with the cluster option specified to take into account the clustering of pupils within schools. This option allows for dependence between pupils in the same school.

RESULTS The analyses were conducted on data from participants who had completed both the baseline and follow-up survey and had complete data on psychological and physical health, as well as health behaviours (unweighted n = 1615 for psychological distress; unweighted n = 1513 for depressive symptoms). Analyses were not stratified by gender, as there were no significant gender interactions between the predictors and outcome measures. Loss to follow-up was greater for respondents from the White UK, White Other and Mixedrace ethnicities but did not differ by age group. Rates of psychological distress and depressive symptoms from baseline to follow-up Table 1 shows the cross-sectional proportion of participants reporting psychological distress and depressive symptoms by gender and age group, at baseline and follow-up. There were no significant gender differences in rates of psychological distress at baseline or follow up (baseline: males 8.9%, females 9.3%; follow up: males 10.1%, females 12.9%). Female participants reported significantly higher rates of depressive symptoms than male participants at baseline and follow-up (baseline 26.4% compared with 18.3%; follow-up 33.7% compared with 20.0%). There were no significant age differences in rates of psychological distress or depressive symptoms at baseline or follow-up. Females were significantly more likely to report depressive symptoms at both baseline and followup, compared with males. Rates of psychological distress and depressive symptoms did not vary significantly by socio-economic status (results not shown), which is consistent with an earlier finding for the baseline data [40]. Table 2 shows the longitudinal association between psychological distress and depressive symptoms at baseline and follow-up. Few young people reported experiencing psychological distress at both baseline and follow-up; however, there was a stronger continuity between

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

0.460 1588

Age group Year 7/9 Year 9/11 P-value for c2 between age group Bases (weighted) 69 (9.8) 77 (8.7)

70 (8.9) 75 (9.3)

620 (87.9) 786 (89.0)

705 (89.9) 700 (87.1)

0.505 1588

0.074 1588 85 (12.1) 97 (11.0)

79 (10.1) 104 (12.9)

496 (78.0) 652 (77.1)

586 (81.7) 564 (73.6)

0.675 1483

< 0.001 1483 141 (22.0) 194 (22.9)

131 (18.3) 202 (26.4)

Case at baseline n (%)

473 (74.4) 608 (71.8)

574 (80.0) 508 (66.3)

Not a case at follow-up n (%)

0.267 1483

< 0.001 1483

Unweighted base = 1615 for SDQ: Strengths and Difficulties Questionnaire (800 for males, 815 for females); unweighted base = 1513 for MFQ: Moods and Feelings Questionnaire (737 for males, 776 for females).

635 (90.2) 807 (91.3)

0.782 1588

714 (91.1) 729 (90.7)

Gender Males Females P-value for c2 between gender Bases (weighted)

Case at follow-up n (%)

Not a case at baseline n (%)

Not a case at follow-up n (%)

Not a case at baseline n (%)

Case at baseline n (%)

MFQ

SDQ

163 (25.6) 239 (28.2)

143 (20.0) 258 (33.7)

Case at follow-up) n (%)

Table 1 Cross-sectional relationship between the proportion of participants with psychological distress (SDQ) and depressive symptoms (MFQ) by gender and age group, at baseline and follow-up (n, %).

Health behaviour predictors of psychological symptoms 129

Addiction, 102, 126–135

130

Charlotte Clark et al.

Table 2 Longitudinal relationship between the proportion of participants with psychological distress (SDQ) and depressive symptoms (MFQ) at baseline and follow-up (n, %). Overall Not a case at follow-up n (%) SDQ Not a case at baseline Case at baseline P-value (McNemar test*) Bases (weighted)† MFQ Not a case at baseline Case at baseline P-value (McNemar test*) Bases (weighted)‡

Males Case at follow-up n (%)

1313 (82.7) 130 (8.2) 93 (5.9) 52 (3.3) 0.016 1588 925 (62.4) 224 (15.1) 156 (10.5) 178 (12.0) < 0.001 1483

Not a case at follow-up n (%)

Females Case at follow-up n (%)

Not a case at follow-up n (%)

Case at follow-up n (%)

