Scholarly article on topic 'The Association Between Social Support and Mental Health Among Vulnerable Adolescents in Five Cities: Findings From the Study of the Well-Being of Adolescents in Vulnerable Environments'

The Association Between Social Support and Mental Health Among Vulnerable Adolescents in Five Cities: Findings From the Study of the Well-Being of Adolescents in Vulnerable Environments Academic research paper on "Psychology"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Adolescent Health
OECD Field of science
Keywords
{Adolescent / "Mental health" / "Social support" / "Suicidal thoughts"}

Abstract of research paper on Psychology, author of scientific article — Yan Cheng, XianChen Li, Chaohua Lou, Freya L. Sonenstein, Amanda Kalamar, et al.

Abstract Purpose Globally, adolescents are at risk of depression, traumatic stress, and suicide, especially those living in vulnerable environments. This article examines the mental health of 15- to 19-year-old youth in five cities and identifies the social support correlates of mental health. Methods A total of 2,393 adolescents aged 15–19 years in economically distressed neighborhoods in Baltimore, MD; New Delhi, India; Ibadan, Nigeria; Johannesburg, South Africa; and Shanghai, China were recruited in 2013 via respondent-driven sampling to participate in a survey using an audio computer-assisted self-interview. Weighted logistic regression and general linear models were used to explore the associations between mental health and social supports. Results The highest levels of depression and posttraumatic stress symptoms were displayed in Johannesburg among females (44.6% and 67.0%, respectively), whereas the lowest were among New Delhi females and males (13.0% and 16.3%, respectively). The prevalence of suicidal ideation ranged from 7.9% (New Delhi female adolescents) to 39.6% (Johannesburg female adolescents); the 12-month prevalence of suicide attempts ranged from 1.8% (New Delhi females) to 18.3% (Ibadan males). Elevated perceptions of having a caring female adult in the home and feeling connected to their neighborhoods were positively associated with adolescents' levels of hope across the sites while negatively associated with depression and posttraumatic stress symptoms with some variation across sites and gender. Conclusions Adolescents living in the very economically distressed areas studied register high levels of depression and posttraumatic stress. Improving social supports in families and neighborhoods may alleviate distress and foster hope. In particular, strengthening supports from female caretakers to their adolescents at home may improve the outlooks of their daughters.

Academic research paper on topic "The Association Between Social Support and Mental Health Among Vulnerable Adolescents in Five Cities: Findings From the Study of the Well-Being of Adolescents in Vulnerable Environments"

ELSEVIER

JOURNAL OF

ADOLESCENT HEALTH

www.jahonline.org

Original article

The Association Between Social Support and Mental Health Among -i.crossMark Vulnerable Adolescents in Five Cities: Findings From the Study of the Well-Being of Adolescents in Vulnerable Environments

Yan Cheng, Ph.D.a, XianChen Li, B.Med.b, Chaohua Lou, M.D.a*, Freya L. Sonenstein, Ph.D.c, Amanda Kalamarc, Shireen Jejeebhoy, Ph.D.d, Sinead Delany-Moretlwe, M.D., Ph.D.e, Heena Brahmbhatt, Ph.D.c, Adesola Oluwafunmilola Olumide, M.B.B.S., M.P.H.f, and Oladosu Ojengbede, B.Sc., M.B.B.S. g

a Department of Epidemiology and Social Science Research on Reproductive Health, Shanghai Institute of Planned Parenthood Research, Shanghai, People's Republic of China b School of Public Health, Fudan University, Shanghai, People's Republic of China

c Department of Population, Family and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland d Population Council, New Delhi, India

e Wits Reproductive Health & HIV Institute, School of Clinical Medicine, University of the Witwatersrand, Johannesburg, South Africa f Institute of Child Health, College of Medicine, University of Ibadan/University College Hospital, Ibadan, Oyo State, Nigeria g Department of Obstetrics and Gynaecology, College of Medicine, University of Ibadan/ University College Hospital, Ibadan, Oyo State, Nigeria

Article history: Received May 14, 2014; Accepted August 20, 2014 Keywords: Adolescent; Mental health; Social support; Suicidal thoughts

ABSTRACT

Purpose: Globally, adolescents are at risk of depression, traumatic stress, and suicide, especially those living in vulnerable environments. This article examines the mental health of 15- to 19-year-old youth in five cities and identifies the social support correlates of mental health. Methods: A total of 2,393 adolescents aged 15—19 years in economically distressed neighborhoods in Baltimore, MD; New Delhi, India; Ibadan, Nigeria; Johannesburg, South Africa; and Shanghai, China were recruited in 2013 via respondent-driven sampling to participate in a survey using an audio computer-assisted self-interview. Weighted logistic regression and general linear models were used to explore the associations between mental health and social supports. Results: The highest levels of depression and posttraumatic stress symptoms were displayed in Johannesburg among females (44.6% and 67.0%, respectively), whereas the lowest were among New Delhi females and males (13.0% and 16.3%, respectively). The prevalence of suicidal ideation ranged from 7.9% (New Delhi female adolescents) to 39.6% (Johannesburg female adolescents); the 12-month prevalence of suicide attempts ranged from 1.8% (New Delhi females) to 18.3% (Ibadan males). Elevated perceptions of having a caring female adult in the home and feeling connected to their neighborhoods were positively associated with adolescents' levels of hope across the sites while negatively associated with depression and posttraumatic stress symptoms with some variation across sites and gender. Conclusions: Adolescents living in the very economically distressed areas studied register high levels of depression and posttraumatic stress. Improving social supports in families and neighborhoods may alleviate distress and foster hope. In particular, strengthening supports from female caretakers to their adolescents at home may improve the outlooks of their daughters.

© 2014 Society for Adolescent Health and Medicine. All rights reserved.

