Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Child Abuse Negl. 2020 Dec 29;112:104904. doi: 10.1016/j.chiabu.2020.104904

The interrelationships of child maltreatment, alcohol use, and suicidal ideation among youth living in the slums of Kampala, Uganda

Rachel Culbreth 1,*, Katherine E Masyn 2, Monica H Swahn 2, Shannon Self-Brown 3, Rogers Kasirye 4
PMCID: PMC7855690  NIHMSID: NIHMS1658976  PMID: 33385928

Abstract

Background:

Youth living in the slums of Kampala face many adversities, such as dire environmental conditions, poverty, and lack of government infrastructure.

Objective:

The purpose of this study is to examine the interplay of alcohol use and child maltreatment on suicidal ideation among youth living in the slums of Kampala, Uganda.

Participants and Setting:

The study sample includes service-seeking youth who were attending Uganda Youth Development Link (UYDEL) drop-in centers in spring 2014 (n=1,134).

Methods:

Indicators of child maltreatment included parental physical abuse, parental neglect, and sexual abuse. Problematic alcohol use was specified using a hybrid structural equation mixture model that distinguished current drinking status with the frequency and intensity of use among current drinkers. This novel approach is more flexible than restricting our analysis to only drinkers or analyzing only current drinking status. The primary outcome of interest was suicidal ideation. All associations controlled for gender and age, and all associations were estimated simultaneously. All analyses were conducted in SAS 9.4 and Mplus 7.4.

Results:

The overall prevalence of suicidal ideation was 23.5% (n=266). Overall, current drinking status (OR: 1.80; 95% CI: 1.31, 2.46), the child maltreatment sum score (OR: 1.88; 95% CI: 1.48, 2.39), and sexual abuse (OR: 2.88; 95% CI: 1.52, 5.47) were statistically significantly associated with reporting suicidal ideation.

Conclusions:

This study highlights a population that would potentially benefit from prevention efforts not only aimed at suicide prevention but also harm reduction in terms of alcohol use and experiences of child maltreatment.

Keywords: alcohol use, child maltreatment, sexual abuse, suicidal ideation, problem drinking

Introduction

Globally, suicide is the third leading cause of death for adolescents ages 15-19 (World Health organization, 2016). Additionally, 79% of suicides occur in low- and middle-income countries (World Health organization, 2016). In sub-Saharan Africa, adolescent suicide is starting to emerge as an important public health problem, but studies examining suicidality among adolescents are limited (Page & West, 2011). Reports of suicidal ideation among adolescents vary across countries in sub-Saharan Africa. An estimated 13% of youth report suicidal ideation among school-attending youth in Malawi (Shaikh, Lloyd, Acquah, Celedonia, & Wilson, 2016). Additionally, the prevalence of suicidal ideation among adolescents in Uganda and Kenya is estimated to be 20% and 28%, respectively (Swahn, Bossarte, Eliman, Gaylor, & Jayaraman, 2010). Youth living in very economically distressed areas may be at a higher risk of suicide (Cheng et al., 2014). Youth living in the slums of Kampala have reported higher rates of suicidal ideation (Culbreth, Swahn, Ndetei, Ametewee, & Kasirye, 2018; Swahn, Palmier, Kasirye, & Yao, 2012) compared to population-based studies examining suicidal behaviors in Uganda (Swahn et al., 2010). Additionally, youth living in the slums or streets in Kampala live in a disadvantaged environment, often characterized by extreme poverty and lack of government infrastructure, which may contribute to the high rates of suicidal ideation among these youth (Mufune, 2000; Swahn, Palmier, et al., 2012; Swahn, Gressard, et al., 2012; Swahn, Dill, Palmier, & Kasirye, 2015; Swahn, Haberlen, & Palmier, 2014).

Theoretical frameworks for suicidal ideation

Predictors for suicidal ideation include substance use (Jones, 1997; King & Merchant, 2008; Sher, Sperling, Zalsman, Vardi, & Merrick, 2006; Sher & Zalsman, 2005; Page & West, 2011; Reifman & Windle, 1995; Schilling, Aseltine, Glanovsky, James, & Jacobs, 2009; Swahn, Palmier, et al., 2012), child maltreatment (Brown, Cohen, Johnson, & Smailes, 1999; Cluver, Orkin, Boyes, & Sherr, 2015; King & Merchant, 2008; Ng et al., 2015), depression and mental illness (Cluver et al., 2015), and negative future expectations (Abramson et al., 1998; Ballard, Patel, Ward, & Lamis, 2015). Several models and theoretical frameworks help explain the associations between these risk factors and suicidal ideation. The Problem Behavior Theory (PBT) states that youth who engage in substance use, such as alcohol, are at an increased risk for the development of depression, which in turn increases risk of suicidal ideation and suicidal behaviors (Jessor & Jessor, 1977). Additionally, the Secondary Mental Disorder Model states that victimization, including child maltreatment victimization, may lead to alcohol use, which in turn may lead to suicidal ideations (Pompili et al., 2010; Marschall-Lévesque et al., 2016). Several longitudinal studies have demonstrated that alcohol use in adolescence is associated with higher suicidal ideation in early adulthood (Borowsky, Ireland, & Resnick, 2001; Duncan, Alpert, Duncan, & Hops, 1997; Fergusson, Woodward, & Horwood, 2000; Reifman & Windle, 1995). This is also consistent with the Stress-Coping Theory, which states that individuals engage in substance use and alcohol use to cope with previous stressful events in life, such as child maltreatment experiences, which then exacerbates risk for suicidal ideation (Kandel, Raveis, & Davies, 1991). However, several studies have reported conflicting directionality results where suicidal ideations and behaviors predict alcohol use and substance use later in life (Fergusson et al., 2000; Steinhausen, Bösiger, & Metzke, 2006).

Other theories have emphasized the importance of negative cognitions and the association with suicidal ideation. The hopeless theory of suicide states that hopeless cognitions and negative future outlooks are directly related to suicidal ideation, specifically when prefaced with adverse events (Abramson et al., 1998; Ballard, Patel, Ward, & Lamis, 2015). Negative future expectations may increase suicidal ideations later in life through the perceptions that negative events are unavoidable, therefore lowering resilience to suicidal thoughts and behaviors (Jamieson & Romer, 2008; Nguyen et al., 2012). Additionally, perceptions of negative future expectations may lead to substance use as a coping mechanism (Jamieson & Romer, 2008; Nguyen et al., 2012). A conceptual model which informs this study is presented in Figure 1.

Figure 1.

Figure 1.

Open in a new tab

Conceptual model of suicidal ideation predictors

Child maltreatment and suicidal ideation

Child maltreatment is a strong predictor for suicidal ideation. Brown and colleagues found that adults who reported child maltreatment were three times more likely to also report suicidal behaviors (1999). A meta-analysis reported there is robust evidence for the association between physical abuse, emotional abuse, and childhood neglect with depressive disorders and suicide attempts (Norman et al., 2012).

In addition to the links between child maltreatment and suicidal ideation, studies have found an association with child sexual abuse (Smith, Smith, & Grekin, 2014; Meyers et al., 2018), emotional abuse (Mills, Alati, Strathearn, & Najman, 2014; Shin, Edwards, & Heeren, 2009; Shin, Miller, & Teicher, 2013), physical abuse, and neglect (Norman et al., 2012) with problematic alcohol use among adolescents. Experiencing multiple types of child maltreatment was associated with a faster progression to heavy episodic drinking, which persisted across young adulthood (Shin et al., 2013). Additionally, overuse of alcohol and binge drinking are known to cause disinhibition, impaired judgment and impulsivity, and these are the mechanisms which may link alcohol use to suicidal behavior (Pompili et al., 2010; Wilsnack, Wilsnack, Kristjanson, Vogeltanz-Holm, & Windle, 2004).