655 (83.5) 59 (7.5) 50 (6.4) 20 (2.6) 0.444 784

658 (81.8) 43 (5.3)

499 (69.7) 86 (12.0) 74 (10.3) 57 (8.0) 0.385 716

426 (55.6) 138 (18.0) 82 (10.7) 120 (15.7) < 0.001 766

71 (8.8) 32 (4.1) 0.011 804

*Compares baseline with follow-up; †unweighted base = 1615 for SDQ: Strengths and Difficulties Questionnaire (800 for males, 815 for females); ‡unweighted base = 1513 for MFQ: Moods and Feelings Questionnaire (737 for males, 776 for females).

depressive symptoms at baseline and follow-up: 12.0% reported depressive symptoms at both phases, while only 3.3% reported psychological distress at both phases. Psychological distress, depressive symptoms and physical health and health-related behaviours Tables 3 and 4 compare the physical health and health behaviours at baseline of young people with psychological distress and depressive symptoms at follow-up. Young people with psychological distress and depressive symptoms at follow-up were more likely to have had poor general health, to have consumed alcohol in the past week, to have tried drugs and to have engaged in two or more health-risk behaviours at baseline. Odds ratios for psychological distress at follow-up were significantly higher for young people who also reported psychological distress at baseline (Table 3). Having consumed alcohol in the past week, having tried drugs and having poorer health at baseline predicted psychological distress at follow-up. Engaging in two or more health-risk behaviours at baseline was associated with increased risk of psychological distress at follow-up. These findings were robust after adjustment for psychological distress at baseline, with the exception of having consumed alcohol, which was borderline significant. Few associations were observed for the subsample with no psychological distress at baseline, and only having tried drugs or engaging in two or more health-risk behaviours increased the risk of psychological distress at follow-up. Odds ratios for depressive symptoms at follow-up were significantly higher for young people who reported depressive symptoms at baseline (Table 4). Engaging in

two or more health-risk behaviours at baseline predicted depressive symptoms for those with and without depressive symptoms at baseline. Having tried drugs at baseline was associated with depressive symptoms at follow-up but not for the subsample without depressive symptoms at baseline.

DISCUSSION Having tried drugs or engaged in two or more health-risk behaviours (smoking, alcohol use or drug use) predicted psychological distress and depressive symptoms over a 2-year period in an ethnically diverse sample of young people. Psychological distress was predicted additionally by general health status and having tried or regularly consumed alcohol at baseline. These findings were robust after adjustment for baseline measures of psychological symptoms with the exception of alcohol use, which was borderline significant. Smoking on its own, long-standing illness, obesity/overweight and activity levels were not associated with increased risk for psychological distress or depressive symptoms. The associations between physical health and health-risk behaviours at baseline and psychological health were moderate and stronger associations existed between psychological health at baseline and follow-up. What the results might mean Overall, health-risk behaviours showed moderate relationships with psychological health over a 2-year period, indicating that the targeting of health-risk behaviours may reduce the burden of disease in terms of adolescent

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

421 433 554 571 333 79 827 154 154 317 474 294

Version A Fair, poor or very poor health Long-standing illness Overweight Tried or regular smoker Not consumed alcohol in past week Consumed alcohol in past week Inactive Tried drugs SDQ case at baseline

Version B Obesity

Version C One health-risk behaviour Two or more health-risk behaviours 36.5 24.4

19.2

24.2 26.6 33.9 35.5 21.1 5.3 48.9 8.7 6.3

Not a case

38.8 40.2

23.1

36.8 30.2 39.0 41.0 31.3 13.4 45.6 16.5 3.3

Case

1.17 2.05**

1.10

1.69** 1.04 1.12 0.99 1.43 2.14* 0.88 1.71* –

OR

Model 1† n = 1615

0.77, 1.76 1.30, 3.23

0.71, 1.69

1.21, 2.36 0.77, 1.39 0.79, 1.58 0.72, 1.35 0.76, 2.68 1.01, 4.51 0.67, 1.16 1.10, 2.64 –