IMPLICATIONS AND CONTRIBUTION

Mental health problems are prevalent among adolescents in vulnerable environments and are significantly associated with family support and neighborhood connection. Given the current limitations of both research and mental health care capacities in resource-poor settings, increasing social support, especially female adult caregivers, may prove effective in addressing the mental health of adolescents.

Conflict of Interest: The authors declare no conflicts of interest. Disclaimer: Publication of this article was supported by the Young Health Programme, a partnership between AstraZeneca, Johns Hopkins Bloomberg School of Public Health, and Plan International. The opinions or views expressed in this article are those of the author and do not necessarily represent the official position of the funders.

1054-139X/© 2014 Society for Adolescent Health and Medicine. All rights reserved. http://dx.doi.org/10.1016/jjadohealth.2014.08.020

* Address correspondence to: Chaohua Lou, M.D., Department of Epidemiology and Social Science Research on Reproductive Health, Shanghai Institute of Planned Parenthood Research, Shanghai, 200237, Laohumin Road 779, Shanghai, People's Republic of China.

E-mail address: Chaohual@yahoo.com (C. Lou).

Mental health is defined as "a state of well-being in which every individual realizes his or her own potential, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to her or his community" [1]. It is noted that about 20% of children and adolescents in the world are estimated to have mental disorders or problems such as schizophrenia, depression, mental retardation, and disorders due to drug abuse, and about half of all mental disorders begin before the age of 14 years [2]. Mental disorders account for a large proportion of the disease burden in young people in all societies, especially in economically disadvantaged areas, where resources for care are scarce and mental health is strongly associated with social determinants of health [3]. People with mental disorders experience disproportionately higher rates of disease burden and mortality [3], and depression is ranked as the leading cause of disability worldwide. Apart from disability, mental disorders also exert a substantial burden on mortality in young people in many communities. Adolescence is a period of heightened risk of suicide [4]. Suicide is a leading cause of death in young people in countries such as China [5] and India [6], and the second most common cause of death among young people worldwide [7,8]. It is well known that mental disorders in adolescents are strongly related to mental disorders in adulthood [9,10]. Thus, it is important to study the risk and protective factors for mental health problems and any impact cultural and geographical variations have on mental health outcomes among adolescents to develop effective prevention strategies.

Mental health, similar to other aspects of health, can be influenced by a range of factors. It is widely recognized that mental health problems are the result of interactions of biological with psychosocial factors [11]. Previous studies have identified risk factors for psychiatric disorders including genetic, cognitive, temperamental, interpersonal, and family environment factors [12,13]. In addition, a strong relationship exists between poor mental health and many other concerns for young people, notably educational achievement, substance use and abuse, violence, and poor reproductive and sexual health [3]. Moreover, poor relationships with peers, feelings of lack of security, and negative life events may also increase the risk of mental disorders in children and adolescents [13-15]. Longitudinal studies have also shown that factors such as a sense of connection, low levels of conflict, and an environment in which the expression of emotions is encouraged protect against the development of behavioral or emotional disorders [3,16].

Social support may be an important psychosocial buffer in the face of other risk factors and may play an even more significant role in the lives of adolescents, which are often characterized by multiple developmental and physical changes [3,16]. Social support is a form of social capital that individuals can draw upon to help them cope with daily stressors [17]. Researchers have consistently documented positive associations between social support and psychological well-being among adult and youth [18,19] and inverse associations between social support and depression [19,20]. In particular, research findings suggest that social support may help protect youth against the negative effects of stressors and promote more positive mental health outcomes [18,19].

So far, much of the work focusing on the positive effects of social support on psychological health has emphasized the role of perceived support with a specific source [21], but there is still limited research about how social support from multiple sources (i.e., parents, peers, and community) differentially predict adolescents' mental health outcomes. Such an understanding is

needed, in part, so that researchers and practitioners can make more informed decisions regarding where to focus prevention and intervention efforts. Furthermore, little attention has been afforded to how social support relates to positive indicators of mental health, such as hope for the future. Such research might elucidate ways to foster optimism among youth. To address these gaps, our study aims to (1) estimate and compare vulnerable youth in different countries with respect to their mental health and social supports and (2) to examine the relationship between social support from multiple important sources and mental health in this population.

Methods

Sample and procedures

Data for these analyses come from a cross-sectional survey of 2,393 males and females aged 15-19 years, conducted in five cities around the world in 2013 in the second phase of the Well-Being of Adolescents in Vulnerable Environments (WAVE) study. Among 2,393 respondents, 476 were from Baltimore, 500 from New Delhi, 465 from Ibadan, 497 from Johannesburg, and 455 were migrants from Shanghai. Only 2,339 eligible cases were included in this analysis because of the exclusion of cases with missing data. For more details on the sampling methodology, see Decker et al. in this volume.

Respondent-driven sampling (RDS) was used to ensure the sample included out-of-school youth and unstably housed youth who are often underrepresented in school-based or household-based samples. The survey was developed by the research team, translated, back-translated, and piloted in each site. Back translations involve having someone other than the original translators translate the culturally adapted measure back into the scale's original language to see if translated items still measured the intended trait. However, using adult translators and backtranslators without an awareness and understanding of how youths will interpret the translated wording is problematic. Therefore, before the formal survey, a pilot with 50 adolescents was conducted in each site, and the measurements were further culturally adapted based on the feedbacks of the pilot. To increase the comparability of data across sites, interviewers received extensive training to follow a standard survey procedure and establish rapport with respondents, and a safe and comfortable environment was provided to conduct the survey to erase the anxiety of respondents. All interviews were conducted through a survey using audio computer-assisted self-interview instruments. All aspects of this study received approval from the Committee on Human Research at the Johns Hopkins University and review committees in the collaborating local organizations.