Suicidal ideation in sub-Saharan Africa

Page and West conducted a review which examined suicidal behaviors and ideation among adolescents in sub-Saharan Africa and reported that 25% of boys and 26% of girls reported suicidal ideation in the past 12 months (2011). Among a sample of adolescents living in southwest Nigeria, suicidal behaviors were statistically significantly associated with childhood sexual abuse (Omigbodun, Dogra, Esan, & Adedokun, 2008). In Uganda, child maltreatment was statistically significantly associated with suicidal behaviors among adolescents in Northern Uganda (Olema, Catani, Ertl, Saile, & Neuner, 2014), while child neglect was associated with suicidal ideation among youth living in the slums of Kampala, Uganda (Swahn, Palmier, et al., 2012). Understanding the mechanisms of suicidal ideation predictors among youth living in the slums of Kampala, Uganda is urgently warranted. These youth may face unique risk factors, and known risk factors may operate differently. For example, this population may have a stronger association between child maltreatment, alcohol use, negative future expectations, and suicidal ideation compared to other populations. These associations may be exacerbated by the dire environmental living conditions these youth face, including poverty, food scarcity, exposure to violence, and a lack of government infrastructure (Culbreth et al., 2018; Swahn, Culbreth, Salazar, Kasirye, & Seeley, 2016; Swahn et al., 2014; Swahn et al., 2015; Swahn, Culbreth, Staton, Self-Brown, & Kasirye, 2017). Additionally, this population has a high prevalence of commercial sex work (13%), which has been previously linked with alcohol use (Swahn et al., 2016) and poor mental health outcomes (Hong, Li, Fang, & Zhao, 2007).

Suicidal ideation and alcohol use among males and females in sub-Saharan Africa

While the prevalence estimates of suicidal ideation are similar for both males (25%) and females (26%) in sub-Saharan Africa (Page & West, 2011), alcohol use differs by gender among youth, which may be an important driver for suicidal ideation in this population. Among a convenience sample of service-seeking youth living in the slums of Kampala, Uganda, estimates of past-year alcohol use, or drinking status, were similar among boys (31.2%) and girls (30.0%) in Kampala, Uganda (Swahn et al., 2020). However, when stratified by age groups, there were statistically significant differences between males and females. Specifically, a higher percentage of males consumed alcohol 4 or more times per week than females; however, a higher percentage of females consumed alcohol 2-3 times a week compared to males (Swahn et al., 2020). Males also reported a younger age of first alcohol consumption compared to females (Swahn et al., 2020).

Analytic framework for examining alcohol use and suicidal ideation

Previous research has established the strong links between alcohol use and suicidal ideation, including problematic and heavy alcohol use with suicidal ideation (Pompili et al., 2010; Wilsnack, Wilsnack, Kristjanson, Vogeltanz-Holm, & Windle, 2004). However, no study to our knowledge has examined both drinking status and problematic alcohol use with suicidal ideation simultaneously. Specifically, researchers have examined drinking status and problematic alcohol use separately (Culbreth et al., 2018; Swahn et al., 2012, 2020). Studies which have examined only drinking status do not account for problematic alcohol use, and studies which examine problematic alcohol use either listwise delete individuals who do not consume alcohol or assign these individuals a “0” for problematic alcohol use behavior. This is a major limitation for alcohol use research by not being able to identify whether drinking status overall predicts an outcome or whether problematic alcohol use specifically predicts an outcome, above and beyond drinking status. When analyzing drinking status and problematic alcohol use separately, the implications for prevention strategies are limited to either drinking status or amount as opposed to a comprehensive approach to examining both drinking status and problematic alcohol use. One solution for this issue is a hybrid structural equation mixture model, which incorporates alcohol use as a latent class variable and problematic alcohol use as a factor model that is only specified in the “current drinker” class of the drinking status latent class variable. This approach would allow effect estimates for both drinking status (the latent class variable) and problematic alcohol use, while not excluding or deleting non-drinkers.

Study objective

While several studies have examined suicidal attempts and ideation among youth living in the slums of Kampala (Culbreth et al., 2018; Swahn, Palmier, et al., 2012), the current study seeks to examine suicidal ideation among this sample of youth in a larger, latent variable framework. No study, to our knowledge, has examined the mechanisms of suicidal ideation predictors among adolescents in Uganda, and more broadly, sub-Saharan Africa. Additionally, this study seeks to apply a hybrid structural equation mixture model to estimate the effects of both drinking status and problem drinking on suicidal ideation, a novel approach to estimating the effects of alcohol use. Using the conceptual model, which is comprised of three theoretical frameworks (Problem Behavior Theory, Secondary Mental Disorder Model, and Hopeless Theory of Suicide), we aim to understand the impact of child maltreatment, drinking status, and negative future expectations simultaneously on suicidal ideation. Additionally, since this study is cross-sectional, we plan to examine the effects of drinking status on suicidal ideation, rather than examining bidirectional effects of suicidal ideation on drinking status. Moreover, this study seeks to determine the specific associations between child maltreatment, problem drinking, and negative future expectations on alcohol use among current drinkers, in addition to the impact of drinking status on suicidal ideation. Therefore, our hypotheses are: 1) drinking status will have a positive, direct effect on suicidal ideation; 2) problematic alcohol use will have a positive, direct effect on suicidal ideation, above and beyond drinking status, 3) negative future expectations will have a positive, direct effect on suicidal ideation, and 4) child maltreatment will have a positive, direct effect on drinking status, problematic alcohol use, negative future expectations, and suicidal ideation. By utilizing a hybrid structural equation mixture model, we will be able to test these hypotheses that incorporate both drinking status and problematic alcohol use. We also plan to adjust for age and gender in these analyses. Understanding the heterogeneity of suicidal ideation predictors among adolescents is critical in creating culturally relevant and effective suicidal interventions (Kinyanda, Wamala, Musisi, & Hjelmeland, 2011).

Methods

Study Design and Participants

The current analysis is based on data collected in Kampala, Uganda, as part of a study known as the “Kampala Youth Survey 2014.” This was a cross-sectional study conducted in 2014 on youth ages 12-18 years of age who live in the streets and the slums of Kampala. The youth comprised a convenience sample who were attending the Uganda Youth Development Link (UYDEL) drop-in centers, which provide may services to youth, including vocational training, HIV/STI testing, and mental health counselling services. The participation rate among youth who were approached to participate was 92%, yielding 1,497 youth. Due to technical issues with the server being offline, 320 surveys were lost, which resulted in 1,134 surveys for the final sample. A total of 1,130 surveys were included in this analytic sample (n=4 had missing data on our outcome suicidal ideation).

The survey was administered face-to-face by social workers and peer educators who were trained in the study methodology and survey administration. Survey questions were translated into the local language, Luganda, if necessary and back translated for improved accuracy. All participants provided verbal informed consent to participate in the study. Youth under 18 who “cater to their own livelihood” are considered independent and emancipated in Uganda, enabling them to provide their own informed consent without parental consent. Youth participants were limited to ages 12-18 on the day of the study, and no other exclusion criteria was applied. IRB approvals were obtained from both sites (Georgia State University and the Uganda National Council for Science and Technology).

Measures

Survey questions for the Kampala Youth Survey 2014 were adapted from measures of youth alcohol use, experiences of violence victimization and perpetration, alcohol marketing exposures and mental health among adolescents. While the survey questions were developed from previously validated measures, these measures were not validated among Ugandan adolescents. Further details on contents of the survey are discussed elsewhere (Swahn et al., 2016), and a detailed description of the measures used are listed in Appendix.