95%CI

1.11 1.95**

1.10

1.46* 0.97 1.15 0.98 1.48 2.05 0.91 1.49 5.19***

OR

Model 2† n = 1615

0.75, 1.64 1.18, 3.22

0.72, 1.80

1.03, 2.09 0.69, 1.38 0.82, 1.61 0.71, 1.36 0.78, 2.80 0.94, 4.48 0.69, 1.20 0.97, 2.31 3.32, 8.11

95%CI

426 257

218

352 379 496 508 300 67 724 127 –

n

1.04 1.86*

0.94

1.34 1.23 1.14 0.98 1.38 1.51 0.90 *1.62 –

OR

Model 3†‡ n = 1461

0.65, 1.66 1.05, 3.29

0.61, 1.46

0.84, 2.15 0.84, 1.82 0.81, 1.60 0.67, 1.43 0.73, 2.59 0.69, 3.29 0.64, 1.24 1.02, 2.59

95%CI

***P < 0.001, ** P < 0.01, *P < 0.05. †All models are adjusted for age, gender, gender ¥ age, ethnicity, eligibility for free school meals. Version A further adjusts for general health status, long-standing illness, overweight, smoking, alcohol use, drug use. Version B as Version A but adjusts for obesity rather than overweight. Version C as Version A but adjusts for clustering of health-risk behaviours rather than the individual smoking, alcohol and drug use variables. ‡Model run on subsample who did not have psychological distress at baseline. §Number of respondents reporting the behaviour at baseline out of the total n = 1615: SDQ Strengths and Difficulties Questionnaire.

n§

Baseline risk factors

Proportions with and without psychological distress at follow-up for each baseline factor

Table 3 Multivariate odds ratios for psychological distress at follow-up predicted by baseline physical health and health related behaviours.

Health behaviour predictors of psychological symptoms 131

Addiction, 102, 126–135

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

35.8 24.4

449 278

19.1

23.6 26.2 33.1 34.5 22.1 5.5 50.2 8.0 10.6

Not a case

32.7

40.5

21.4

33.2 27.9 38.9 40.4 23.9 8.0 44.5 15.0 12.1

Case

1.74***

1.42

1.04

1.42*** 1.03 1.20 1.17 1.10 1.38 0.93 1.84*** –

OR

Model 1† n = 1513

1.25, 2.41

0.96, 2.09

0.80, 1.36

1.15, 1.76 0.84, 1.25 0.95, 1.53 0.88, 1.56 0.78, 1.55 0.84, 2.26 0.69, 1.25 1.27, 2.66

95%CI

1.53*

1.28

0.99

1.22 0.96 1.16 1.07 1.03 1.52 0.96 1.79*** 4.34***

OR

Model 2† n = 1513

1.07, 2.19

0.86, 1.90

0.75, 1.30

0.96, 1.55 0.78, 1.18 0.90, 1.50 0.80, 1.43 0.70, 1.50 0.90, 2.56 0.71, 1.29 1.25, 2.57 3.32, 5.68

95%CI

195

335

214

267 286 385 358 228 59 585 102 –

n

1.67*

1.49*

0.92

1.14 0.95 1.20 1.29 0.88 1.39 1.01 1.72 –

OR

Model 3†‡ n = 1170

1.11, 2.51

1.02, 2.18

0.57, 1.48

0.83, 1.56 0.74, 1.22 0.91, 1.57 0.96, 1.72 0.53, 1.47 0.81, 2.36 0.71, 1.45 0.95, 3.10

95%CI

***P < 0.001, *P < 0.05. †All models are adjusted for age, gender, gender ¥ age, ethnicity, eligibility for free school meals. Version A further adjusts for general health status, long-standing illness, overweight, smoking, alcohol use, drug use. Version B as Version A but adjusts for obesity rather than overweight. Version C as Version A but adjusts for clustering of health-risk behaviours rather than the individual smoking, alcohol and drug use variables. ‡Model run on subsample who did not have depressive symptoms at baseline. §Number of respondents reporting the behaviour at baseline out of the total n = 1513: MFQ Moods and Feelings Questionnaire.