Measures

Sociodemographic characteristics include age, sex, current school enrollment, perceived relative wealth (same as others, better than others, or worse than others in the neighborhood), family situation in which the adolescent was raised (two parents—including one or both adoptive or step parents; one parent; or other relatives or nonrelatives), and unstably housed (not having a regular place to stay or staying an average of 3-4 nights a week or less in the regular place over the last 30 days and staying overnight in more than one place in the last 7 days). These measures were used as covariates in the multivariate analyses.

The extent to which respondents reported social support from family, peers, and neighborhood was measured using four scales. All scale scores were derived by summing the items in each scale (item range, 0-3) and were treated as continuous measures. The caring male adult in home measure consisted of a scale composed of four items that asked the extent to which a male adult in family expected the respondent to follow rules, be successful, and would help and listen to the respondent. Scale scores ranged from 0 to 12, with a values ranging from .78 (in Ibadan) to .92 (in Johannesburg). A similar scale of four items constitutes the caring female adult in home measure. Alpha values ranged from .82 (in Ibadan) to .92 (in Johannesburg). For both scales, the higher the score, the higher was the perception of support from a caring adult. Peer support asked about certain aspects of friendship such as trust, being able to lend money, providing help when needed, being able to talk about personal problems and sex, and unconditional acceptance. The range of scores is from 0 to 18 and a values ranged from .78 (in Shanghai) to .88 (in Baltimore). The higher the score, the greater the sense of peer support. Finally, the connection to the neighborhood measure identified whether the respondents perceived support on a community level and consisted of three items asking how much the respondents believed that they were connected to people in the neighborhood, if they knew most of the people in the neighborhood, and whether they felt like they were a part of this community. Scale scores ranged from 0 to 9, with a values ranging from .74 (in New Delhi) to .82 (in Shanghai) across sites. As the score increased, so did the perception of being connected to the neighborhood.

Mental health measures were identified through self-report scales assessing the respondents' perception of hope for the future, depression and posttraumatic stress, and suicidal ideation (Table 1). All scale scores were derived by summing items in each scale. The hope measure consisted of three items that asked respondents about the extent to which they believed "good things will happen" and "the future will turn out well" and they "felt excited about future." Each item ranged from 0 to 2 and the range of scale scores is from 0 to 6, and the higher the score, the more positive the hopes for the future [22]. Alpha values for this measure ranged from .70 (in Johannesburg) to .89 (in Baltimore). The depression symptoms measure was assessed by the short 10-item version of the Center for Epidemiological Studies Depression Scale, which is a screening measure developed to identify current depressive symptomatology related to major or clinical depression in adults and adolescents. The range of scores is 0-30, and higher scores indicate more symptoms; a score of 11 or higher indicates mild depressive symptomatology for the short version of the scale [23]. Alpha values for this measure ranged from .74 (in New Delhi) to .87 (in Shanghai). The posttraumatic stress measure, consisted of a scale made up of six items from the Post-Traumatic Stress Disorder (PTSD) Checklist-Civilian version (PCL-C, 17 items) [24,25], a brief screening instrument for detecting PTSD symptomatology and severity among men and women in primary care; its reliability and validity have been established [26,27]. In this study, six items based on analyses of data collected in an earlier study of Baltimore youth were used. The items contributing to the highest Cronbach alpha in the earlier study were selected for inclusion in the WAVE study. Alpha values for this measure ranged from .80 (in New Delhi) to .95 (in Baltimore). The range of scores is 6-30 with each item ranging from 1 to 5 and a score of 13 or higher indicating PTSD symptoms. This cut point was determined based on the Baltimore data (combining the whole PCL-C and adapted six-item version)

Table 1

Score sheet for mental health measures

Mental health measure Code

Depression symptoms

a. You were bothered by things that usually do not bother you in the past 7 days

b. You had trouble keeping your mind on what you were doing in the past 7 days

c. You felt depressed in the past 7 days

d. You felt that everything you did was an effort in the past 7 days

e. You felt hopeful about the future in the past 7 days

f. You felt fearful in the past 7 days

g. Your sleep was restless in the past 7 days

h. You were happy in the past 7 days

i. You felt lonely in the past 7 days

j. You could not get "going" in the past 7 days

Posttraumatic stress

a. In the past 30 days, how much did you suddenly act or feel like a stressful experience was happening again (like you were reliving it)?

b. In the past 30 days, how much did you feel very upset when something reminded you of a stressful experience from the past?

c. In the past 30 days, how much did you lose interest in things that you used to enjoy?

d. In the past 30 days, how much did you feel emotionally numb or unable to have loving feelings for people close to you?

e. In the past 30 days, how much did you feel like your future would be cut short somehow?

f. In the past 30 days, how much did you have difficulty concentrating?

a. I expect good things will happen to me.

b. I am excited about my future.

c. I trust my future will turn out well.

Suicide ideation

a. Have you ever seriously thought about attempting suicide?

b. During the past 12 months, did you make a plan about how you would attempt suicide?

c. During the past 12 months, did you ever attempt suicide?

using receiver operating characteristic curves in STATA; a cutoff score of 13 was calculated through Youden's index according to the standard cutoff score of 30 for the PCL-C, with the sensitivity of .82, the specificity of .95, and the area under the curve of .9469. A consistency test was then performed between PCL-C and adapted PTSD, and the Kappa coefficient was .80 [24,28,29]. The suicidal thoughts, plans, and attempts measures were binary variables based on questions, "Have you ever seriously thought about attempting suicide?" "During the past 12 months, did you make a plan about how you would attempt suicide?" and "During the past 12 months, did you ever attempt suicide?"