Suicidal ideation.

For the current analysis, suicidal ideation was the main outcome of interest. Youth were asked, “In the past year, did you ever think of killing yourself?” Response options were binary: (0) “No” or (1) “Yes.”

Child maltreatment.

Three questions used to measure child maltreatment (lifetime) included parental neglect, parental abuse, and sexual abuse. Parental neglect was attributed to parental alcohol use, and was measured using, “Did your parents/caretakers’ alcohol use make them not able to care for you?” Sexual abuse was measured using, “Has someone ever raped you or forced you to have sex with him or her?” Parental physical abuse was measured using, “Did your parents ever beat you so hard that you had bruises/marks?” Responses to all three questions were binary: (0) “No” or (1) “Yes.”

Negative future expectations.

Three questions measured negative future expectations. Participants were asked, “Overall, what do you think about the following statements? I will probably die before I am thirty; I will be unhappy; Bad things happen to people like me.” Responses were binary: (0) “No/disagree” or (1) “Yes/agree.”

Current drinking status.

Two questions of alcohol use were used to measure current drinking status, and all participants were asked these two questions. The first alcohol use question was, “How old were you when you had your first full drink of alcohol?” Respondents could answer 1-12, 13-14, 15-16, 17-18, and never. The second question was, “Have you had a drink of alcohol in the past year?” Responses were binary: (0) “No” or (1) “Yes.”

Problematic alcohol use.

Problematic alcohol use consists of four measures: frequency, amount and alcohol use adverse behavior. Youth who reported not drinking in the past year were missing on all of the problematic alcohol questions since a skip pattern was present in the survey. Alcohol use frequency was measured using, “How often do you have a drink containing alcohol?” The timeframe for this question was not specified. Responses consisted of “Monthly or Less”, “2-4 times a month”, “2-3 times a week”, and “4 or more times a week.” Alcohol use amount was measured using, “How many full drinks containing alcohol do you have in a typical day when you are drinking?” Responses consisted of “1-2 drinks,” “3-4 drinks,” and “5 or more drinks.” Alcohol use adverse behavior was measured using two questions, “Have you been seriously injured or hurt due to your drinking?” and “Has someone else been seriously injured or hurt because of our drinking?” Responses were binary for both questions: (0) “No” or (1) “Yes.”

Control variables.

Control variables included the analysis included gender (female/male) and age (in years).

Data Analysis

Descriptive statistics and bivariate associations were examined among the variables of interest. A hybrid structural equation mixture model was utilized to examine all variables. First, factor models for negative future expectations and problematic alcohol use were each constructed separately. Then, these two factor models were analyzed with the child maltreatment variables. Child maltreatment variables were tested using a series of nested model tests to determine the optimal operationalization of these variables.

Since we utilized a hybrid structural equation mixture model, we were able to analyze problematic alcohol use among current drinkers only. A latent class variable was used to comprise drinking status (current, non-active, and never drinkers). Then, problematic alcohol use was estimated in the class of current drinkers only. Current drinkers were operationalized by using two questions. If youth reported a specific age for initiating alcohol use and responded “Yes” to having a full drink of alcohol in the past year, they were classified as current drinkers. If youth reported a specific age for initiating alcohol use but responded “No” to having a full drink of alcohol in the past year, they were classified as non-active drinkers. Lastly, for youth who reported “Never” to initiating alcohol use and “No” to having a full drink of alcohol in the past year, they were classified as never drinkers. Then, the problematic alcohol use factor model (which included frequency, amount, and alcohol use adverse behaviors) was only estimated within the current drinking class. Therefore, this hybrid structural equation mixture model allowed us to examine our hypotheses in the context of both drinking status and problematic alcohol use. This approach is more flexible compared to only analyzing current drinking status or analyzing problematic alcohol use among current drinkers and listwise deleting non-drinkers. Additionally, this approach allows more flexibility in the modeling process of problematic drinking compared to typical practices of setting all problematic drinker indicators to zero for missing values. This method allows for the inclusion of all participants for the analysis of both current drinking status and problematic alcohol use. Our complete analytic model for the hybrid structural equation mixture model is presented in Figure 2.

Figure 2.

Figure 2.

Open in a new tab

Analytic model for the impact of child maltreatment, negative future expectations, alcohol use, and suicidal ideation among youth living in the slums of Kampala, Uganda

Measurement invariance was assessed for all latent factors. Full information maximum likelihood (FIML) estimation was used to estimate the model under the missing-at-random (MAR) assumption. Descriptive and bivariate analyses were conducted in SAS 9.4 (SAS Institute, Cary, NC), and the measurement and structural equation mixture models were estimated using MPlus 7.4 (Muthén, L. K., & Muthén, B. O. Los Angeles, CA: Muthén & Muthén).

Results

Descriptive statistics among reported suicidal ideation are displayed in Table 1. Among all youth participants (n=1,130), the prevalence of suicidal ideation is 23.5% (n=266). A higher percentage of females reported suicidal ideation compared to males (27% vs. 19%, respectively).

Table 1.

Demographics and Predictors of Suicidal Ideation among Youth Living in the Slums of Kampala

Suicidal Ideation
Total
n=1130		 Yes
n=266 (24%)		 No
n=864 (76%)
Demographic variables, n (%)
Age, mean (SD) 16.15 (1.79) 16.41 (1.69) 16.07 (1.81)
Gender
  Female 635 (56%) 172 (27%) 463 (73%)
  Male 494 (44%) 94 (19%) 400 (81%)
Child maltreatment experiences, n (%)
Physical abuse 380 (34%) 137 (36%) 243 (64%)
Sexual abuse 191 (17%) 77 (40%) 114 (60%)
Parental neglect 212 (20%) 89 (42%) 123 (58%)
Child maltreatment sum score
  0 595 (53%) 69 (12%) 526 (88%)
  1 349 (31%) 117 (34%) 232 (67%)
  2 151 (13%) 62 (41%) 89 (59%)
  3 35 (3%) 18 (51%) 17 (49%)
Alcohol use, n (%)
Age at first alcohol consumption
  Never 718 (64%) 120 (17%) 598 (83%)
  1-12 58 (5%) 25 (43%) 33 (57%)
  13-14 116 (10%) 42 (36%) 74 (64%)
  15-16 165 (15%) 57 (35%) 108 (65%)
  17-18 66 (6%) 19 (29%) 47 (71%)
Alcohol use in past year
 Yes 346 (31%) 129 (37%) 217 (63%)
 No 784 (69%) 137 (17%) 647 (83%)
Alcohol frequency
  Monthly or less 70 (20%) 26 (37%) 44 (63%)
  2-4 times a month 104 (30%) 35 (34%) 69 (66%)
  2-3 times a week 128 (37%) 53 (41%) 75 (59%)
  4 or more times a week 44 (13%) 15 (34%) 29 (66%)
Amount of alcohol consumed
 1-2 drinks 195 (57%) 65 (33%) 130 (67%)
 3-4 drinks 118 (24%) 49 (42%) 69 (59%)
 5 or more drinks 32 (9%) 15 (47%) 17 (63%)
Ever hurt yourself due to drinking 132 (38%) 64 (49%) 68 (52%)
Ever hurt someone else due to drinking 95 (28%) 45 (47%) 50 (53%)
Negative future expectations, n (%)
Unhappy about future 158 (14%) 74 (47%) 84 (53%)
Anticipating bad events 344 (30%) 126 (37%) 218 (63%)
Anticipating early death 146 (13%) 65 (45%) 80 (55%)
Open in a new tab

Among all youth in the sample, physical abuse was the highest reported type of abuse among youth (34% of total sample), and 36% of youth who experienced physical abuse also reported experiencing suicidal ideation. Among youth who experienced sexual abuse and parental neglect, a high percentage of youth reported suicidal ideation (40% and 42%, respectively). Additionally, higher child maltreatment sum scores corresponded to higher percentages of reported suicidal ideation. For example, approximately half (51%) of youth who reported experiencing all three types of child maltreatment experienced suicidal ideation.