Two or more health-risk behaviours

One health-risk behaviour

Version C

Obesity

296

402 401 521 533 320 76 772 148 343

Version A Fair, poor or very poor health Long-standing illness Overweight Tried or regular smoking Not consumed alcohol in past week Consumed alcohol in past week Inactive Tried drugs MFQ case at baseline

Version B

n§

Baseline risk factors

Proportions with and without depressive symptoms at follow-up for each baseline factor

Table 4 Multivariate odds ratios for depressive symptoms at follow-up predicted by baseline physical health and health related behaviours. †

132 Charlotte Clark et al.

Addiction, 102, 126–135

Health behaviour predictors of psychological symptoms

psychological health and early adulthood psychological health. Stronger associations existed between psychological health at baseline and follow-up. Young people with psychological ill-health at baseline were four to five times more likely to experience psychological distress or depressive symptoms at follow-up, suggesting that prevention strategies may be more successful in reducing the burden of disease by focusing on young people with a history of psychological ill-health, as well as health-risk behaviours. In terms of individual health-risk behaviours, drug use showed the strongest associations with later psychological health and our findings support the argument for drug use preceding psychological distress [8,21]. Drug use in this sample, particularly the younger age group (aged 11–12) is indicative of non-normative behaviour, so we should not be surprised that it relates to later psychological health problems. The association between having consumed alcohol at baseline and psychological distress at follow-up was of borderline significance after adjustment for psychological distress at baseline, supporting a previous study which found no prospective association between alcohol use and depression after adjustment for prior psychological health [22]. Together these studies provide limited support for the argument that alcohol use precedes depressive symptoms [20]. Surprisingly, no associations were demonstrated between smoking and later psychological health, contrary to previous findings [9–12]. The lack of replication may be because our sample are younger than in previous studies; few respondents reported smoking regularly (> 1 cigarette a week), and in our analyses regular smokers were combined with respondents who reported having tried smoking, which may have limited the power of the analyses. No gender differences were found in the associations between health-risk behaviours and psychological health, suggesting that risk does not differ for males and females. While our findings are in contrast to those of some recent studies [20,21,47], our sample is younger and it is possible that gender differences in these associations develop in later adolescence. Prevention strategies targeting young people engaging in co-occurring substance use, rather than specific, individual substance use behaviours, may reduce the burden of disease. This study is unable to cast light upon theories of the co-occurrence of health-risk behaviours. Study of the development of the co-occurrence of health-risk factors would require more frequent longitudinal data collection than the 2-year time-frame of this study. From our data it is unclear whether use of one substance leads to experimentation with other substances or whether there are shared risk factors for substance use. Previous studies have found that the key

133

pathways for accessibility to substances are through peers and siblings who are already users [48–52]. Our findings partially support previous findings relating to physical health behaviours and risk of psychological ill-health. We found a significant association between general health and psychological health, supporting previous cross-sectional findings [29,30]. Interestingly, no association was found for long-standing illness, which we might expect to show a similar relationship with psychological health. One reason for this is that perceived general health may encompass a physical and psychological evaluation of health, whereas long-standing illness reflects only physical health. It is likely that an evaluation of health with a psychological component would be associated with mental health bias. Our findings suggest that young people with poor general health may be a vulnerable group at risk of developing psychological ill-health. Measures of overweight and obesity showed little relationship with psychological health over a 2-year period. Previous studies have also failed to demonstrate an association between obesity and later depression, but have found an association between depression and later obesity [33,35,36]. We found no association between levels of physical activity at baseline and psychological health at follow-up, unlike a recent cross-sectional study that found an association in males [33]. The lack of associations observed between physical health and psychological health could reflect the relatively short follow-up period of the study. Pathways of risk between obesity and physical activity, and psychological health may occur later in adolescence or take longer to develop than the 2-year time frame of this study.

