[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure.
Hypothesized Mediating Effects of Depressive Symptoms on the Association Between Baseline Nonmedical Use of Prescription Drugs (NMUPD) and Subsequent Suicidality
Hypothesized Mediating Effects of Depressive Symptoms on the Association Between Baseline Nonmedical Use of Prescription Drugs (NMUPD) and Subsequent Suicidality

Suicidality included suicidal ideation and suicidal attempts. Depressive symptoms were measured using the Center for Epidemiology Scale for Depression.

Table 1.  
Baseline Sample Characteristics of 3273 Patients Stratified by NMUPD
Baseline Sample Characteristics of 3273 Patients Stratified by NMUPD
Table 2.  
Longitudinal Relationships Between NMUPD, Suicidal Ideation, and Suicidal Attempts: Multivariable Logistic Regression Analyses in 3145 Patients
Longitudinal Relationships Between NMUPD, Suicidal Ideation, and Suicidal Attempts: Multivariable Logistic Regression Analyses in 3145 Patients
Table 3.  
Associations of Interaction Items With Suicidal Ideation and Suicidal Attempts: Univariable Logistic Regression Analyses in 3145 Patients
Associations of Interaction Items With Suicidal Ideation and Suicidal Attempts: Univariable Logistic Regression Analyses in 3145 Patients
Table 4.  
Mediating Effects of Depressive Symptoms on the Relationships Between Baseline NMUPD and Suicidal Ideation and Attempts at Follow-up in 3145 Patients
Mediating Effects of Depressive Symptoms on the Relationships Between Baseline NMUPD and Suicidal Ideation and Attempts at Follow-up in 3145 Patients
1.
He  ZX, ed.  Suicidology. Beijing, China: China Press of Traditional Chinese Medicine; 1997.
2.
Nock  MK, Borges  G, Bromet  EJ, Cha  CB, Kessler  RC, Lee  S.  Suicide and suicidal behavior.  Epidemiol Rev. 2008;30:133-154.PubMedGoogle ScholarCrossref
3.
Miranda  R, Scott  M, Hicks  R, Wilcox  HC, Harris Munfakh  JL, Shaffer  D.  Suicide attempt characteristics, diagnoses, and future attempts: comparing multiple attempters to single attempters and ideators.  J Am Acad Child Adolesc Psychiatry. 2008;47(1):32-40.PubMedGoogle ScholarCrossref
4.
Ruengorn  C, Sanichwankul  K, Niwatananun  W, Mahatnirunkul  S, Pumpaisalchai  W, Patumanond  J.  Incidence and risk factors of suicide reattempts within 1 year after psychiatric hospital discharge in mood disorder patients.  Clin Epidemiol. 2011;3:305-313.PubMedGoogle Scholar
5.
Zhang  J, Sun  L, Liu  Y, Zhang  J.  The change in suicide rates between 2002 and 2011 in China.  Suicide Life Threat Behav. 2014;44(5):560-568.PubMedGoogle ScholarCrossref
6.
Cui  S, Cheng  Y, Xu  Z, Chen  D, Wang  Y.  Peer relationships and suicide ideation and attempts among Chinese adolescents.  Child Care Health Dev. 2011;37(5):692-702.PubMedGoogle ScholarCrossref
7.
Liu  X, Tein  JY, Zhao  Z, Sandler  IN.  Suicidality and correlates among rural adolescents of China.  J Adolesc Health. 2005;37(6):443-451.PubMedGoogle ScholarCrossref
8.
World Health Organization. Preventing suicide: a global imperative. http://apps.who.int/iris/bitstream/10665/131056/1/9789241564779_eng.pdf. Accessed July 12, 2016.
9.
Falbo  T, Poston  DL.  How and why the one-child policy works in China.  Adv Popul. 1994;2:205-229.PubMedGoogle Scholar
10.
Moscicki  EK.  Epidemiology of completed and attempted suicide: toward a framework for prevention.  Clin Neurosci Res. 2001;1(5):310-323.Google ScholarCrossref
11.
Hedden  SL, Kennet  J, Lipari  R,  et al. Substance Abuse and Mental Health Services Administration. Behavioral health trends in the United States: results from the 2014 National Survey on Drug Use and Health. 2015. http://www.samhsa.gov/data/sites/default/files/NSDUH-FRR1-2014/NSDUH-FRR1-2014.pdf. Accessed July 12, 2016.
12.
Ford  JA, Perna  D.  Prescription drug misuse and suicidal ideation: Findings from the National Survey on Drug Use and Health.  Drug Alcohol Depend. 2015;157:192-196.PubMedGoogle ScholarCrossref
13.
Guo  L, Xu  Y, Deng  J,  et al.  Nonmedical use of prescription pain relievers among high school students in China: a multilevel analysis.  BMJ Open. 2015;5(7):e007569.PubMedGoogle ScholarCrossref
14.
Wang  H, Deng  J, Zhou  X,  et al.  The nonmedical use of prescription medicines among high school students: a cross-sectional study in Southern China.  Drug Alcohol Depend. 2014;141:9-15.PubMedGoogle ScholarCrossref
15.
Juan  W, Jian-Xiong  D, Lan  G,  et al.  Nonmedical use of psychoactive drugs in relation to suicide tendencies among Chinese adolescents.  Addict Behav. 2015;51:31-37.PubMedGoogle ScholarCrossref
16.
Dvorak  RD, Lamis  DA, Malone  PS.  Alcohol use, depressive symptoms, and impulsivity as risk factors for suicide proneness among college students.  J Affect Disord. 2013;149(1-3):326-334.PubMedGoogle ScholarCrossref
17.
Skarbø  T, Rosenvinge  JH, Holte  A.  Alcohol problems, mental disorder and mental health among suicide attempters 5-9 years after treatment by child and adolescent outpatient psychiatry.  Nord J Psychiatry. 2006;60(5):351-358.PubMedGoogle ScholarCrossref
18.
Nrugham  L, Holen  A, Sund  AM.  Associations between attempted suicide, violent life events, depressive symptoms, and resilience in adolescents and young adults.  J Nerv Ment Dis. 2010;198(2):131-136.PubMedGoogle ScholarCrossref