Measurement model for problematic alcohol use and negative future expectations

The measurement models for problematic alcohol use and negative future expectations are presented in Table 2. The model for negative future expectations is just-identified, and the problematic alcohol use measurement model had adequate fit. A residual correlation was added between the two alcohol behavior items due to the high similarity between the two questions. All standardized loadings for both latent variables are above 0.60, except the two alcohol behavior indicators.

Table 2.

Measurement models for alcohol use and negative future expectations

Est.
loadings		 SE Standardized
est.		 R-
Square 		Thresholds
Alcohol use 1 2 3
Alcohol frequency 1.00 --- .63 .40 −1.90 −.01 2.59
Alcohol amount 1.59 .61 .79 .63 .51 3.93
Alcohol behavior (self) 1.04 .30 .47 .69 .98
Alcohol behavior (others) 1.28 .36 .55 .73 2.06
Negative future expectations
Unhappy 1.00 --- .77 .60 3.06
Bad events .81 .17 .70 .50 1.27
Open in a new tab

Note. Model fit statistics for alcohol use model: (χ2=58.06, df=35, p=0.009), Loglikelihood: −1102.347, RMSEA: 0.00, (90% CI: 0.00, 0.09), CFI: 1.00, TLI: 1.02.

Structural model

Structural associations are presented in Table 3. All structural associations were adjusted for gender and age, and measurement invariance held for all latent factors. After testing child maltreatment variable patterns using nested model tests, the model that incorporated a sexual abuse indicator, a child maltreatment sum score for physical abuse and neglect (0, 1, and 2), and an interaction term between sexual abuse and the sum score fit the data better than alternative models. Regarding the sum score, for youth who reported only sexual abuse, they received a “0” for maltreatment sum score (and a “1” for the sexual abuse variable). Participants who experienced either physical abuse alone or neglect alone each received a “1” for the maltreatment sum score, whereas participants who experienced both physical abuse and neglect received a “2” for the maltreatment sum score. Alternative models that were compared included only the child maltreatment sum score (all three types of child maltreatment) as well as a model with each unique child maltreatment experience type separately (physical abuse, sexual abuse, and neglect in the model as separate terms with all possible interactions). Table 4 presents the structural associations for the different patterns of child maltreatment.

Table 3.

Structural associations of child maltreatment, drinking status, problematic alcohol use, and negative future expectations on suicidal ideation among youth living in the slums of Kampala, Uganda

Negative future
expectations
(Difference in means)		 Problematic alcohol use
(Difference in means)		 Drinking status
(Conditional log odds ratios)		 Suicidal ideation
(Log odds ratios)
Current (vs. Never) Non-active (vs. Never)
Est.
(95% CI)		 P-
value		 Est.
(95% CI)		 P-
value		 Est.
(95% CI)		 P-
value		 Est.
(95% CI)		 P-value Est.
(95% CI)		 P-
value
Sex abuse → .53
(−.06, 1.13)		 .14 1.91
(1.09, 2.74) 		<.001 .84
(.41, 1.27) 		.001 1.11
(.37, 1.85) 		.01 1.06
(.42, 1.70) 		.01
Maltreatment sum→ .99
(.69, 1.29) 		<.001 .17
(−.13, .48)		 .34 .82
(.61, 1.03) 		<.001 .45
(.04, .86)		 .07 .63
(.39, .87) 		<.001
Sex abuse x Maltreatment sum→ −.22
(−.72, .27)		 .46 −.86
(−1.45, −.26) 		.02 .02
(−.43, .46)		 .95 −0.01
(−.81, .80)		 .99 −.72
(−1.20, −.24) 		.01
Negative future expectations→ -- -- .06
(−.07, .20)		 .43 .11
(.01, .20)		 .07 −.12
(−.32, .07)		 .30 .37
(.24, .50) 		<.001
Drinking status
Current drinker→ -- -- -- -- -- --- -- -- .59
(.27, .90) 		.002
Non-active drinker→ -- -- -- -- -- -- -- -- .24
(−.43, .91)		 .55
Problematic drinking -- -- -- -- -- -- -- -- .05
(−.19, .30)		 .72
Open in a new tab

Note. All statistically significant associations are bolded. All structural associations adjusted for gender and age.

Maltreatment sum score= sum score includes parental neglect and parental physical abuse (Min: 0, Max: 2); EST=estimate; SE=standard error. Model fit statistics: Loglikelihood: −3624.743. Number of parameters: 53.

Table 4.

Structural associations between patterns of child maltreatment and negative future expectations, problematic alcohol use, drinking status, and suicidal ideation among youth living in the slums of Kampala, Uganda

Negative future
expectations		 Problematic
alcohol use		 Drinking status Suicidal ideation
Current (vs. Never) Non-active (vs. Never)
Means
(95% CI)		 Means
(95% CI)		 Odds ratio
(95% CI)		 Odds ratio
(95% CI)		 Odds ratio
(95% CI)
No s. abuse 0 0 Ref. Ref. Ref.
S. abuse only .53
(− 0.59, 1.13)		 1.91
(1.09, 2.74)		 2.32
(1.50, 3.55)		 3.03
(1.45, 6.39)		 2.89
(1.52, 5.47)
S. abuse + sum (1) 1.28
(.85, 1.70)		 1.24
(.67, 1.80)		 5.34
(3.65, 7.81		 4.71
(2.33, 9.54)		 2.64
(1.65, 4.22)
S. abuse + sum (2) 2.03
(1.41, 2.65)		 .55
(.11, 1.21)		 12.35
(6.44, 23.71)		 7.30
(2.17, 24.51)		 2.41
(1.33, 4.43)
No s. abuse + sum (1) .99
(0.69, 1.29)		 .17
(−.13, 0.48)		 2.28
(1.84, 2.81)		 1.57
(1.04, 2.35)		 1.88
(1.48, 2.39)
No s. abuse + sum (2) 1.94
(1.36, 2.52)		 .35
(−.26, .96)		 5.16
(3.38, 7.86)		 2.44
(1.08, 5.53)		 3.52
(2.19, 5.73)
Open in a new tab

Note. All structural associations adjusted for gender and age.

S. abuse= sexual abuse; Maltreatment sum score= sum score includes parental neglect and parental physical abuse (Min: 0, Max: 2)

Child maltreatment and negative future expectations

For the association between maltreatment and negative future expectations, the maltreatment sum score was statistically significantly associated with having negative future expectations (mean difference: 0.99; 95% CI: 0.69, 1.29, p<0.001) when sexual abuse was not experienced, after adjusting for other covariates (Table 3). Additionally, reporting both physical abuse and neglect corresponded to a 1.98 positive difference in the mean of negative future expectations compared to no maltreatment. Sexual abuse was not statistically significantly associated with experiencing negative future expectations.