Strengths and limitations Earlier studies have been limited in the range of physical health and health-risk behaviours examined. This study of a representative ethnically diverse sample of East London adolescents has used validated questions to investigate the risk associated with a wide range of physical health and health-risk behaviours for adolescent psychological health. This study has a high participation rate at baseline and follow-up. The follow-up period of the study was limited to 2 years, which may affect the strength of associations observed for some predictors and limit the conclusions that can be drawn about pathways of causality between general health, health-risk behaviours and psychological health. In addition, further risk factors such as peer and sibling influences on substance availability and sexual activity were not assessed at baseline. The analyses are based on self-report and are also limited to participants with complete data for all the outcomes. We may have

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135

134

Charlotte Clark et al.

lost psychologically distressed participants from the sample, but this seems less likely over a 2-year period; in addition, psychologically distressed participants may be more likely to have been absent from school during data collection. While we have controlled for previous psychological health at baseline, the life-course history of psychological health for these participants is unknown. CONCLUSIONS Building on previous findings that indicate that individual substance use increases the risk for young people’s psychological health [20,21], the results of this study suggest that prevention strategies that target young people engaging in co-occurring substance use may also be a beneficial strategy for reducing the level of psychological disorders in young people. In addition, this study illustrates, with a community sample, that the most vulnerable young people are those with previous psychological ill-health, indicating that prevention strategies should not focus solely on health-risk behaviours. Acknowledgements We are grateful for the support of the schools, parents and students involved in this study, as well as the Community Advisory Board. We also thank the field team, including Wendy Isenwater, Giash Ahmed, Sarah Brentnall, Sultana Choudry-Dormer, Franca Davenport, Davina Woodley-Jones, Amanda Lawrence, Rachel Cameron and Hannah Bennett. Charlotte Clark is supported by an Academic Fellowship from the EPSRC; Muna Arephin is supported by an ESRC award ‘Development and persistence of human capability and resilience in the social and geographical context’, reference number L326253061. Phase 1 of the RELACHS study was commissioned by the East London and the City Health Authority and Phase 2 by the Teenage Pregnancy Unit at the Department of Health: we thank them for their support. We also thank Tower Hamlets, City and Hackney and Newham Primary Care Trusts for additional funding. The study has been carried out independently from the funders. References 1. Insel T. R., Fenton W. S. Psychiatric epidemiology. It’s not just about counting anymore. Arch Gen Psychiatry 2005; 62: 590–2. 2. Pine D. S., Cohen P., Gurley D., Brook J., Ma Y. The risk for early adulthood anxiety and depressive disorders in adolescents with anxiety and depressive disorders. Arch Gen Psychiatry 1998; 55: 56–64. 3. Lewinsohn P. M., Rohde P., Klein D. N., Seeley J. R. Natural course of adolescent major depressive disorder I. Continuity into young adulthood. J Am Acad Child Adolesc Psychiatry 1999; 38: 56–63.