19.
Zullig  KJ, Divin  AL.  The association between nonmedical prescription drug use, depressive symptoms, and suicidality among college students.  Addict Behav. 2012;37(8):890-899.PubMedGoogle ScholarCrossref
20.
Kessler  RC, McGonagle  KA, Zhao  S,  et al.  Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey.  Arch Gen Psychiatry. 1994;51(1):8-19.PubMedGoogle ScholarCrossref
21.
Volkow  ND.  The reality of comorbidity: depression and drug abuse.  Biol Psychiatry. 2004;56(10):714-717.PubMedGoogle ScholarCrossref
22.
Shimamoto  A, Debold  JF, Holly  EN, Miczek  KA.  Blunted accumbal dopamine response to cocaine following chronic social stress in female rats: exploring a link between depression and drug abuse.  Psychopharmacology (Berl). 2011;218(1):271-279.PubMedGoogle ScholarCrossref
23.
Cluver  L, Orkin  M, Boyes  ME, Sherr  L.  Child and adolescent suicide attempts, suicidal behavior, and adverse childhood experiences in south africa: a prospective study.  J Adolesc Health. 2015;57(1):52-59.PubMedGoogle ScholarCrossref
24.
Miller  AB, Jenness  JL, Oppenheimer  CW, Gottleib  AL, Young  JF, Hankin  BL.  Childhood emotional maltreatment as a robust predictor of suicidal ideation: a 3-year multi-wave, prospective investigation  [published online March 31, 2016].  J Abnorm Child Psychol. doi:10.1007/s10802-016-0150-z.PubMedGoogle Scholar
25.
Kwok  SY, Shek  DT.  Socio-demographic correlates of suicidal ideation among Chinese adolescents in Hong Kong.  Int J Adolesc Med Health. 2008;20(4):463-472.PubMedGoogle Scholar
26.
Cheng  CP, Yen  CF, Ko  CH, Yen  JY.  Factor structure of the Center for Epidemiologic Studies Depression Scale in Taiwanese adolescents.  Compr Psychiatry. 2012;53(3):299-307.PubMedGoogle ScholarCrossref
27.
Lee  SW, Stewart  SM, Byrne  BM,  et al.  Factor structure of the Center for Epidemiological Studies Depression Scale in Hong Kong adolescents.  J Pers Assess. 2008;90(2):175-184.PubMedGoogle ScholarCrossref
28.
Radloff  LS.  The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults.  J Youth Adolesc. 1991;20(2):149-166.PubMedGoogle ScholarCrossref
29.
Firth  D.  Bias reduction of maximum likelihood estimates.  Biometrika. 1993;80:27-38.Google ScholarCrossref
30.
Wang  X.  Firth logistic regression for rare variant association tests.  Front Genet. 2014;5:187.PubMedGoogle Scholar
31.
Preacher  KJ, Hayes  AF.  Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models.  Behav Res Methods. 2008;40(3):879-891.PubMedGoogle ScholarCrossref
32.
Gunzler  D, Chen  T, Wu  P, Zhang  H.  Introduction to mediation analysis with structural equation modeling.  Shanghai Arch Psychiatry. 2013;25(6):390-394.PubMedGoogle Scholar
33.
Mackinnon  DP, Lockwood  CM, Williams  J.  Confidence limits for the indirect effect: distribution of the product and resampling methods.  Multivariate Behav Res. 2004;39(1):99.PubMedGoogle ScholarCrossref
34.
Campbell  G, Bruno  R, Darke  S,  et al.  Prevalence and correlates of suicidal thoughts and suicide attempts in people prescribed pharmaceutical opioids for chronic pain.  Clin J Pain. 2016;32(4):292-301.PubMedGoogle ScholarCrossref
35.
Wichstrøm  L.  Predictors of adolescent suicide attempts: a nationally representative longitudinal study of Norwegian adolescents.  J Am Acad Child Adolesc Psychiatry. 2000;39(5):603-610.PubMedGoogle ScholarCrossref
36.
Lewinsohn  PM, Rohde  P, Seeley  JR.  Psychosocial risk factors for future adolescent suicide attempts.  J Consult Clin Psychol. 1994;62(2):297-305.PubMedGoogle ScholarCrossref
37.
Rhoades  H, Winetrobe  H, Rice  E.  Prescription drug misuse among homeless youth.  Drug Alcohol Depend. 2014;138:229-233.PubMedGoogle ScholarCrossref
38.
Crumley  FE.  Substance abuse and adolescent suicidal behavior.  JAMA. 1990;263(22):3051-3056.PubMedGoogle ScholarCrossref
39.
Noori  R, Rafiey  H, Azizabadi-Farahani  M,  et al.  Risk factors of suicidal ideation and attempt in women with drug user spouses.  J Chin Med Assoc. 2013;76(11):648-652.PubMedGoogle ScholarCrossref
40.
Brockie  TN, Dana-Sacco  G, Wallen  GR, Wilcox  HC, Campbell  JC.  The relationship of adverse childhood experiences to PTSD, depression, poly-drug use and suicide attempt in reservation-based native american adolescents and young adults.  Am J Community Psychol. 2015;55(3-4):411-421.PubMedGoogle ScholarCrossref
41.
Bukstein  OG, Brent  DA, Perper  JA,  et al.  Risk factors for completed suicide among adolescents with a lifetime history of substance abuse: a case-control study.  Acta Psychiatr Scand. 1993;88(6):403-408.PubMedGoogle ScholarCrossref
42.
Brodsky  BS, Malone  KM, Ellis  SP, Dulit  RA, Mann  JJ.  Characteristics of borderline personality disorder associated with suicidal behavior.  Am J Psychiatry. 1997;154(12):1715-1719.PubMedGoogle ScholarCrossref
43.
Liu  X, Tein  JY, Zhao  Z, Sandler  IN.  Suicidality and correlates among rural adolescents of China.  J Adolesc Health. 2005;37(6):443-451.PubMedGoogle ScholarCrossref
44.
Wong  MM, Brower  KJ.  The prospective relationship between sleep problems and suicidal behavior in the National Longitudinal Study of Adolescent Health.  J Psychiatr Res. 2012;46(7):953-959.PubMedGoogle ScholarCrossref
Original Investigation
October 2016