Child maltreatment and current drinking status

Regarding alcohol use, sexual abuse only and the child maltreatment sum score were statistically significantly associated with being in the current drinker class compared to the never drinker class, after adjusting for covariates and negative future expectations. Additionally, experiencing sexual abuse alongside other types of maltreatment was associated with higher odds of being in the current drinking class compared to the never drinking class. For example, the odds ratio for being in the current drinker class (compared to the never drinker class) for youth reporting sexual abuse only was 2.32 (95% CI 1.50, 3.55), and the odds ratio for being in the current drinker class for youth reporting all three types of abuse was 12.35 (95% CI 6.44, 23.71) (Table 4). Sexual abuse only was also associated with being in the non-active drinker class compared to the never drinker class; however, this association was not observed among participants reporting physical abuse and neglect (maltreatment sum score).

Child maltreatment and problematic alcohol use

Problematic alcohol use among current drinkers was statistically significantly associated with experiencing sexual abuse (Est: 1.91; 95% CI: 1.09, 2.74; p<0.001) when the maltreatment sum score is 0 (physical abuse and neglect not present), controlling for covariates and negative future expectations (Table 3). However, the association between the sum score and problematic alcohol use was not statistically significant when sexual abuse was not present. Reporting sexual abuse only corresponded with a 1.91 (95% CI: 1.09, 2.74) positive difference in means for problematic alcohol use compared to maltreatment (Table 4). Furthermore, experiencing sexual abuse and one other type of maltreatment (either physical abuse or neglect alone) corresponded with a 1.24 (95% CI: 0.67, 1.80) positive difference in means for problematic alcohol use. Experiencing all three types of maltreatment corresponded with a 0.55 (95% CI: 0.11, 1.21) positive difference in means of problematic alcohol use.

Suicidal ideation associations

Regarding associations with suicidal ideation, negative future expectations (OR: 1.45), current drinking status (OR: 1.80), sexual abuse only (OR: 2.89), and the maltreatment sum score (OR: 1.88) all were statistically significantly associated with suicidal ideation. However, problematic alcohol use was not a statistically significant predictor of suicidal ideation. The highest odds ratio among different patterns of child maltreatment for suicidal ideation was observed among participants reporting both physical abuse and neglect without sexual abuse (Table 4).

Discussion

Among the youth in our study, 23.5% reported suicidal ideation in the past year. This estimate was lower than previously reported suicidal ideation among youth living in the slums of Kampala (30%) (Swahn, Palmier, et al., 2012) but higher than the national prevalence of suicidal ideation among youth in Uganda (Swahn et al., 2010). Consistent with previous studies, negative future expectations had a direct effect on suicidal ideation (Abramson et al., 1998; Ballard, Patel, Ward, & Lamis, 2015), however, these effects were not observed via alcohol use.

Current drinking status (vs. never) was associated with suicidal ideation. However, problematic alcohol use was not associated with suicidal ideation. Our study presented a unique approach of estimating problematic alcohol use within classes of drinking behavior, without listwise deleting non-drinkers when examining problematic drinking behaviors. This analytic method is more flexible than restricting the analysis to only drinkers, analyzing only current drinking status among all participants, or coding all missing values on problematic alcohol use indicators to zero. Our finding of any alcohol use and suicidal ideation is consistent with the literature (Duncan et al., 1997; Borowsky et al., 2001) but inconsistent with the literature that demonstrates the association between problematic alcohol use and suicidal ideation (Fergusson et al., 2000; Reifman & Windle, 1995). However, this inconsistency might be due to a difference in populations assessed. Additionally, the inconsistency may also be due to the previous studies including all non-drinkers as a “0” on their problematic alcohol use measure, rather than including both current drinking status and problematic alcohol use together. Including all non-drinkers as “0” violates a crucial assumption in the model because problematic drinking cannot be assessed among non-drinkers who do not consume alcohol, in addition to violating the distributional assumption. Furthermore, our analytic method allowed us to examine the unique direct pathways of variables on the different aspects of the drinking process, further contributing to literature by utilizing this approach.

Regarding child maltreatment and problematic alcohol use, reporting sexual abuse only and sexual abuse alongside other types of abuse was statistically significantly associated with problematic alcohol use. Thus, the effects of child maltreatment on problematic alcohol use were only statistically significant when sexual abuse was present, and the effects of sexual abuse depended on the other types of maltreatment experienced alongside sexual abuse. The strong association between sexual abuse and problematic alcohol use has previously been demonstrated in the literature (Smith, Smith, & Grekin, 2014; Meyers et al., 2018). However, the interaction term between sexual abuse and the maltreatment sum score was in the opposite direction than expected (Shin et al., 2013). Youth who experienced sexual abuse alongside other types of abuse had a slightly lower association with problematic alcohol use compared to youth who only experienced sexual abuse; however, all combinations of sexual abuse alongside other types of maltreatment were associated with positive mean differences for problematic alcohol use. It should be noted that the context of the sexual abuse measure in this study involves any perpetrator, while the context of the physical abuse and neglect questions involve familial perpetrators. Additionally, the neglect measure incorporated neglect due to parental alcohol use. Also, the strong association between experiencing sexual abuse only with problematic alcohol use in this study might be partially explained by youth engaging in commercial sex work. The prevalence of commercial sex work in this sample among sexually active youth is 14%, and the majority of sex workers (90%) report previously being sexually abused (68%) (Swahn et al., 2016). While this study did not assess the prevalence of engaging in commercial sex work among youth who only report sexual abuse, commercial sex work may be one underlying mechanism driving the strong association between experiencing only sexual abuse with problematic alcohol use. Future research is needed to investigate the differences in outcomes related to child maltreatment patterns in this population. Moreover, it would be beneficial to determine the source of the maltreatment (familial vs. other) and other contextual information around the maltreatment experiences and long-term consequences in this population.

Additionally, the child maltreatment sum score (experiencing physical abuse and/or neglect without sexual abuse) was statistically significantly associated with negative future expectations. Sexual abuse was not statistically significantly associated with negative future expectations. While previous research has found an association with early adverse events and negative future expectations, the specific type of adverse event has not been extensively examined (Abramson et al., 1998; Ballard et al., 2015). Again, the physical abuse and neglect measures in our study both involved familial perpetrators, whereas the sexual abuse measure involved any perpetrator. There may be an underlying mechanism where the familial perpetration is driving the association with negative future expectations, compared to the sexual abuse measure where the perpetrator is not specified.

Our results also showed a statistically significant association between all patterns of child maltreatment and suicidal ideation. For youth who experienced both physical abuse and neglect, without sexual abuse, the suicidal ideation odds ratio was the highest. A previous meta-analysis showed robust evidence for the association for physical abuse and neglect with suicidal ideation (Norman et al., 2012). Youth who reported only sexual abuse experienced the second highest odds ratio for suicidal ideation. Youth who experienced sexual abuse in addition to another type of abuse only had a slightly lower odds ratio for suicidal ideation compared to youth who experienced only sexual abuse. These differences in child maltreatment patterns may be partially explained by the differences in perpetrators (familial perpetrators in physical abuse and neglect and non-specific perpetrator in sexual abuse). However, all types of child maltreatment were associated with an increased odds of suicidal ideation, consistent with previous literature (Brown et al., 1999; King & Merchant, 2008; Ng et al., 2015). As mentioned previously, identifying the context of child maltreatment would be beneficial for future studies.