4. Fergusson D. M., Woodward L. J. Mental health, educational and social role outcomes of adolescents with depression. Arch Gen Psychiatry 2002; 59: 225–31. 5. Kim-Cohen J., Caspi A., Moffitt T. E., Harrington H., Milne B. J., Poulton R. Prior juvenile diagnoses in adults with mental disorder. Arch Gen Psychiatry 2003; 60: 709–17. 6. Collishaw S., Maughan B., Goodman R., Pickles A. Time trends in adolescent mental health. J Child Psychol Psychiatry 2004; 45: 1350–62. 7. Bonomo U., Proimos J. Substance misuse: alcohol, tobacco, inhalants, and other drugs. BMJ 2005; 330: 777–80. 8. Fergusson D. M., Horwood L. J. Early onset cannabis use and psychosocial adjustment in young adults. Addiction 1997; 92: 279–96. 9. Breslau N., Peterson E. L., Schultz L. R., Chilcoat H. D., Andreski P. Major depression and stages of smoking. Arch Gen Psychiatry 1998; 55: 161–6. 10. Hanna E. Z., Grant B. F. Parallels to early onset alcohol use in the relationship of early onset smoking with drug use and DSM-W drug and depressive disorders: findings from the National Longitudinal Epidemiologic Survey. Alcohol Clin Exp Res 1999; 23: 513–22. 11. Breslau N., Novak S. P., Kessler R. C. Daily smoking and the subsequent onset of psychiatric disorders. Psychol Med 2004; 34: 323–33. 12. Brook J. S., Schuster E., Zhang C. Cigarette smoking and depressive symptoms: a longitudinal study of adolescents and young adults. Psychol Reports 2004; 95: 159–66. 13. Kandel D. B., Davies M., Karus D., Yamaguchi K. The consequences in young adulthood of adolescent drug involvement: an overview. Arch Gen Psychiatry 1986; 26: 117– 41. 14. Patton G. C., Hibbert M., Rosier M. J., Carlin J. B., Caust J., Bowes G. Is smoking associated with depression and anxiety in teenagers? Am J Public Health 1996; 86: 225–30. 15. Patton G. C., Carlin J. B., Coffey C., Wolfe R., Hibbert M., Bowes G. Depression, anxiety and smoking initiation: a prospective study over 3 years. Am J Public Health 1998; 88: 1518–22. 16. Milberger S., Biederman J., Faraone S. V., Chen L., Jones J. ADHD is associated with early initiation of cigarette smoking in children and adolescents. J Am Acad Child Adolesc Psychiatry 1997; 36: 37–44. 17. Escobedo L. G., Reddy M., Giovino G. A. The relationship between depressive symptoms and cigarette smoking in US adolescents. Addiction 1998; 93: 433–40. 18. Wang M. Q., Fitzhugh E. C., Green B. L., Turner L. W., Eddy J. M., Westerfield R. C. Prospective social-psychological factors of adolescent smoking progression. J Adolesc Health 1999; 24: 2–9. 19. Ismail K., Sloggett A., De Stavola B. Do common mental disorders increase cigarette smoking? Results from five waves of a population-based panel cohort study. Am J Epidemiol 2000; 152: 651–7. 20. McGee R., Williams S., Poulton R., Moffitt T. A longitudinal study of cannabis use and mental health from adolescence to early adulthood. Addiction 2000; 95: 491–503. 21. Patton G. C., Coffey C., Carlin J. B., Degenhardt L., Lynskey M., Hall W. Cannabis use and mental health in young people: cohort study. BMJ 2002; 325: 1195–8. 22. Wells J. E., Horwood L. J., Fergusson D. M. Drinking patterns in mid-adolescence and psychosocial outcomes in late adolescence and early adulthood. Addiction 2004; 99: 1529– 41.

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135

Health behaviour predictors of psychological symptoms

23. Jackson K. M., Sher K. J., Cooper L., Wood P. Adolescent alcohol and tobacco use: onset, persistence and trajectories of use across two samples. Addiction 2002; 97: 517–31. 24. Viner R. M., Haines M. M., Head J. A., Bhui K., Taylor S. J. C., Stansfeld S. A. et al. Variations in associations of health risk behaviors among ethnic minority early adolescents. J Adolesc Health 2006; 38: 55. 25. Wagner F. A., Anthony J. C. Into the world of illegal drug use. exposure opportunity and other mechanisms linking the use of alcohol, tobacco, marijuana and cocaine. Am J Public Health 2002; 155: 918–25. 26. Lynskey M. T., Fergusson D. M., Horwood L. J. The origins of the correlations between tobacco, alcohol and cannabis use during adolescence. J Child Psychol Psychiatry 1998; 39: 995–1005. 27. Morral A. R., McCaffrey D. F., Paddock S. M. Reassessing the marijuana gateway effect. Addiction 2003; 97: 1493–504. 28. Fergusson D. M., Boden J. M., Horwood J. Cannabis use and other illicit drug use: testing the cannabis gateway hypothesis. Addiction 2006; 101: 556–69. 29. Steptoe A., Butler N. Sports participation and emotional wellbeing in adolescents. Lancet 1996; 347: 1789–92. 30. Sawyer M. G., Arney F. M., Baghurst P. A., Clark J. J., Graetz B. W., Kosky B. et al. The mental health of young people in Australia: key findings from the child and adolescent component of the national survey of mental health and well-being. Aust NZ J Psychiatry 2001; 35: 806– 14. 31. Kirkcaldy B. D., Shephard R. J., Siefen R. G. The relationship between physical activity and self-image and problem behaviour among adolescents. Soc Psychiatry Psychiatr Epidemiol 2002; 37: 544–50. 32. Tomson L. M., Pangrazi R. P., Friedman G., Hutchison N. Childhood depressive symptoms, physical activity and health related fitness. J Sport Exer Psychol 2003; 25: 419– 39. 33. Fulkerson J. A., Sherwood N. E., Perry C. L., NeumarkSztainer D., Story M. Depressive symptoms and adolescent eating and health behaviours: a multifaceted view in a population-based sample. Prev Med 2004; 38: 865–75. 34. Motl R. W., Birnbaum A. S., Kubik M. Y., Dishman R. K. Naturally occurring changes in physical activity are inversely related to depressive symptoms during early adolescence. Psychosom Med 2004; 66: 336–42. 35. Goodman. E., Whitaker. R. C. A prospective study on the role of depression on the development and persistence of adolescent obesity. Pediatrics 2002; 110: 497–504. 36. Richardson L. P., Davis R., Poulton R., McCauley E., Moffitt T. E., Caspi A. et al. A longitudinal evaluation of adolescent depression and adult obesity. Arch Pediatr Adolesc Med 2003; 157: 739–45. 37. Mustillo S., Worthman C., Erkanli A., Keeler G., Angold A., Costello E. J. Obesity and psychiatric disorder: developmental trajectories. Pediatrics 2003; 111: 851–9.