Data analysis

The sample was weighted to accommodate the RDS design before the analysis. Weights were generated via the RDS II estimator to account for the intercluster correlation and were used in the multivariate analysis. Given differences in the age distribution across sites, a poststratification age weight was also developed and combined with the RDS weight. The age-adjusted RDS

0 = Rarely or none of the time (less than1 day)

1 = Some or a little of the time (1-2 days)

2 = Occasionally or a moderate amount of the time (3-4 days)

3 = Most of or all the time (5-7 days)

1 = Not at all

2 = A little bit

3 = Moderately

4 = Quite a bit

5 = Extremely

0 = Not at all like me

1 = Somewhat like me

2 = Exactly like me

0 = No

1 = Yes

weights were used in the univariate analysis to compare the difference across sites, but the RDS weight without the age adjustment was used in the multivariate analyses because these analyses were stratified by gender in each site Weighted means and standard errors and unweighted means were used to describe the continuous variables Weighted percentages, unweighted percentages, and sample sizes were used to describe the categorical variables . Chi-square tests were used to detect differences in the distribution of demographic characteristics, prevalence of depression symptoms, and posttraumatic stress, and the information regarding suicide across sites by gender Analysis of variance tests were used to detect differences in social support scores and hope scores across sites Multivariate analyses (logistic regression model and general linear regression model) were used to examine the associations between social support and hope, depression symptoms, posttraumatic stress, and suicidal thoughts after controlling for covariates Data were analyzed with STATA 12 .0 statistical software using complex design procedures to accommodate the nonindependence of observations related to intercluster correlation within recruitment chains

Results

Sample characteristics across sites

Table 2 presents the sociodemographic characteristics of the respondents across study sites, stratified by sex A demographic description of the samples across the sites was presented in Marshall et al in this volume Significant differences between the sites on these measures, which have been shown to be associated with mental health outcomes from past literature, suggested that they should be included as controls in multivariate analyses

Mental health and social support across the sites

Regarding the hope scale, both male and female respondents from Ibadan scored the highest across the five sites, with a mean of 5 . 2 for males and 5 . 6 for females, approaching the maximum of 6 . Male and female respondents from Shanghai reported the lowest levels of hope, with mean of 3.4 for males and 3. 5 for females .

The male and female respondents from Johannesburg reported the highest percentages of depressive symptoms with 41 1% of males and 44 6% of females scoring above the cut point (>11) . The percentage with elevated depression symptoms was the lowest among Shanghai male and New Delhi female participants, with 17.1% and 13.0% scoring elevated symptoms, respectively. Regarding posttraumatic stress, Johannesburg adolescents also showed the highest percentage with elevated symptoms, 54.5% among males and 67.0% among females. New Delhi respondents had the lowest percentage of elevated post-traumatic stress symptoms among males and Ibadan had the lowest among females, but the levels still reached 16.3% and 24.8%, respectively. The differences in depression and post-traumatic stress measures across sites are statistically significant (p < .01). Gender differences were also found in four sites except for Ibadan. Shanghai female respondents had a significantly higher percentage of depression symptoms compared with males, and the percentage of posttraumatic stress symptoms was significantly higher among Baltimore, New Delhi, and Johannesburg females compared with males (p < .05).

Among both males and females, adolescents from Johannesburg reported significantly higher percentages of suicidal

E i« P.

E if P.

E if P.

v-' CH ™ g

CD cri ^ o

m to cn ^

CJ1 CD >

m to > P

K cn ™ g rn to ^ O

in t- f": n

rn^S £

OO r- >

^ Ln > D

™ 00 ^ <N > _

m ^ > D

oo <n H o

rn to ^ O

to m <o ^

<N oo ^

r" > —

m ^ > D

to rn 1 ^ in ;

OO (N^

^ in w ■ in ^ to

r^ oo ^

(N > _

m oo (N csi oo (NIO ^

t< t< ^ mLnm

^ m 00 Ln

, , cn

(N (N r*

ioio^ r-: 00 °° mmt^

m t^ 00 ^ (N <0

t ^ 10 cn-^ , romio

CO ^ ^

S oq £

tOCntN

tO r* C

rn ^ r-:

(N en in 00 fi 00

00 m t^ ¡vi r^ ^ m(N(N

n en 't csi ^ r<"i ■imM

•sf (.0

•sf (.0 to (vi to m

(N ^ r^ 00

thoughts and suicide plans in the past 12 months compared with other sites, with 32.6% and 19.9% for males, and 39.6% and 25% for females, respectively. When asked about whether a respondent attempted suicide in the past 12 months, the top percentages are displayed among Ibadan respondents, with 18.3% among males and 14.3% among females saying that they had. The lowest percentages of suicide thoughts and plans emerged among New Delhi male and female respondents, respectively (13.9% and 4.1% for males, 7.9% and 2.7% for females), respectively. As for suicide attempts, the lowest percentages are from Shanghai male respondents and New Delhi female adolescents, with 2.8% and 1.8%, respectively. The differences across sites are statistical significant (p < .001). The differences in suicidal thoughts between sexes only appeared in Johannesburg and Shanghai; females reported higher percentages of suicidal ideation (p < .05).

The social support scores from family, peers, and neighborhood display similar patterns among male and female respondents across the sites. Further description of these measures across sites is included in Marshall et al. in this volume (Table 3).

The associations between respondents' perceived social support and hope score

Table 4 displays the associations between respondents' perception of social support and their hope score by gender after controlling for covariates. The bivariate analyses are not presented because of space limitations.

The most important positive predictor of hope among males was perceived connection to the neighborhood, significant in all

sites (b = .09-.17) except Baltimore. Among female adolescents, social support from a caring female adult showed a strong positive association with hope across all sites (b=.07-.13), the exception being New Delhi. However, among male adolescents, support from a caring female adult was not significantly associated with hope.