Association Between Nonmedical Use of Prescription Drugs and Suicidal Behavior Among Adolescents

Author Affiliations
  • 1Department of Medical Statistics and Epidemiology, School of Public Health, Sun Yat-sen University, Guangzhou 510080, People’s Republic of China
  • 2Epidemiology, Biostatistics and Clinical Research Center, School of Public Health, Université Libre de Bruxelles (ULB), Belgium
  • 3Center for ADR Monitoring of Guangdong, Guangzhou 510080, People’s Republic of China
 

Copyright 2016 American Medical Association. All Rights Reserved.

JAMA Pediatr. 2016;170(10):971-978. doi:10.1001/jamapediatrics.2016.1802
Key Points

Question  Is nonmedical use of prescription drugs by adolescents associated with subsequent suicidal ideation and attempts while controlling for depressive symptoms?

Findings  In this longitudinal study, adolescent baseline opioids misuse, sedatives misuse, and nonmedical use of any prescription drug were associated with suicidal ideation at follow-up. Additionally, baseline opioids misuse and any prescription drug misuse were also associated with subsequent suicide attempts.

Meaning  Youths with nonmedical use of prescription drugs should be closely monitored for suicidal ideation and suicide attempts.

Abstract

Importance  Suicidal behavior is a leading cause of injury and death worldwide, and previous cross-sectional studies have demonstrated that nonmedical use of prescription drugs (NMUPD) was associated with suicidality. However, there is not any study in China having examined the longitudinal relationships between NMUPD, suicidal ideation, and suicidal attempts, as well as explored the potential mediating effects of depressive symptoms.

Objective  To determine whether baseline NMUPD was associated with subsequent suicidal ideation and attempts while controlling for depressive symptoms and to determine whether the increased risks were mediated by depressive symptoms.

Design, Setting, and Participants  In this longitudinal study, a total of 3273 students in randomly selected schools in Guangzhou were surveyed from 2009 to 2010 (response rate, 96.8%) and followed up at 1 year (2011-2012; retention rate, 96.1%). The dates of data analysis were October 9, 2015, to October 15, 2015; additional data analysis occurred March 23, 2016, to March 29, 2016.

Main Outcomes and Measures  Suicidal ideation, suicidal attempts, NMUPD, depressive symptoms, and alcohol-related problems.

Results  Overall, 3273 adolescents (mean [SD] age, 13.7 [1.0] years) were recruited for this study. The final results showed that after controlling for sociodemographic information (including sex, age, household socioeconomic status, and living arrangements), baseline depressive symptoms, baseline alcohol-related problems, baseline suicidal ideation, and baseline suicidal attempts, baseline opioids misuse (adjusted odds ratio [AOR], 2.31; 95% CI, 1.30-4.11), sedatives misuse (AOR, 4.46; 95% CI, 1.54-7.94), and nonmedical use of any prescription drug (AOR, 1.97; 95% CI, 1.21-3.23) were positively associated with suicidal ideation at follow-up. Additionally, baseline opioid misuse (AOR, 3.39; 95% CI, 1.33-5.63) and nonmedical use of any prescription drug (AOR, 2.91; 95% CI, 1.26-3.71) were also associated with subsequent suicidal attempts after controlling for sex, age, household socioeconomic status, living arrangements, depressive symptoms, alcohol-related problems, suicidal ideation, and suicidal attempts at baseline. There were significant standardized indirect effects of baseline opioids misuse on subsequent suicidal ideation (standardized β estimate = 0.020; 95% CI, 0.010-0.030) and suicidal attempts (standardized β estimate = 0.009; 95% CI, 0.004-0.015) through depressive symptoms; the standardized indirect effect of baseline sedatives misuse on subsequent suicidal ideation through depressive symptoms was also significant (standardized β estimate = 0.016; 95% CI, 0.005-0.026).

Conclusions and Relevance  In this study, NMUPD at baseline was associated with subsequent suicidal ideation and attempts. These findings support that proper surveillance systems with the potential to reduce adolescent suicidality should be established to control and supervise suicidality and NMUPD among Chinese adolescents.

Introduction

Suicidal behavior is a range of behaviors that include suicidal ideation, suicidal attempts, and suicide itself and is a leading cause of injury and death worldwide.1,2 Moreover, suicidal ideation and suicidal attempts have shown an increased risk of recurrence.3,4 Although the overall rate of suicide in China has been significantly reduced from the rate in the 1990s,5 suicidal ideation and attempts are still currently serious problems among adolescents in China.6 Liu et al7 reported that 19% of Chinese rural adolescents reported having had suicidal ideation, and 7% reported having made a suicidal attempt during the past 6 months. Cui et al6 showed that 16.9% of adolescents reported having had suicidal ideation, and 8.0% reported having actually attempted suicide during the past 12 months. Additionally, suicidal ideation and suicidal attempts among youth can lead to a substantial economic, social, and psychological burden for the individuals, families, and communities.8 In China, most families have only 1 child as a result of the government’s previous 1-child policy9; however, suicidal ideation and attempts in Chinese youth have rarely been studied.