Limitations

While this is the first study to our knowledge to document the associations between child maltreatment, alcohol use, negative future expectations, and suicidal ideation among youth living in the slums of Kampala, this study has several limitations. First, the sample is a convenience sample of youth, which may limit generalizability to service-seeking youth living in the slums who are attending UYDEL drop-in centers. Second, the survey is cross-sectional, and directionality of effects cannot be determined using this data alone. Future research would greatly benefit from longitudinal studies of this population. Caution should be used in evaluating the results of this study as to not infer causality from this data. One alternative model includes the possible reciprocal effect of alcohol use on negative future expectations. Instead of negative future expectations predicting alcohol use, alcohol use could also predict negative future expectations (Pompili et al., 2010). Additionally, suicidal ideation could also predict alcohol use, and this study did not examine those reciprocal effects. As mentioned previously, the timeline and context of the abuse variables cannot be ascertained from this data. For example, this study cannot determine whether abuse happened before or after alcohol use behaviors, negative future expectations, and suicidal ideations. Future studies should also seek to tease apart the timing of abuse, effects of different types of abuse, and incorporating the perpetrator source along with the frequency and severity of abuse. Again, caution should be used in interpreting these effects as causal due to the cross-sectional data type.

Implications

Despite the limitations, this study contributes to the growing body of literature on suicidality among youth living in sub-Saharan Africa. Approximately 23.5% of our sample reported experiencing suicidal ideation, and suicide prevention programs should be tailored to this population. Multi-level suicide prevention campaigns have demonstrated efficacy in decreasing suicide attempts among youth (Harris et al., 2016; Hegerl, Althaus, Schmidtke, & Niklewski, 2006). Multi-level suicide prevention programs focus on high-risk adolescents and training youth about coping and self-help skills, equipping community leaders on suicide prevention tools, and implementing a widespread media awareness campaign on suicide prevention (Harris et al., 2016; Hegerl et al., 2006). Additionally, best practice recommendations to reduce suicide attempts at the population level in low- and middle-income countries include restricting access to lethal weapons and substances used in suicide for that particular region (Petersen et al., 2016). Also at the population level, best practices recommend reducing and enforcing alcohol restriction among youth due to the strong connection between alcohol use, depression, and suicidality (Petersen et al., 2016). Since child maltreatment is also strongly associated with alcohol use and suicidality, some research from low- and middle-income countries on mental health support the implementation of child protection laws to protect children at high risk for child maltreatment (Fluke et al., 2012; Petersen et al., 2016).

Additionally, this study provided a unique approach to modeling alcohol use that allowed inclusion of all participants to examine both current drinking status as well as problematic alcohol use among current drinkers. This approach provides flexibility over previously utilized methods. Furthermore, this study found that current drinking status and not problematic alcohol use was associated with suicidal ideation. Interventions which delay alcohol use or target the initiation of alcohol use may be useful to incorporate in suicide prevention programs for this population.

Currently, Uganda Youth Development Link (UYDEL) provides child protection services, substance use counseling and rehabilitation, mental health counseling, HIV and sexually transmitted infection testing and counseling, and vocational training to youth living in the slums of Kampala. Future research should evaluate the feasibility and efficacy of implementing a tailored suicide prevention and mental health program in this population alongside current services offered at UYDEL. Additionally, the associations between child maltreatment, alcohol use, and suicidal ideation in this study should be evaluated in a longitudinal framework for future studies.

Acknowledgments

Funding

This work was supported by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health [R21AA22065 to Dr. Swahn].

Appendix

Appendix 1.

List of measures used in analysis

Child maltreatment
Physical abuse: “Did your parents ever beat you so hard you had bruises or marks?”
 Yes (n=325, 28.7%)
 No (n=805, 71.0%)
Sexual abuse: “Has someone ever raped you or forced you to have sex with him or her?”
 Yes (n=191, 16.8%)
 No (n=939, 82.8%)
Parental neglect (due to alcohol use): “Did a parent beat you when they were drunk?”
 Yes (n=140, 12.3%)
 No (n=988, 87.1%)
Alcohol use (Current drinking status)
“How old were you when you had your first full drink of alcohol?”
 1-12 (n=58, 5.1%)
 13-14 (n=116, 10.2%)
 15-16 (n=165, 14.6%)
 17-18 (n=66, 5.8%)
 Never (n=721, 63.6%)
“Have you had a drink of alcohol in the past year?”
 Yes (n=346, 30.5%)
 No (n=65, 5.7%)
Alcohol use (Problematic alcohol use)
Alcohol frequency: “How often do you have a drink containing alcohol?”
 Monthly or less (n=70, 6.2%)
 2-4 times a month (n=104, 9.2%)
 2-3 times a week (n=128, 11.3%)
 4 or more times a week (n=44, 3.9%)
Alcohol amount: “How many full drinks containing alcohol do you have in a typical day when you are drinking?”
 1-2 drinks (n=195, 17.2%)
 3-4 drinks (n=118, 10.4%)
 5 or more drinks (n=32, 2.8%)
Alcohol behavior (1): “Have you ever been seriously injured or hurt due to your drinking?”
 Yes (n=132, 11.6%)
 No (n=214, 18.9%)
Alcohol behavior (2): “Has someone else been seriously injured or hurt because of your drinking?”
 Yes (n=95, 8.4%)
 No (n=251, 22.1%)
Negative future expectations
Anticipating unhappiness: “Overall, what do you think about the following statements-I will be unhappy.”
 Yes (agree) (n=158, 13.9%)
 No (disagree) (n=972, 85.7%)
Anticipating bad things: “Overall, what do you think about the following statements- Bad things happen to people like me.”
 Yes (agree) (n=344, 30.3%)
 No (disagree) (n=786, 69.3%)
Anticipating early death: “Overall, what do you think about the following statements- I will probably die before I am thirty.”
 Yes (agree) (n=146, 12.9%)
 No (disagree) (n=985, 86.9%)
Suicidal ideation
“In the past year, did you ever think of killing yourself?”
 Yes (n=266, 23.5%)
 No (n=864, 76.2%)
Open in a new tab

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The authors have no conflicts of interest.

Disclaimer: The views expressed in this article are the author’s own and not an official positive of the National Institute on Alcohol Abuse and Alcoholism.