The associations between respondents' perceived social support and depression symptoms

Table 5 summarizes the associations between perceived social support from caring adults in the home, peers, connection to the neighborhood, and depressive symptoms. Among male respondents, the only factors that reduced the odds of depression were having caring male adults in the home in Johannesburg (adjusted odds ratio [AOR] = .88), with peer support in New Delhi (AOR = .89), and with perceived connection to neighborhood in Ibadan and Shanghai (AOR = .88 and .72, respectively). The social support variables were not significantly associated with depression among Baltimore male respondents. Among female respondents, having a caring female adult in the home was correlated with a significant decrease in depression in Baltimore, New Delhi, and Johannesburg (AOR = .85-.89, p < .05). Similarly, having a caring male adult in the home was also correlated with decreased report of depression among females in Baltimore and Shanghai (AOR = .85 and .74, respectively). Also, perceived connection to the neighborhood among females in Johannesburg and Shanghai was associated with a lower odds of depression (AOR=.92 and .75, respectively).

Table 3

Respondents perceived social support scores and mental health status across sites (adjusted by age-respondent-driven sampling combined weight)

Variable Baltimore, New Delhi, Ibadan, Johannesburg, Shanghai,

W mean W mean W mean (SE), W mean (SE), W mean (SE),

(SE), U mean (SE), U mean U mean U mean U mean

Social support scores

Caring male adult in home (0-12)*** 7.5 (.07), 8.0AC 9.0 (.13), 9.2B 7.8 (.44), 8.0a 7.6 (.37), 7.7a,c 6.7 (.41), 6.6C

Caring female adult in home (0-12)*** 10.1 (.34), 10.0A 8.5 (.08), 8.8B 8.1 (.39), 8.5B 8.5 (.31), 8.8B 7.6 (.55), 7.4B

Peer support (0-18)*** 14.7 (.32), 14.8a 13.6 (.26), 13.8B 13.8 (.13), 13.7ab 13.1 (.56), 13.6B 15.2 (.06), 15.8a

Perceived connection to neighborhood (0-9)*** 5.6 (.07), 5.7a 7.4 (.10), 7.4B 7.1 (.14), 7.1B 6.3 (.24), 6.3C 5.4 (.16), 5.6a

Mental health

Hope score (0-6)*** 4.9 (.06), 4.9AB 4.8 (.03), 4.9a 5.2 (.13), 5.4B 4.9 (.04), 5.0a b 3.4 (.11), 3.4C

Depression symptoms-percent scoring >11, 26.4 (27.0,71) 18.7(17.2, 43) 29.4 (27.3, 60) 41.1 (40.1, 109) 17.1 (17.1, 38)

W% (U%, N)***

Posttraumatic stress scoring >13, W% (U%, N)** 30.6 (32.3, 85) 16.3 (15.6, 39) 33.6 (32.7, 72) 54.5 (55.5, 151) 35.6 (31.5, 70)

Suicidal thoughts, W% (U%, N)** 16.7 (18.0, 45) 13.9(13.2, 33) 23.3 (19.7, 43) 32.6 (28.4, 77) 19.7 (13.1, 29)

Suicidal plans, W% (U%, N)*** 6.4 (8.8, 23) 4.1 (3.6, 9) 17.9(15.5, 34) 19.9 (14.7, 40) 6.7 (3.6, 8)

Suicidal attempts, W% (U%, N)*** 5.0 (8.4, 22) 4.2 (3.2,8) 18.3 (14.1, 3) 11.8 (9.6, 26) 2.8 (1.4, 3)

Females

Social support scores

Caring male adult in home (0-12)* 7.1 (.18), 6.8ABa 7.3 (.21), 7.7A,a 6.7 (.26), 6.7A,a 6.8 (.26), 6.8Aa 6.6 (.45), 6.2B

Caring female adult in home (0-12)*** 9.8 (.16), 10.1A 8.6 (.23), 8.9BC 8.6 (.21), 8.6B,C 9.2 (.33), 9.4C a 7.7 (.52), 7.7B

Peer support (0-18)*** 14.9 (.23), 15.6Aa 13.6 (.14), 13.6B 13.2 (.10), 13.2B 15.1 (.16), 14.7Aa 14.9 (.14), 15.3a

Perceived connection to neighborhood (0-9)*** 5.0 (.17), 5.5a 7.0 (.11), 7.0B,a 6.9 (.12), 7.0B 6.0 (.33), 5.7C,a 4.7 (.05), 4.9Aa

Mental health

Hope score (0-6)*** 5.2 (.01), 5.3Aa 4.7 (.08), 4.6B 5.6 (.07), 5.6a 5.5 (.05), 5.4A,a 3.5 (.12), 3.7C

Depression symptoms-percent scoring >11, 31.0 (36.3, 70) 13.0 (14.8, 37) 28.5 (29.3, 67) 44.6 (37.9, 85) 24.9 (26.4, 57)a

W% (U%, N)***

Posttraumatic stress-percent scoring >13 37.5 (45.1, 87)a 35.1 (38.0, 95)a 24.8 (26.6, 61) 67.0 (67.4, 151)a 39.2 (38.4, 83)

W% (U%, N)***

Suicidal thoughts, W% (U%, N)*** 28.6 (23.8, 45) 7.9 (8.8, 22) 19.1 (20.9, 47) 39.6 (38.4, 86)a 38.7 (29.3, 63)a

Suicidal plans W% (U%, N)*** 14.7 (13.5, 26) 2.7 (2.4, 6) 14.3 (14.9, 34) 25.0 (21.9, 49) 11.3 (9.7, 21)

Suicidal attempts, W% (U%, N)*** 7.3 (7.3, 14) 1.8 (2.0, 5) 14.3 (14.4, 33) 10.0 (13.0,29) 5.9 (5.1, 11)

*p < .05, **p < .01, ***p < .001. Means marked with the same A, B, or C superscripts are not significantly different across five sites. a p < .05. Means and percentages are significantly different between females and males on social support score and mental health outcomes of each site.