Although suicidal behavior is multifactorial, approximately 90% of the individuals who die from suicide have substance use, depression or other mental disorders, or a combination of both.10 According to the 2014 National Survey on Drug Use and Health (NSDUH) in the United States, prescription drugs were the second most popular drug among adolescents, and 655 000 adolescents aged 12 to 17 years reported nonmedical use of prescription drugs (NMUPD).11 A cross-sectional study using the 2012 NSDUH data showed that the respondents who admitted misusing prescription drugs in the past year were more likely to report having seriously thought about suicide.12 Our 3 previous cross-sectional studies13-15 in Chongqing and Guangdong Provinces also demonstrated that NMUPD was prevalent among Chinese adolescents, and one of them15 illustrated that NMUPD was positively associated with suicidal ideation and suicidal attempts. Although drug use is a known correlate of suicidality, there is a relative dearth of studies on the relationships between NMUPD, suicidal ideation, and suicidal attempts. Furthermore, most prior studies on this topic used a cross-sectional design, and could not determine whether NMUPD preceded suicidality.

Depression is one of the risk factors that has been implicated in suicidal behavior. Previous cross-sectional studies and longitudinal investigations have demonstrated that depressive symptoms increase the risk of suicidal behavior among adolescents.16-18 Moreover, prior studies15,19 have also reported a significant association between NMUPD and depressive symptoms, and the association between depression and drug use may be related to similarities in the neurobiology.20-22 Therefore, depressive symptoms may play a mediator role in the relationship between NMUPD and suicidality, but to our knowledge no study has examined the potential mediating effect.

We thus designed this prospective study among Chinese adolescents to examine the longitudinal relationships between NMUPD, suicidal ideation, and suicidal attempts while controlling for the effects of important covariates and to test for the potential mediating effects of depressive symptoms.

Methods
Participants and Procedures

This study analyzed data from a school-based survey that investigated adolescents NMUPD and suicidality in Guangzhou between 2009 and 2012. In 2009, all middle schools in Guangzhou (n = 121) were stratified by type (public and private). A random sample of 12 middle schools was selected with a probability proportional to size. All students grade 7 and 8 within the selected schools were invited to participate in this study voluntarily. The baseline sample included 3273 students (10-18 years of age) who were interviewed at baseline (2009-2010; response rate, 96.8%) and followed up at 1 year (2011-2012; retention rate, 96.1%); the follow-up sample included 3145 students. To protect the privacy of the students, the anonymity of the self-reported questionnaires was guaranteed, and the questionnaires were administered by research assistants in the classrooms without the presence of teachers (to avoid any potential information bias). After the procedure had been fully explained, written informed consent was obtained from each participating student who was at least 18 years of age. If the student was younger than 18 years of age, written informed consent was obtained from one of the student’s parents (or legal guardian). The study was approved by the Sun Yat-Sen University School of Public Health institutional review board. All adolescents and their parents were provided with a list of resources in case of needing help.

Measures
Outcome

Suicidal ideation was defined as responding “1 or more times” to the question “During the past 12 months, how many times did you seriously consider attempting suicide?” Suicidal attempt was assessed by the question “During the past 12 months, how many times did you actually attempt suicide?” and asked that the student rate on a scale of zero or once or more.

Predictor Variables

Nonmedical use of prescription drugs was defined as taking medications without a physicians’s prescription or solely for the feeling or experience caused by the drug (0 = no and 1 = yes). In this study, 4 measures of NMUPD included nonmedical use of opioids, sedatives, and stimulants separately, as well as nonmedical use of any of the above-mentioned prescription drugs in the past 12 months. Opioids included compound cough syrup with codeine (codeine), compound liquorice tablets (opium), tramadol hydrochloride, and diphenoxylate. Sedatives included compound aminopyrine phenacetin tablets (barbiturates), diazepam or triazolam (benzodiazepines), and phenobarbital and scopolamine hydrobromide tablets (barbiturates). Stimulants included paracetamol caffeine aspirin powder (caffeine).

Other Covariates

Factors that have previously been reported to be associated with suicidal behavior or NMUPD were taken into account.23-25 The 20-item Chinese version of the Center for Epidemiology Scale for Depression (CES-D) was used to identify whether individuals had depressive symptoms; the Chinese version of this scale has been previously validated among Chinese adolescents.26,27 Higher scores indicated more severe depressive symptoms, with a maximum score of 60.28 Alcohol-related problems were measured by asking students one question, “How often did you experience the following problems in the past 12 months due to drinking alcohol: got into trouble with your parents, had problems at school, had problems with friends, and got into a physical fight?” The responses included “never,” “once,” “twice,” “3-4 times,” or “5 or more times.” Data from all 5 responses were later recorded into a dichotomous variable, alcohol-related problems (0 = no, 1 = yes).

Sociodemographic information included age, sex (0 = male, 1 = female), household socioeconomic status (HSS), and living arrangement. Household socioeconomic status was measured by asking about the student’s perceptions of his or her household’s current socioeconomic status (responses were coded as excellent or very good = 1, good = 2, and fair or poor = 3). Living arrangement was assessed by asking who lived in the student’s primary home (responses were coded as living with both parents = 1, living with father only = 2, living with mother only = 3, living with others = 4).