135

38. Goodman R. The strengths and difficulties questionnaire: a research note. J Child Psychol Psychiatry 1997; 38: 581–6. 39. Goodman R., Meltzer H., Bailey V. The strengths and difficulties questionnaire: a pilot study on the validity of the self-report version. Eur Child Adolesc Psychiatry 1998; 7: 125–30. 40. Stansfeld S. A., Haines M. M., Head J. A., Bhui K., Viner R., Taylor S. J. C. et al. Ethnicity, social deprivation and psychological distress in adolescents. Br J Psychiatry 2004; 185: 233–8. 41. Angold A., Costello E. J., Messer S. C., Pickles A., Winder F., Silver D. Development of a short questionnaire for use in epidemiological studies of depression in children and adolescents. Int J Meth Psychiatr Res 1995; 5: 237–49. 42. Booth M. L., Okely A. D., Chey T., Bauman A. The reliability and validity of the physical activity questions in the WHO health behavior in school children (HBSC) survey: a population study. Br J Sports Med 2001; 35: 263–7. 43. Goddard E., Higgins V. Smoking, Drinking and Drug Use Among Young Teenagers in 1998, vols 1 and 2. London: The Stationery Office; 1999. 44. Taylor S. J. C., Viner R., Booy R., Head J., Tate H., Brentnall S. L. et al. Ethnicity, socio-economic status, overweight and underweight in East London adolescents. Ethn Health 2005; 10: 113–28. 45. Cole T. J., Freeman J. V., Preece M. A. Body mass index reference curves for the UK, 1990. Arch Dis Child 1995; 73: 25–9. 46. Sammons P., West A., Hind A. Accounting for variation in pupil attainment at the end of Key Stage 1. Br Educ Res J 1997; 23: 489–511. 47. Poulin C., Hand D., Boudreau B., Santor D. Gender differences in the association between substance use and elevated depressive symptoms in a general adolescent population. Addiction 2005; 100: 525–35. 48. Needle R., McCubbin H., Wilson M., Reineck R., Lazar A., Mederer H. Interpersonal influences in adolescent drug use—the role of older siblings, parents and peers. Int J Addict 1986; 21: 739–66. 49. Fergusson D. M., Swain-Campbell N. R., Horwood L. J. Deviant peer affiliations, crime and substance use: a fixed effects regression analysis. J Abnorm Child Psychol 2002; 30: 419–30. 50. von Sydow K., Lieb R., Pfister H., Höfler M., Wittchen H. What predicts incident use of cannabis and progression to abuse and dependence? A 4-year prospective examination of risk factors in a community sample of adolescents and young adults. Drug Alcohol Depend 2002; 68: 49–64. 51. Patton G. C., McMorris B. J., Toumbourou J. W., Hemphill S. A., Donath S., Catalano R. F. Puberty and the onset of substance use and abuse. Pediatrics 2004; 114: 300–6. 52. van den Bree M. B., Pickworth W. B. Risk factors predicting changes in marijuana involvement in teenagers. Arch Gen Psychiatry 2005; 62: 311–9.

© 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction

Addiction, 102, 126–135