Table 4

The association between social support and hope score across sites (linear regression model adjusted by respondent-driven sampling weight, b [95% CI])

Social support Baltimore New Delhi Ibadan Johannesburg Shanghai

N 236 219 185 245 193

Caring male adult in home .05 (-.02, .11) .19* (.15, .22) .06 (-.02, .13) .05* (.01, .10) .04 (-.13, .21)

Caring female adult in home .01 (-.15, .17) -.03 (-.14, .17) .02 (-.07, .11) .10 (.04, .17) .04 (-.09, .18)

Peer support .09 (.00, .19) .01 (-.08, .11) .03 (-.04, .10) -.01 (-.06, .03) .03 (-.09, .16)

Perceived connection to neighborhood -.03 (-.07, .02) .16* (.12, .20) .09* (.01, .17) .11* (.07, .15) .17* (.06, .28)

Females

N 185 216 182 217 190

Caring male adult in home .10* (.05, .15) .08 (-.01, .16) .01 (-.03, .06) .05* (.02, .07) -.02 (-.17, .13)

Caring female adult in home .08* (.04, .13) .10 (-.02, .21) .08* (.06, .10) .07* (.04, .10) .13* (.09, .18)

Peer support .00 (-.02, .02) .06 (-.02, .14) .04 (.00, .08) .01 (.00, .03) .10* (.03, .17)

Perceived connection to neighborhood .00 (-.05, .05) -.06 (-.21, .10) -.03 (-.05, .00) .05* (.02, .08) .05 (-.05, .15)

*p < .05, adjusted by age, current school enrollment, perceived relative wealth, raised by, and unstably housed. New Delhi site only adjusted age, current school enrollment, perceived relative wealth, and raised by.

The associations between respondents' perceived social support and posttraumatic stress symptoms

After controlling for covariates, among male respondents, having a caring male adult in the home and perceiving a connection to the neighborhood showed a strong negative association with posttraumatic stress symptoms only in Shanghai (AOR = .86 and .80, respectively). Among female respondents, the odds of posttraumatic stress were significantly reduced by having a caring female adult in the home in Baltimore (AOR = .85) and perceived connection to the neighborhood in Johannesburg (AOR = .84; Table 5).

The associations between respondents' perceived social support and suicide thoughts

Table 5 also summarizes the associations between respondents' perceived social support and suicide thoughts by gender after controlling for covariates. The results from the analyses of the associations between social support and planning or attempting suicide are not presented because of few significant outcomes. Suicidal ideation in female respondents' was much more strongly associated with not having caring male and female adults, peer support, and connection to neighborhood than that in males in Baltimore, New Delhi, Johannesburg, and Shanghai. More support from caring male adults was associated with lower levels of suicidal thoughts among New Delhi and Shanghai female respondents (AOR = .90 and .87, respectively). Also, higher scores for caring adult females in the home were significantly associated with a lower odds of suicidal ideas among Baltimore and Johannesburg female respondents and Shanghai male respondents (AOR = .79-.90). More peer support was significantly associated with lower levels of suicidal thoughts among Baltimore and Ibadan male respondents (AOR = .88). Higher perceived neighborhood connection was also associated with lower odds of suicide thoughts among Johannesburg female adolescents and Shanghai female and male respondents (AOR = .79-.90). There were no statistically significant associations between perceiving support from male or female adults in the home, peer support, and connection to neighborhood among Ibadan female respondents.

Discussion

The findings indicate that, except for Shanghai, adolescents from the other four cities reported high levels of hope for the

future (scoring close to the highest level). There were differences in the proportions of adolescents who reported depressive symptoms in the five cities; the report of depressive symptoms was highest among adolescents in Johannesburg, followed by adolescents in Ibadan and Baltimore, and the lowest proportions were in Shanghai and New Delhi. However, even in settings where the reports of depressive symptoms were lower, 13% and 18% of the Indian female and male participants did report some depressive symptoms. These findings are similar to previous findings from the Youth in India study, where 14% of males and females aged 15-24 years reported three or more symptoms on the General Health Questionnaire—12 [30]. When assessing different types of mental health outcomes, reports of post-traumatic stress symptoms (PTSD) were more common compared with depression. This was especially true of males and females in Johannesburg where a majority reported symptoms of PTSD. Suicide ideation was also quite prevalent in our population with almost one-tenth of females in New Delhi to almost 40% of females in Johannesburg reporting thoughts of suicide. The 12-month prevalence of suicidal attempts was as high as almost 20% among males in Ibadan. Female adolescents appeared to bear a greater burden of mental distress than males, which is consistent with previous literature [31,32] and may be explained by the weaker social connections and stronger social expectations for the role of young women. Moreover, our results demonstrate the importance of support from caring family members and connection to neighborhood, which were shown to be associated with less depression, posttraumatic stress symptoms, suicidal ideation, and, conversely, more hope in some sites. The variations across sites on mental health and its association with social supports might be explained by various sociocultural factors, which merit further research.

The severity of depression and posttraumatic stress symptoms found across the sites underscores how vulnerable these adolescents in very economically distressed city neighborhoods are. A study in a French high school for example, reported 19% and 34% of boys and girls, respectively, had moderate-to-severe depressive symptomatology in 2009 [31]. In an Australian high school, the percentages were 22% for girls and 16% for boys. In the United States, 15% of adolescent girls and 7.5% of adolescent boys were reported to have had a major depressive disorder [33,34]. The percentages reported in the Wave study are substantially higher, especially in Johannesburg. Similarly, the prevalence of posttraumatic stress symptoms is higher in the WAVE study than that previous studies have found. The rates in Johannesburg

^ O £ O ™

„ • oo • t^

oo to o oo o oo oo

_ iN * tO

o q « §8

(N t—

CXT3 ™ £ ^

(N o cn oo oo oo r^

cn o to

Ln oo m

(N o o

o o t^ oo

, m o <N

, , CO ,

№ (N O M

oo oo oo

t^ O o fN O

o (N oo q co o (N m in q o

LO oq (N o* oq cn oq ^f cn q oq oo

q oo oq oo oq f q i>. m (N O O o q m q oo

r" r" r- r-

q if oo q CO if r^ (N q N

oq cn oq oo o* 00 •sf O* 00 00 cn 00 in 00

f q o m q oo 00 o oo cn o q O oo q q to oo

, , o ,

® o o

oo oo oo

<u ^ > '

t o ■

CD ^ & T3 ,

e & a?