Statistical Analysis

First, descriptive analyses stratified by baseline NMUPD were used to describe the sample characteristics at baseline. Second, logistic regression analyses were conducted that accounted for the stratified cluster survey design and were weighted for nonresponse. For the outcome with low event rate, logistic regression models used the Firth penalized likelihood approach.29,30 Univariable logistic analyses were performed to explore the potential factors associated with suicidal ideation and attempts at follow-up. The variables that were significant at .10 level in univariable analyses or widely reported in the literature (eg, sex and age) were entered in the multivariable logistic regression models simultaneously to examine the independent longitudinal relationships between NMUPD, suicidal ideation, and suicidal attempts. Third, the interactions between NMUPD and depressive symptoms were tested using univariable logistic regression models before the mediation analyses. Structural equation modeling (SEM) analyses using the robust maximum likelihood method were performed to measure the mediating effects of depressive symptoms on the relationships between NMUPD and suicidal behaviors (Figure). As the depressive symptoms score was a continuous variable and the measure of NMUPD were dichotomized, standardized probit coefficients, standardized indirect effects, and standardized total effects were reported, and the bias-corrected 95% CIs were estimated using 1000 bootstrap samples.31-33 All data were analyzed using SAS 9.2 (SAS Institute Inc) and Mplus version 7.0 (Muthén and Muthén). P values less than .05 were considered statistically significant (tested 2-sided).

Results
Baseline Sample Characteristics Stratified by NMUPD

The sample characteristics at baseline are shown in Table 1. Of the total sample, 48.9% (1599 of 3273) were boys, and 51.1% (1674 of 3273) were girls. The mean (SD) age of the total students was 13.7 (1.0) years. The proportion of students who reported nonmedical use of opioids, sedatives, stimulants, and any prescription drug was 1.8% (60 of 3273), 0.8% (26 of 3273), 1.8% (58 of 3273), and 2.8% (91 of 3273), respectively.

Univariable Analyses of Suicidal Ideation and Suicidal Attempts at Follow-up

Overall, 17.0% (534 of 3145) and 3.0% (94 of 3145) students reported suicidal ideation and attempts at follow-up, respectively. Without adjusting for other variables, baseline nonmedical use of opioids (odds ratio [OR], 3.03; 95% CI, 1.47-6.22), sedatives misuse (OR, 4.60; 95% CI, 2.09-10.15), stimulants misuse (OR. 2.38; 95% CI, 1.07-5.84), and nonmedical use of any prescription drug (OR, 1.93; 95% CI, 1.19-3.11) were significantly associated with suicidal ideation at follow-up. Additionally, without adjusting for other variables, subsequent suicidal attempts were more common in students who reported baseline opioids misuse (OR, 5.96; 95% CI, 2.78-9.53), sedatives misuse (OR, 4.68; 95% CI, 1.86-10.83), stimulants misuse (OR, 2.67; 95% CI, 1.76-10.82), and any prescription drug misuse (OR, 4.20; 95% CI, 2.07-8.56). Moreover, female students were more likely to be involved in subsequent suicidal ideation than male students (OR, 1.58; 95% CI, 1.31-1.91). Other baseline characteristics including HSS, living arrangement, depressive symptoms, alcohol-related problems, suicidal ideation, and suicidal attempt were also associated with suicidal ideation and attempts at follow-up (eTable in the Supplement).

Longitudinal Relationships Between NMUPD and Suicidality: Multivariable Analyses

As shown in Table 2, our final multivariable logistic regression models demonstrated that baseline opioids misuse (adjusted odds ratio [AOR], 2.31; 95% CI, 1.30-4.11), sedatives misuse (AOR, 4.46; 95% CI, 1.54-7.94), and nonmedical use of any prescription drug (AOR, 1.97; 95% CI, 1.21-3.23) were positively associated with suicidal ideation at follow-up after controlling for sociodemographic information (including sex, age, HSS, and living arrangement), baseline depressive symptoms, baseline alcohol-related problems, baseline suicidal ideation, and baseline suicidal attempts.

Additionally, after controlling for sex, age, HSS, living arrangement, depressive symptoms, alcohol-related problems, suicidal ideation, and suicidal attempts at baseline, the final results showed that students who reported baseline misuse of opioids (AOR, 3.39; 95% CI, 1.33-5.63) and nonmedical use of any prescription drug (AOR, 2.91; 95% CI, 1.26-3.71) were more likely to be involved in suicidal attempts 1 year later.

Mediating Effects of Depressive Symptoms on the Relationship Between NMUPD and Suicidality

First, univariable regression models were conducted to test the associations between the interaction items (between nonmedical use opioids/sedatives/any prescription drugs and depressive symptoms) and suicidality; these interaction items were not statistically significant in the models (Table 3).

Next, Table 4 shows that the standardized indirect effects of baseline opioids misuse on subsequent suicidal ideation and attempts through depressive symptoms were significant (suicidal ideation at follow-up: standardized β estimate = 0.020; 95% CI, 0.010-0.030; suicidal attempts at follow-up: standardized β estimate = 0.009, 95% CI, 0.004-0.015), indicating that depressive symptoms partially mediated the associations between baseline nonmedical use of opioids and suicidal ideation and attempts at follow-up. Additionally, there was a significant standardized indirect effect of baseline sedatives misuse on subsequent suicidal ideation through depressive symptoms (standardized β estimate = 0.016; 95% CI, 0.005-0.026), suggesting that the association between baseline sedatives misuse and subsequent suicidal ideation was partially mediated by depressive symptoms.

Discussion

To our knowledge, this is the first large-scale, prospective longitudinal study to investigate the relationships between NMUPD, suicidal ideation, and suicidal attempts and to test the potential mediating effects of depressive symptoms.

Consistent with previous studies,16,23,34,35 our univariable analyses first reported that female students were at a higher risk of suicidal ideation and that those having alcohol-related problems were more likely to be involved in suicidality. For the adolescents who reported previous suicidal attempts, the likelihood of future attempts was greatly increased. These findings are useful for identifying adolescents who may be at risk of suicidality, and we should focus on high-risk students who present with the adverse characteristics mentioned above.