References

  1. Abramson LY, Alloy LB, Hogan ME, Whitehouse WG, Cornette M, Akhavan S, & Chiara A (1998). Suicidality and cognitive vulnerability to depression among college students: a prospective study. Journal of Adolescence, 21(4), 473–487. 10.1006/jado.1998.0167 [DOI] [PubMed] [Google Scholar]
  2. Ballard ED, Patel AB, Ward M, & Lamis DA (2015). Future disposition and suicidal ideation: mediation by depressive symptom clusters. Journal of Affective Disorders, 170, 1–6. 10.1016/j.jad.2014.08.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Borowsky IW, Ireland M, & Resnick MD (2001). Adolescent suicide attempts: risks and protectors. Pediatrics, 107(3), 485–493. [DOI] [PubMed] [Google Scholar]
  4. Brown J, Cohen P, Johnson JG, & Smailes EM (1999). Childhood abuse and neglect: specificity of effects on adolescent and young adult depression and suicidality. Journal of the American Academy of Child and Adolescent Psychiatry, 38(12), 1490–1496. 10.1097/00004583-199912000-00009 [DOI] [PubMed] [Google Scholar]
  5. Cheng Y, Li X, Lou C, Sonenstein FL, Kalamar A, Jejeebhoy S, … Ojengbede O (2014). 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 Journal of Adolescent Health: Official Publication of the Society for Adolescent Medicine, 55(6 Suppl), S31–38. 10.1016/j.jadohealth.2014.08.020 [DOI] [PubMed] [Google Scholar]
  6. Cluver L, Orkin M, Boyes ME, & Sherr L (2015). Child and Adolescent Suicide Attempts, Suicidal Behavior, and Adverse Childhood Experiences in South Africa: A Prospective Study. The Journal of Adolescent Health:Official Publication of the Society for Adolescent Medicine, 57(1), 52–59. 10.1016/j.jadohealth.2015.03.001 [DOI] [PubMed] [Google Scholar]
  7. Culbreth R, Swahn MH, Ndetei D, Ametewee L, & Kasirye R (2018). Suicidal Ideation among Youth Living in the Slums of Kampala, Uganda. International Journal of Environmental Research and Public Health, 15(2). 10.3390/ijerph15020298 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Duncan SC, Alpert A, Duncan TE, & Hops H (1997). Adolescent alcohol use development and young adult outcomes. Drug and Alcohol Dependence, 49(1), 39–48. [DOI] [PubMed] [Google Scholar]
  9. English DJ, Upadhyaya MP, Litrownik AJ, Marshall JM, Runyan DK, Graham JC, & Dubowitz H (2005). Maltreatment’s wake: the relationship of maltreatment dimensions to child outcomes. Child Abuse & Neglect, 29(5), 597–619. 10.1016/j.chiabu.2004.12.008 [DOI] [PubMed] [Google Scholar]
  10. Fergusson DM, Woodward LJ, & Horwood LJ (2000). Risk factors and life processes associated with the onset of suicidal behaviour during adolescence and early adulthood. Psychological Medicine, 30(1), 23–39. [DOI] [PubMed] [Google Scholar]
  11. Fluke JD, Goldman PS, Shriberg J, Hillis SD, Yun K, Allison S, & Light E (2012). Systems, strategies, and interventions for sustainable long-term care and protection of children with a history of living outside of family care. Child Abuse & Neglect, 36(10), 722–731. 10.1016/j.chiabu.2012.09.005 [DOI] [PubMed] [Google Scholar]
  12. Harris FM, Maxwell M, O’Connor R, Coyne JC, Arensman E, Coffey C, … Hegerl U (2016). Exploring synergistic interactions and catalysts in complex interventions: longitudinal, mixed methods case studies of an optimised multi-level suicide prevention intervention in four european countries (Ospi-Europe). BMC Public Health, 16, 268 10.1186/s12889-016-2942-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hegerl U, Althaus D, Schmidtke A, & Niklewski G (2006). The alliance against depression: 2-year evaluation of a community-based intervention to reduce suicidality. Psychological Medicine, 36(9), 1225–1233. 10.1017/S003329170600780X [DOI] [PubMed] [Google Scholar]
  14. Hong Y, Li X, Fang X, & Zhao R (2007). Correlates of suicidal ideation and attempt among female sex workers in China. Health Care for Women International, 28(5), 490–505. 10.1080/07399330701226529 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jamieson PE, & Romer D (2008). Unrealistic fatalism in U.S. youth ages 14 to 22: prevalence and characteristics. The Journal of Adolescent Health: Official Publication of the Society for Adolescent Medicine, 42(2), 154–160. 10.1016/j.jadohealth.2007.07.010 [DOI] [PubMed] [Google Scholar]
  16. Jessor R, & Jessor S (1977). Problem behavior and psychosocial development: A longitudinal study of youth. New York: Acad Press. [Google Scholar]
  17. Jones GD (1997). The role of drugs and alcohol in urban minority adolescent suicide attempts. Death Studies, 21(2), 189–202. 10.1080/074811897202074 [DOI] [PubMed] [Google Scholar]
  18. Kandel DB, Raveis VH, & Davies M (1991). Suicidal ideation in adolescence: Depression, substance use, and other risk factors. Journal of Youth and Adolescence, 20(2), 289–309. 10.1007/BF01537613 [DOI] [PubMed] [Google Scholar]
  19. King CA, & Merchant CR (2008). Social and Interpersonal Factors Relating to Adolescent Suicidality: A Review of the Literature. Archives of Suicide Research : Official Journal of the International Academy for Suicide Research, 12(3), 181–196. 10.1080/13811110802101203 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kinyanda E, Wamala D, Musisi S, & Hjelmeland H (2011). Suicide in urban Kampala, Uganda: a preliminary exploration. African Health Sciences, 11(2), 219. [PMC free article] [PubMed] [Google Scholar]
  21. Manly JT, Kim JE, Rogosch FA, & Cicchetti D (2001). Dimensions of child maltreatment and children’s adjustment: contributions of developmental timing and subtype. Development and Psychopathology, 13(4), 759–782. [PubMed] [Google Scholar]
  22. Marschall-Lévesque S, Castellanos-Ryan N, Parent S, Renaud J, Vitaro F, Boivin M, … Séguin JR (2016). Victimization, Suicidal Ideation, and Alcohol Use From Age 13 to 15 Years: Support for the Self-Medication Model. The Journal of Adolescent Health: Official Publication of the Society for Adolescent Medicine. 10.1016/j.jadohealth.2016.09.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Meyers JL, Sartor CE, Werner KB, Koenen KC, Grant BF, & Hasin D (2018). Childhood interpersonal violence and adult alcohol, cannabis, and tobacco use disorders: variation by race/ethnicity? Psychological Medicine, 1–11. 10.1017/S0033291717003208 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mills R, Alati R, Strathearn L, & Najman JM (2014). Alcohol and tobacco use among maltreated and non-maltreated adolescents in a birth cohort. Addiction (Abingdon, England), 109(4), 672–680. 10.1111/add.12447 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Mufune P (2000). Street Youth in Southern Africa. International Social Science Journal, 52(164), 233–243. 10.1111/1468-2451.00254 [DOI] [Google Scholar]
  26. Ng LC, Kirk CM, Kanyanganzi F, Fawzi MCS, Sezibera V, Shema E, … Betancourt TS (2015). Risk and protective factors for suicidal ideation and behaviour in Rwandan children. The British Journal of Psychiatry: The Journal of Mental Science, 207(3), 262–268. 10.1192/bjp.bp.114.154591 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Nguyen QC, Villaveces A, Marshall SW, Hussey JM, Halpern CT, & Poole C (2012). Adolescent expectations of early death predict adult risk behaviors. PloS One, 7(8), e41905 10.1371/journal.pone.0041905 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Norman RE, Byambaa M, De R, Butchart A, Scott J, & Vos T (2012). The long-term health consequences of child physical abuse, emotional abuse, and neglect: a systematic review and meta-analysis. PLoS Medicine, 9(11), e1001349 10.1371/journal.pmed.1001349 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Olema DK, Catani C, Ertl V, Saile R, & Neuner F (2014). The hidden effects of child maltreatment in a war region: correlates of psychopathology in two generations living in Northern Uganda. Journal of Traumatic Stress, 27(1), 35–41. 10.1002/jts.21892 [DOI] [PubMed] [Google Scholar]
  30. Omigbodun O, Dogra N, Esan O, & Adedokun B (2008). Prevalence and correlates of suicidal behaviour among adolescents in southwest Nigeria. The International Journal of Social Psychiatry, 54(1), 34–46. 10.1177/0020764007078360 [DOI] [PubMed] [Google Scholar]