(U ffl

^ T3 ^ <D 3 >

M (D ffl

j= e s

1 ^ MM

(54% for males and 67% for females) were much higher than past reports of 38% among South African adolescents in 2005 [35] and 4%-10% in other studies among adolescents. [32,36].

Across sites, Shanghai respondents received the lowest scores for caring family members and neighborhood connection, which might be related to their migration status. In this study, all Shanghai respondents were migrants. Being an outsider may have made it more challenging for these young migrants to be fully integrated into their new communities. Besides, instable housing, loss of supportive networks, and absence from family members due to employment may have resulted in young people giving a lower score to family and neighborhood connections. These findings are supported by the qualitative findings in the phase 1 of the WAVE study [37]. It is possible that difficulties in receiving support from family and neighborhood may have made young migrants turn to peers; the perceived peer support scores were almost the highest among Shanghai male and female respondents across the sites.

Caring female adults in the home emerged as a consistent protective factor for mental health, especially among Baltimore and Johannesburg female adolescents. Although many females in these two cities are in single-parent households or raised by other relatives or nonrelatives, these results indicate that adolescents will benefit from having a caring female adult in their life. Caring male adults in the home also displayed an important role for Shanghai female adolescents, which underscores the saliency of the supportive family context in facilitating mental wellness. In general, the present study's findings regarding significant links between family support and adolescent mental health are consistent with the body of research that suggests supportive relationships co-occur with optimal mental health [38,39]. Results of the present study indicate that connection to the neighborhood is also implicated in alleviating adolescents' depression and stress syndrome and even suicidal thoughts in Johannesburg female adolescents and Shanghai male and female adolescents, above and beyond the substantial influence of perceived support from family adults. Regarding positive indicators of mental health, social support from family and neighborhood resources are positively associated with adolescents' hope score universally, suggesting that the presence of social support is positive in an adolescent's life and an important resilience factor from which to build. These results are similar to earlier findings, which indicated that higher perceptions of support from parents co-occur with higher ratings of global life satisfaction [40]. Moreover, the data suggest that family adult and community support were much more significantly associated with the mental health of female adolescents compared with male adolescents.

The findings of this study are subject to limitations. First, this study is cross sectional, so any attempt to make a causal link between social support and mental health outcomes should be avoided. Second, the data gathered for this study were from adolescents aged 15-19 years living in poor urban neighborhoods, only migrants were recruited in Shanghai site, and the data were collected using RDS methods. These sample characteristics limit the degree to which we can make any inferences for adolescents in general. Also important, these analyses do not examine a significant source of social support identified in the literature, support from caring adults in school, as over 60% of respondents from Shanghai were not students and did not respond to this part of questionnaire. A major limitation is that our measures of mental health, with the exception of depression and suicidality, have not been used in cross-cultural studies, They

were developed in the United States, and in spite of our efforts to accurately translate their concepts, they may not be applicable in other settings. Our shortened version of the PTSD scale could be particularly problematic and probably assesses generalized distress rather than clinical trauma. Nonetheless our measures appear reliable and differentiate the adolescents.

Implications

Despite these limitations, our findings suggest that mental health problems are prevalent among adolescents in the vulnerable environments and are significantly associated with family support and neighborhood connection. Given the prevalence of poor mental health shown by our imperfect measures, more attention needs to be paid to developing and testing instruments that are cross culturally appropriate. In addition, there is little evidence that the mental health of young people is a priority in many resource-poor settings that are already challenged with competing health and social burdens. However, the mental health of young people should not be neglected, especially in cities where young people have experienced a lot of stress and trauma because of loss of family members from infectious diseases, such as AIDS, or higher levels of violence in the community environment as in Johannesburg and Baltimore. Increasing social support from neighborhood and family, especially female adult caregivers, may prove effective in addressing the mental health of adolescents, and interventions to do this should receive increased attention from practitioners and researchers.

Funding Sources

This research was supported by Young Health Programme, a partnership between AstraZeneca, Johns Hopkins Bloomberg School of Public Health, and Plan International, a leading global children's charity. In Ibadan, the study was funded by The Bill and Melinda Gates Institute at Johns Hopkins Bloomberg School of Public Health through its funding to The Centre for Population and Reproductive Health, University of Ibadan.

References

[1] World Health Organization. Mental health: A state of well-being. Available at: http://www.who.int/features/factfiles/mental_health/en. Accessed March 14, 2014.

[2] World Health Organization. 10 Facts on mental health. Available at: http:// www.who.int/features/factfiles/mental_health/mental_health_facts/en. Accessed March 14, 2014.

[3] Patel V, Flisher AJ, Hetrick S, et al. Mental health of young people: A global public-health challenge. Lancet 2007;369:1302-13.

[4] Vijayakumar L, John S, Pirkis J, Whiteford H. Suicide in developing countries (2): Risk factors. Crisis 2005;26:112-9.

[5] Phillips MR, Li X, Zhang Y. Suicide rates in China, 1995-99. Lancet 2002; 359:835-40.