As depressive symptoms, alcohol-related problems, and prior suicidality are well-known risk factors for suicidal behavior,17,23,36 and as the associations of these factors with suicidality were significant in our univariable analyses, the final multivariable logistic regression models found that baseline opioids misuse, sedatives misuse, and nonmedical use of any prescription drug were significantly associated with suicidal ideation at follow-up after controlling for sociodemographic information (including sex, age, HSS, and living arrangement), baseline depressive symptoms, baseline alcohol-related problems, baseline suicidal ideation, and baseline suicidal attempts. Similarly, after controlling for sex, age, HSS, living arrangement, depressive symptoms, alcohol-related problems, suicidal ideation, and suicidal attempts at baseline, the relationships between baseline opioids misuse, baseline any prescription drug misuse, and subsequent suicidal attempts were statistically significant in our final logistic regression models. These results were consistent with some previous cross-sectional findings, suggesting that there are significant relationships between the misuse of any prescription drug and suicidal ideation12,19,37 and suicidal attempts.13,38-40 Therefore, NMUPD is an important risk factor for suicidal ideation and suicidal attempts in adolescence. This finding may be related to several possible explanations: (1) drug use (including NMUPD) may impair judgment through its intoxicating effects, (2) it may have acute and chronic mood-altering effects, or (3) it may facilitate the route to suicidal behavior through inhibitions loss, impulsiveness, and/or aggression.41,42 These findings suggest the following interventions to prevent adolescent suicidal behavior: (1) establishing a proper surveillance system to manage and oversee suicidality among adolescents, (2) promoting resilience among current nonmedical users of prescription drugs, and (3) strengthening regulations to limit sales of prescription drugs containing opioids, sedatives, or stimulants to adolescents.

In addition, our study also showed that the effect of baseline nonmedical use of opioids on suicidal ideation or suicidal attempts 1 year later was partially mediated by depressive symptoms. Moreover, the association between baseline sedatives misuse and subsequent suicidal ideation was also partially mediated by depressive symptoms. The neurobiological similarities between NMUPD and depressive disorder might be a possible explanation for the mediating effects of depressive symptoms.20-22 Furthermore, it has been well established that most suicidal behavior is associated with depressive symptoms among culturally diverse teens,16,18,40,43 and previous studies15,19 have also reported a significant association between NMUPD and depressive symptoms. Moreover, some prior studies have demonstrated mediating effects between adverse child experiences and suicide attempts through depressive symptoms23 and between sleep problems and suicidal attempts through depressive symptoms.44 To our knowledge, few studies have discussed the mediating effects of depressive symptoms on the relationship between NMUPD and suicidality. Our findings highlighted the importance of depressive mood and NMUPD on suicidality among adolescents. Furthermore, an accessible and effective adolescent mental health service and a suicidality monitoring system may be essential components in reducing Chinese youth suicidality.

Several limitations of the study should be noted when interpreting the results. First, our study sample included only school students and did not include adolescents who had dropped out of school or were not present in school on the day the survey was administered; suicidality or depressive symptoms may be more common among adolescents who were absent. Second, we used a structured self-rating questionnaire to collect data; although self-reporting is a common and accepted method, we could not completely rule out the possibility of recall bias. Third, suicidal behavior was measured by single items. Although these items had face validity, the reliability of these measures is unknown. Future studies should use multiple items to measure suicidality. Fourth, we may lose some information by dichotomizing the variables of suicidal behavior and NMUPD. Fifth, although this is the first known Chinese study to collect prospective data to study the relationships between NMUPD and suicidality, the longitudinal sample was limited to 1 year.

Despite these limitations, the primary strengths of our study are that it provides the first known longitudinal evidence for the relationships between baseline NMUPD and subsequent suicidal ideation and attempts among Chinese adolescents and is the first to report the mediating effects of depressive symptoms on the above-mentioned relationships.

Conclusions

Our study identified that baseline nonmedical use of opioids was significantly associated with suicidal ideation and suicidal attempts at follow-up, baseline nonmedical use of sedatives was related to suicidal ideation at follow-up, and baseline nonmedical use of any prescription drug was also associated with subsequent suicidal ideation and suicidal attempts. Moreover, depressive symptoms significantly partially mediated those significant relationships. Suicide is a complex behavior, and further studies are needed to better understand the respective roles of these factors (ie, interactions and mediation). Based on the findings of our study, effective prevention and intervention programs should be established. First, educational campaigns directed at students are needed to improve their awareness of the serious consequences of NMUPD. Second, we should pay close attention to the adolescents who have depressive symptoms, are involved in NMUPD, or have suicidal thoughts or attempts. Third, a proper surveillance system needs to be developed by learning from the experiences of the NSDUH in the United States to control and oversee the NMUPD and suicidality among Chinese adolescents.

Back to top
Article Information

Corresponding Author: Ciyong Lu, MD, PhD (Chinese supervisor), Department of Medical Statistics and Epidemiology, School of Public Health, Sun Yat-sen University, No. 74 Zhongshan Rd 2, Guangzhou 510080, People’s Republic of China (luciyong@mail.sysu.edu.cn).

Correction: This article was corrected online for typographical errors September 19, 2016, and was corrected for an addition to author affiliations November 14, 2016.

Accepted for Publication: May 1, 2016.

Published Online: August 15, 2016. doi:10.1001/jamapediatrics.2016.1802

Author Contributions: Drs Guo and Lu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Guo and Xu contributed equally to this work, and Profs Lu and Zhang contributed equally to this work.

Study concept and design: Guo, Gao, Wu, Pan, Lu.

Acquisition, analysis, or interpretation of data: Guo, Xu, Deng, J. Huang, G. Huang, Zhang, Lu.