  31. Page RM, & West JH (2011). Suicide Ideation and Psychosocial Distress in Sub-Saharan African Youth. American Journal of Health Behavior, 35(2), 129–141. 10.5993/AJHB.35.2.1 [DOI] [PubMed] [Google Scholar]
  32. Petersen I, Evans-Lacko S, Semrau M, Barry MM, Chisholm D, Gronholm P, … Thornicroft G (2016). Promotion, prevention and protection: interventions at the population- and community-levels for mental, neurological and substance use disorders in low- and middle-income countries. International Journal of Mental Health Systems, 10, 30 10.1186/s13033-016-0060-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Pompili M, Serafini G, Innamorati M, Dominici G, Ferracuti S, Kotzalidis GD, … Lester D (2010). Suicidal behavior and alcohol abuse. International Journal of Environmental Research and Public Health, 7(4), 1392–1431. 10.3390/ijerph7041392 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Reifman A, & Windle M (1995). Adolescent suicidal behaviors as a function of depression, hopelessness, alcohol use, and social support: a longitudinal investigation. American Journal of Community Psychology, 23(3), 329–354. [DOI] [PubMed] [Google Scholar]
  35. Schilling EA, Aseltine RH, Glanovsky JL, James A, & Jacobs D (2009). Adolescent alcohol use, suicidal ideation, and suicide attempts. The Journal of Adolescent Health: Official Publication of the Society for Adolescent Medicine, 44(4), 335–341. 10.1016/j.jadohealth.2008.08.006 [DOI] [PubMed] [Google Scholar]
  36. Shaikh MA, Lloyd J, Acquah E, Celedonia KL, & L Wilson M (2016). Suicide attempts and behavioral correlates among a nationally representative sample of school-attending adolescents in the Republic of Malawi. BMC Public Health, 16(1), 843 10.1186/s12889-016-3509-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sher L, Sperling D, Zalsman G, Vardi G, & Merrick J (2006). Alcohol and suicidal behavior in adolescents. Minerva Pediatrica, 58(4), 333–339. [PubMed] [Google Scholar]
  38. Sher Leo, & Zalsman G (2005). Alcohol and adolescent suicide. International Journal of Adolescent Medicine and Health, 17(3), 197–203. [DOI] [PubMed] [Google Scholar]
  39. Shin Sunny H., Chung Y, & Rosenberg RD (2016). Identifying Sensitive Periods for Alcohol Use: The Roles of Timing and Chronicity of Child Physical Abuse. Alcoholism, Clinical and Experimental Research, 40(5), 1020–1029. 10.1111/acer.13038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Shin Sunny H., Miller DP, & Teicher MH (2013). Exposure to childhood neglect and physical abuse and developmental trajectories of heavy episodic drinking from early adolescence into young adulthood. Drug and Alcohol Dependence, 127(1–3), 31–38. 10.1016/j.drugalcdep.2012.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shin Sunny Hyucksun, Edwards EM, & Heeren T (2009). Child abuse and neglect: relations to adolescent binge drinking in the national longitudinal study of Adolescent Health (AddHealth) Study. Addictive Behaviors, 34(3), 277–280. 10.1016/j.addbeh.2008.10.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Smith KZ, Smith PH, & Grekin ER (2014). Childhood sexual abuse, distress, and alcohol-related problems: Moderation by drinking to cope. Psychology of Addictive Behaviors: Journal of the Society of Psychologists in Addictive Behaviors, 28(2), 532–537. 10.1037/a0035381 [DOI] [PubMed] [Google Scholar]
  43. Steinhausen H-C, Bösiger R, & Metzke CW (2006). Stability, correlates, and outcome of adolescent suicidal risk. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 47(7), 713–722. 10.1111/j.1469-7610.2005.01569.x [DOI] [PubMed] [Google Scholar]
  44. Culbreth R, Swahn MH, Ndetei D, Ametewee L, & Kasirye R (2018). Suicidal Ideation among Youth Living in the Slums of Kampala, Uganda. International Journal of Environmental Research and Public Health, 15(2). 10.3390/ijerph15020298 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Swahn MH, Culbreth R, Salazar LF, Tumwesigye NM, Jernigan DH, Kasirye R, & Obot IS (2020). The Prevalence and Context of Alcohol Use, Problem Drinking and Alcohol-Related Harm among Youth Living in the Slums of Kampala, Uganda. International Journal of Environmental Research and Public Health, 17(7). 10.3390/ijerph17072451 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Swahn MH, Palmier JB, Kasirye R, & Yao H (2012). Correlates of Suicide Ideation and Attempt among Youth Living in the Slums of Kampala. International Journal of Environmental Research and Public Health, 9(2), 596–609. 10.3390/ijerph9020596 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Swahn MH, Culbreth RE, Staton CA, Self-Brown SR, & Kasirye R (2017). Alcohol-Related Physical Abuse of Children in the Slums of Kampala, Uganda. International Journal of Environmental Research and Public Health, 14(10). 10.3390/ijerph14101124 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Swahn MH, Culbreth R, Salazar LF, Kasirye R, & Seeley J (2016). Prevalence of HIV and Associated Risks of Sex Work among Youth in the Slums of Kampala. AIDS Research and Treatment, 2016, 5360180 10.1155/2016/5360180 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Swahn MH, Dill LJ, Palmier JB, & Kasirye R (2015). Girls and Young Women Living in the Slums of Kampala. SAGE Open, 5(2), 2158244015580853. 10.1177/2158244015580853 [DOI] [Google Scholar]
  50. Swahn MH, Gressard L, Palmier JB, Kasirye R, Lynch C, & Yao H (2012). Serious Violence Victimization and Perpetration among Youth Living in the Slums of Kampala, Uganda. Western Journal of Emergency Medicine, 13(3). Retrieved from http://escholarship.org/uc/item/9h4660nd [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Swahn MH, Palmier JB, Kasirye R, & Yao H (2012). Correlates of Suicide Ideation and Attempt among Youth Living in the Slums of Kampala. International Journal of Environmental Research and Public Health, 9(2), 596–609. 10.3390/ijerph9020596 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Swahn MH, Bossarte R, Eliman D, Gaylor E, & Jayaraman S (2010). Prevalence and correlates of suicidal ideation and physical fighting: A comparison of students in Botswana, Kenya, Uganda, Zambia and the United States. 2, 195–205. [Google Scholar]
  53. Swahn Monica, Haberlen M, & Palmier JB (2014). Alcohol and drug use and other high-risk behaviors among youth in the slums of Kampala, Uganda: Perceptions and contexts obtained through focus groups. The International Journal of Alcohol and Drug Research, 3(4), 289–295. (Convenience sample). [Google Scholar]
  54. Villodas MT, Litrownik AJ, Thompson R, Jones D, Roesch SC, Hussey JM, … Dubowitz H (2015). Developmental transitions in presentations of externalizing problems among boys and girls at risk for child maltreatment. Development and Psychopathology, 27(1), 205–219. 10.1017/S0954579414000728 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wasserman D (Ed.). (2009). Oxford Textbook of Suicidology and Suicide Prevention: A Global Perspective. Oxford: Oxford University Press. [Google Scholar]
  56. Weiland BJ, Nigg JT, Welsh RC, Yau W-YW, Zubieta J-K, Zucker RA, & Heitzeg MM (2012). Resiliency in adolescents at high risk for substance abuse: flexible adaptation via subthalamic nucleus and linkage to drinking and drug use in early adulthood. Alcoholism, Clinical and Experimental Research, 36(8), 1355–1364. 10.1111/j.1530-0277.2012.01741.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wilsnack SC, Wilsnack RW, Kristjanson AF, Vogeltanz-Holm ND, & Windle M (2004). Alcohol use and suicidal behavior in women: longitudinal patterns in a U.S. national sample. Alcoholism, Clinical and Experimental Research, 28(5 Suppl), 38S–47S. [DOI] [PubMed] [Google Scholar]
  58. Wlodarczyk O, Schwarze M, Rumpf H-J, Metzner F, & Pawils S (2017). Protective mental health factors in children of parents with alcohol and drug use disorders: A systematic review. PloS One, 12(6), e0179140 10.1371/journal.pone.0179140 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. World Health organization. (2016). Adolescent Health Epidemiology. Retrieved from http://www.who.int/maternal_child_adolescent/epidemiology/adolescence/en/

RESOURCES