[6] Aaron R, Joseph A, Abraham S, et al. Suicides in young people in rural southern India. Lancet 2004;363:1117-8.

[7] Patton GC, Coffey C, Sawyer SM, et al. Global patterns of mortality in young people: A systematic analysis of population health data. Lancet 2009;374: 881-92.

[8] Hawton K, Saunders KE, O'Connor RC. Self-harm and suicide in adolescents. Lancet 2012;379:2373-82.

[9] Himle JA, Baser RE, Taylor RJ, et al. Anxiety disorders among African Americans, blacks of Caribbean descent, and non-Hispanic whites in the United States. J Anxiety Disord 2009;23:578-90.

[10] Williams DR, Haile R, González HM, et al. The mental health of Black Caribbean immigrants: Results from the National Survey of American Life. Am J Public Health 2007;97:52-9.

Huang JP, Xia W, Sun CH, et al. Psychological distress and its correlates in Chinese adolescents. Aust N Z J Psychiatry 2009;43:674-80. quiz 681.

Greszta E. Family environment risk factors of depression in adolescence. Psychiatr Pol 2006;40:719-30.

Purper-Ouakil D, Michel G, Mouren-Simeoni MC. Vulnerability to depression in children and adolescents: Updates and perspectives. Encephale 2002;28(3 Pt 1):234-40.

Adewuya AO, Ologun YA, Ibigbami OS. Post-traumatic stress disorder after childbirth in Nigerian women: Prevalence and risk factors. BJOG 2006;113: 284-8.

Abela JR, Hankin BL, Haigh EA, et al. Interpersonal vulnerability to depression in high-risk children: The role of insecure attachment and reassurance seeking. J Clin Child Adolesc Psychol 2005;34:182-92. Joiner TE, ed. Mood disorders: Diagnostic & therapeutic issues. Tallahassee, FL: Center Circle Press; 2002.

Dominguez S, Watkins C. Creating networks for survival and mobility: Social capital among African-American and Latin-American low-income mothers. Social Probl 2003;50:111-35.

Hussong AM. Perceived peer context and adolescent adjustment. J Res Adolesc 2000;10:391-415.

Newman BM, Newman PR, Griffen S, et al. The relationship of social support to depressive symptoms during the transition to high school. Adolescence 2007;42:441-59.

Peirce RS, Frone MR, Russell M, et al. A longitudinal model of social contact, social support, depression, and alcohol use. Health Psychol 2000;19:28.

Bierman A, Aneshensel CS, eds. Handbook of the Sociology of Mental Health. New York: Kluwer Academic/Plenum Publishers; 1999. Lippman LH, Anderson Moore K, Guzman L, et al. Flourishing children: Defining and testing indicators of positive development. Defining and testing indicators of positive development. Dordrecht: Springer Netherlands, Dordrecht; 2014.

Radloff LS. The use of the Center for Epidemiological Studies of Depression Scale in adolescents and young adults. J Youth Adolesc 1991; 20:149-66.

Freedy JR, Steenkamp MM, Magruder KM, et al. Post-traumatic stress disorder screening test performance in civilian primary care. Fam Pract 2010; 27:615-24.

Weathers FW, Litz BT, Huska JA, Keane TM. PTSD Checklist—Civilian version. Boston: National Center for PTSD, Behavioral Science Division; 1994.

Blanchard EB, Jones-Alexander J, Buckley TC, et al. Psychometric properties of the PTSD Checklist (PCL). Behav Res Ther 1996;34:669-73. Ruggiero KJ, Del Ben K, Scotti JR, et al. Psychometric properties of the PTSD checklist civilian version. J Trauma Stress 2003;16:495-502. VA National Center for PTSD. Using the PTSD Checklist (PCL). Available at http://www.ptsd.va.gov/professional/pages/assessments/assessment-pdf/ pcl-handout.pdf. Accessed March 10, 2014. Youden WJ. Index for rating diagnostic tests. Cancer 1950;3:32-5. Youth in India: Situation and needs 2006-2007. Mumbai: International Institute for Population Sciences (HPS) and Population Council; 2010. Chabrol H, Choquet M. Relationship between depressive symptoms, hopelessness and suicidal ideation among 1547 high school students. Encephale 2009;35:443-7.

Kessler RC, Sonnega A, Bromet E, et al. Posttraumatic stress disorder in the National Comorbidity Survey. Arch Gen Psychiatry 1995;52:1048. Sawyer MG, Pfeiffer S, Spence SH. Life events, coping and depressive symptoms among young adolescents: A one-year prospective study. J Affect Disord 2009;117:48-54.

National Institute of Mental Health. Major depressive disorder in children. Available at http://www.nimh.nih.gov/statistics/1mdd_child.shtml. Accessed March 30, 2014.

Suliman S, Kaminer D, Seedat S, et al. Assessing post-traumatic stress disorder in South African adolescents: Using the child and adolescent trauma survey (CATS) as a screening tool. Ann Gen Psychiatry 2005;4:2. National Institute of Mental Health. Post-Traumatic Stress Disorder Among Children Available at http://www.nimh.nih.gov/statistics/1AD_PTSD_ CHlLD.shtml. Accessed March 30, 2014.

Mmari K, Blum R, Sonenstein F, et al. Adolescents' perceptions of health from disadvantaged urban communities: Findings from the WAVE study. Social Sci Med 2013;104:124-32.

Suldo SM, Huebner ES. ls extremely high life satisfaction during adolescence advantageous? Social Indicators Res 2006;78:179-203. Wight RG, Botticello AL, Aneshensel CS. Socioeconomic context, social support, and adolescent mental health: A multilevel investigation. J Youth Adolesc 2006;35:109-20.

Danielsen AG, Samdal O, Hetland J, et al. School-related social support and students' perceived life satisfaction. J Educ Res 2009;102:303-20.