Drafting of the manuscript: Guo.

Critical revision of the manuscript for important intellectual content: Xu, Deng, J. Huang, G. Huang, Gao, Wu, Pan, Zhang, Lu.

Statistical analysis: Guo, Lu.

Obtained funding: Guo, Xu, Deng, J. Huang, G. Huang, Gao, Wu, Pan, Lu.

Administrative, technical, or material support: Guo, Xu, Deng, J. Huang, G. Huang, Gao, Wu, Pan, Lu.

Study supervision: Guo, Deng, Gao, Wu, Pan, Zhang, Lu.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by the Guangdong Food and Drug Administration.

Role of the Funder/Sponsor: The Guangdong Food and Drug Administration had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
He  ZX, ed.  Suicidology. Beijing, China: China Press of Traditional Chinese Medicine; 1997.
2.
Nock  MK, Borges  G, Bromet  EJ, Cha  CB, Kessler  RC, Lee  S.  Suicide and suicidal behavior.  Epidemiol Rev. 2008;30:133-154.PubMedGoogle ScholarCrossref
3.
Miranda  R, Scott  M, Hicks  R, Wilcox  HC, Harris Munfakh  JL, Shaffer  D.  Suicide attempt characteristics, diagnoses, and future attempts: comparing multiple attempters to single attempters and ideators.  J Am Acad Child Adolesc Psychiatry. 2008;47(1):32-40.PubMedGoogle ScholarCrossref
4.
Ruengorn  C, Sanichwankul  K, Niwatananun  W, Mahatnirunkul  S, Pumpaisalchai  W, Patumanond  J.  Incidence and risk factors of suicide reattempts within 1 year after psychiatric hospital discharge in mood disorder patients.  Clin Epidemiol. 2011;3:305-313.PubMedGoogle Scholar
5.
Zhang  J, Sun  L, Liu  Y, Zhang  J.  The change in suicide rates between 2002 and 2011 in China.  Suicide Life Threat Behav. 2014;44(5):560-568.PubMedGoogle ScholarCrossref
6.
Cui  S, Cheng  Y, Xu  Z, Chen  D, Wang  Y.  Peer relationships and suicide ideation and attempts among Chinese adolescents.  Child Care Health Dev. 2011;37(5):692-702.PubMedGoogle ScholarCrossref
7.
Liu  X, Tein  JY, Zhao  Z, Sandler  IN.  Suicidality and correlates among rural adolescents of China.  J Adolesc Health. 2005;37(6):443-451.PubMedGoogle ScholarCrossref
8.
World Health Organization. Preventing suicide: a global imperative. http://apps.who.int/iris/bitstream/10665/131056/1/9789241564779_eng.pdf. Accessed July 12, 2016.
9.
Falbo  T, Poston  DL.  How and why the one-child policy works in China.  Adv Popul. 1994;2:205-229.PubMedGoogle Scholar
10.
Moscicki  EK.  Epidemiology of completed and attempted suicide: toward a framework for prevention.  Clin Neurosci Res. 2001;1(5):310-323.Google ScholarCrossref
11.
Hedden  SL, Kennet  J, Lipari  R,  et al. Substance Abuse and Mental Health Services Administration. Behavioral health trends in the United States: results from the 2014 National Survey on Drug Use and Health. 2015. http://www.samhsa.gov/data/sites/default/files/NSDUH-FRR1-2014/NSDUH-FRR1-2014.pdf. Accessed July 12, 2016.
12.
Ford  JA, Perna  D.  Prescription drug misuse and suicidal ideation: Findings from the National Survey on Drug Use and Health.  Drug Alcohol Depend. 2015;157:192-196.PubMedGoogle ScholarCrossref
13.
Guo  L, Xu  Y, Deng  J,  et al.  Nonmedical use of prescription pain relievers among high school students in China: a multilevel analysis.  BMJ Open. 2015;5(7):e007569.PubMedGoogle ScholarCrossref
14.
Wang  H, Deng  J, Zhou  X,  et al.  The nonmedical use of prescription medicines among high school students: a cross-sectional study in Southern China.  Drug Alcohol Depend. 2014;141:9-15.PubMedGoogle ScholarCrossref
15.
Juan  W, Jian-Xiong  D, Lan  G,  et al.  Nonmedical use of psychoactive drugs in relation to suicide tendencies among Chinese adolescents.  Addict Behav. 2015;51:31-37.PubMedGoogle ScholarCrossref
16.
Dvorak  RD, Lamis  DA, Malone  PS.  Alcohol use, depressive symptoms, and impulsivity as risk factors for suicide proneness among college students.  J Affect Disord. 2013;149(1-3):326-334.PubMedGoogle ScholarCrossref
17.
Skarbø  T, Rosenvinge  JH, Holte  A.  Alcohol problems, mental disorder and mental health among suicide attempters 5-9 years after treatment by child and adolescent outpatient psychiatry.  Nord J Psychiatry. 2006;60(5):351-358.PubMedGoogle ScholarCrossref
18.
Nrugham  L, Holen  A, Sund  AM.  Associations between attempted suicide, violent life events, depressive symptoms, and resilience in adolescents and young adults.  J Nerv Ment Dis. 2010;198(2):131-136.PubMedGoogle ScholarCrossref
19.
Zullig  KJ, Divin  AL.  The association between nonmedical prescription drug use, depressive symptoms, and suicidality among college students.  Addict Behav. 2012;37(8):890-899.PubMedGoogle ScholarCrossref
20.
Kessler  RC, McGonagle  KA, Zhao  S,  et al.  Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey.  Arch Gen Psychiatry. 1994;51(1):8-19.PubMedGoogle ScholarCrossref
21.
Volkow  ND.  The reality of comorbidity: depression and drug abuse.  Biol Psychiatry. 2004;56(10):714-717.PubMedGoogle ScholarCrossref
22.
Shimamoto  A, Debold  JF, Holly  EN, Miczek  KA.  Blunted accumbal dopamine response to cocaine following chronic social stress in female rats: exploring a link between depression and drug abuse.  Psychopharmacology (Berl). 2011;218(1):271-279.PubMedGoogle ScholarCrossref
23.
Cluver  L, Orkin  M, Boyes  ME, Sherr  L.  Child and adolescent suicide attempts, suicidal behavior, and adverse childhood experiences in south africa: a prospective study.  J Adolesc Health. 2015;57(1):52-59.PubMedGoogle ScholarCrossref
24.
Miller  AB, Jenness  JL, Oppenheimer  CW, Gottleib  AL, Young  JF, Hankin  BL.  Childhood emotional maltreatment as a robust predictor of suicidal ideation: a 3-year multi-wave, prospective investigation  [published online March 31, 2016].  J Abnorm Child Psychol. doi:10.1007/s10802-016-0150-z.PubMedGoogle Scholar
25.
Kwok  SY, Shek  DT.  Socio-demographic correlates of suicidal ideation among Chinese adolescents in Hong Kong.  Int J Adolesc Med Health. 2008;20(4):463-472.PubMedGoogle Scholar
26.
Cheng  CP, Yen  CF, Ko  CH, Yen  JY.  Factor structure of the Center for Epidemiologic Studies Depression Scale in Taiwanese adolescents.  Compr Psychiatry. 2012;53(3):299-307.PubMedGoogle ScholarCrossref
27.
Lee  SW, Stewart  SM, Byrne  BM,  et al.  Factor structure of the Center for Epidemiological Studies Depression Scale in Hong Kong adolescents.  J Pers Assess. 2008;90(2):175-184.PubMedGoogle ScholarCrossref
28.
Radloff  LS.  The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults.  J Youth Adolesc. 1991;20(2):149-166.PubMedGoogle ScholarCrossref
29.
Firth  D.  Bias reduction of maximum likelihood estimates.  Biometrika. 1993;80:27-38.Google ScholarCrossref
30.
Wang  X.  Firth logistic regression for rare variant association tests.  Front Genet. 2014;5:187.PubMedGoogle Scholar
31.
Preacher  KJ, Hayes  AF.  Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models.  Behav Res Methods. 2008;40(3):879-891.PubMedGoogle ScholarCrossref
32.
Gunzler  D, Chen  T, Wu  P, Zhang  H.  Introduction to mediation analysis with structural equation modeling.  Shanghai Arch Psychiatry. 2013;25(6):390-394.PubMedGoogle Scholar
33.
Mackinnon  DP, Lockwood  CM, Williams  J.  Confidence limits for the indirect effect: distribution of the product and resampling methods.  Multivariate Behav Res. 2004;39(1):99.PubMedGoogle ScholarCrossref
34.
Campbell  G, Bruno  R, Darke  S,  et al.  Prevalence and correlates of suicidal thoughts and suicide attempts in people prescribed pharmaceutical opioids for chronic pain.  Clin J Pain. 2016;32(4):292-301.PubMedGoogle ScholarCrossref
35.
Wichstrøm  L.  Predictors of adolescent suicide attempts: a nationally representative longitudinal study of Norwegian adolescents.  J Am Acad Child Adolesc Psychiatry. 2000;39(5):603-610.PubMedGoogle ScholarCrossref
36.
Lewinsohn  PM, Rohde  P, Seeley  JR.  Psychosocial risk factors for future adolescent suicide attempts.  J Consult Clin Psychol. 1994;62(2):297-305.PubMedGoogle ScholarCrossref
37.
Rhoades  H, Winetrobe  H, Rice  E.  Prescription drug misuse among homeless youth.  Drug Alcohol Depend. 2014;138:229-233.PubMedGoogle ScholarCrossref
38.
Crumley  FE.  Substance abuse and adolescent suicidal behavior.  JAMA. 1990;263(22):3051-3056.PubMedGoogle ScholarCrossref
39.
Noori  R, Rafiey  H, Azizabadi-Farahani  M,  et al.  Risk factors of suicidal ideation and attempt in women with drug user spouses.  J Chin Med Assoc. 2013;76(11):648-652.PubMedGoogle ScholarCrossref
40.
Brockie  TN, Dana-Sacco  G, Wallen  GR, Wilcox  HC, Campbell  JC.  The relationship of adverse childhood experiences to PTSD, depression, poly-drug use and suicide attempt in reservation-based native american adolescents and young adults.  Am J Community Psychol. 2015;55(3-4):411-421.PubMedGoogle ScholarCrossref
41.
Bukstein  OG, Brent  DA, Perper  JA,  et al.  Risk factors for completed suicide among adolescents with a lifetime history of substance abuse: a case-control study.  Acta Psychiatr Scand. 1993;88(6):403-408.PubMedGoogle ScholarCrossref
42.
Brodsky  BS, Malone  KM, Ellis  SP, Dulit  RA, Mann  JJ.  Characteristics of borderline personality disorder associated with suicidal behavior.  Am J Psychiatry. 1997;154(12):1715-1719.PubMedGoogle ScholarCrossref
43.
Liu  X, Tein  JY, Zhao  Z, Sandler  IN.  Suicidality and correlates among rural adolescents of China.  J Adolesc Health. 2005;37(6):443-451.PubMedGoogle ScholarCrossref
44.
Wong  MM, Brower  KJ.  The prospective relationship between sleep problems and suicidal behavior in the National Longitudinal Study of Adolescent Health.  J Psychiatr Res. 2012;46(7):953-959.PubMedGoogle ScholarCrossref
×