OBJECTIVE: Chronic conditions may be associated with suicide risk. This study aimed to specify the extent to which youth chronic conditions are at risk for suicidality and self-harm.
METHODS: Logistic regression was used to estimate odds of self-harm, suicidal ideation, and suicide attempts in 10- to 19-year-olds with and without chronic physical and/or mental health conditions.
RESULTS: Independent of race, socioeconomic status, absent parent, special education status, substance use, and emotional distress, youth with co-occurring chronic physical and mental conditions (n = 4099) had significantly higher odds of self-harm (odds ratio [OR]: 2.5 [99% confidence interval (CI): 2.3–2.8), suicidal ideation (OR: 2.5 [99% CI: 2.3–2.8), and suicide attempts (OR: 3.5 [99% CI: 3.1–3.9]) than healthy peers (n = 106 967), as did those with chronic mental conditions alone (n = 8752). Youth with chronic physical conditions alone (n = 12 554) were at slightly elevated risk for all 3 outcomes. Findings were similar among male and female youth, with a risk gradient by grade.
CONCLUSIONS: Chronic physical conditions are associated with a slightly elevated risk for self-harm, suicidal thinking, and attempted suicide; chronic mental conditions are associated with an increased risk for all 3 outcomes. Co-occurring chronic physical and mental conditions are associated with an increased risk for self-harm and suicidal ideation that is similar to the risk in chronic mental conditions and with an attempted suicide risk in excess of that predicted by the chronic mental health conditions alone. Preventive interventions for these youth should be developed and evaluated.
- chronic conditions
- psychiatric issues in primary care
- psychological impact
- social-emotional problems
WHAT'S KNOWN ON THIS SUBJECT:
Although youth with chronic health conditions are at higher risk than healthy peers for having emotional or behavioral problems, the risk for self-harm and suicide in these youth has been unclear.
WHAT THIS STUDY ADDS:
Other than slightly higher odds of attempted suicide, there is little difference in suicidality between youth with co-occurring physical and mental conditions and those with mental conditions alone, whereas youth with physical conditions have slightly higher risk than healthy peers.
Children and youth with special health care needs (CYSHCN) “have or are at risk for a chronic physical, developmental, behavioral, or emotional condition and who also require health and related services of a type or amount beyond that required by children generally.”1,2 CYSHCN thus represent a subset of all children with chronic health conditions. Approximately 13 million US children are CYSHCN, with rates varying from 12.8% to 19.3% in national surveys of caregivers in the past decade.3,–,5 Similarly, anxious or depressed feelings and emotional-behavioral problems are more common in CYSHCN than in those without special health care needs.4,6,7
Little is yet known about suicide- and self-harm–related morbidity and mortality in CYSHCN or in the broader population of children with chronic health conditions, although some investigators have found older adolescents' overall “health status” to be a risk factor for suicidality.8 To clarify the extent to which having chronic health conditions might contribute to suicidal thinking, behavior, and purposeful self-injury, we used a large population-based survey of youth. We hypothesized that the risk for these outcomes remains elevated in children with chronic conditions even after controlling for covariate risks; that this risk is highest among youth with co-occurring chronic physical and mental health conditions; and that the risk for suicide in children with chronic health conditions occurs in a developmental context, varying by gender and across age levels.
The 2007 Minnesota Student Survey (MSS) data were used for the current study. Minnesota's Departments of Health (MDH), Human Services, Education, and Public Safety have administered this voluntary and anonymous survey every 3 years since 1989 to students in sixth, ninth, and 12th grades statewide. Surveys were administered in schools during classes, usually homeroom; no individually linked information was collected. Passive parental consent was allowed in most districts (ie, opt-out); active parental consent (ie, opt-in) was used when required by independent school district research protocols. This study included data from all surveyed students who were not institutionalized in correctional facilities or alternative learning centers at the time of the survey.
Specific details of sample characteristics, procedures, instrumentation, psychometrics, consent, and data processing are described elsewhere.9 Ninety-one percent of state public schools participated in the 2007 MSS, along with private, tribal, and charter schools; 72% of eligible public school students participated. Analysts at MDH deemed responses to the MSS reliable and valid9; ∼3% of surveys were eliminated from the final data set by MDH because of absent gender data, highly inconsistent responses, or response patterns suggesting exaggeration or falsification. Not every student answered every question, and ∼4% of participants' responses were missing data that were relevant to this study; these were excluded from subsequent analysis by the investigators. The final sample used in this study thus consisted of data from 136549 students who ranged in age from 10 to 19 years, inclusive, at the time of data collection. This sample had a mean age of 14.4 years, with a trimodal distribution at 12, 15, and 18 years.
Type of Chronic Condition
The independent variable self-reported chronic health condition was composed of 4 categories. Youth who had a chronic physical health condition answered, “Yes,” to, “Do you have a physical condition or problem that has lasted at least 12 months?” Youth who had a chronic mental health condition answered, “Yes,” to, “Do you have a mental or emotional health problem that has lasted at least 12 months?” Youth who had co-occurring chronic physical and mental health conditions answered, “Yes,” to both of these questions; youth without any chronic conditions answered, “No,” to both.
Self-harm and Suicide Risk
The 3 outcome variables were self-reported history of suicidal thinking, suicide attempt, or purposeful self-harm (ie, self-injurious behavior). Youth who had a history of self-harm answered, “Yes, during the last year,” or, “Yes, more than a year ago,” to, “Have you ever hurt yourself on purpose (‘cutting,’ burns, bruises)?” Youth with a history of suicidal ideation answered, “Yes, during the last year,” or, “Yes, more than a year ago,” to, “Have you thought about killing yourself?” Youth with a history of attempted suicide answered, “Yes, during the last year,” or, “Yes, more than a year ago,” to, “Have you tried to kill yourself?”
The 12 emotional-behavioral items in the MSS were taken directly from the short version of the Pediatric Symptom Checklist for internalizing and externalizing symptoms10 and from the K6 screening scales for internalizing symptoms.11 Factor analysis revealed that responses to 5 of the 12 items were highly correlated (Cronbach's α = .86): “I am often unhappy, depressed, or tearful” (0 = agree; 4 = disagree); “During the last 30 days, have you felt you were under any stress or pressure?” (0 = yes, almost more than I could take; 5 = no); “During the last 30 days, have you felt sad?” (0 = all the time; 5 = none of the time); “During the last 30 days, have you felt so discouraged or hopeless that you wondered if anything was worthwhile?” (0 = not at all; 5 = extremely so, to the point I have just about given up); and “During the last 30 days, have you felt nervous, worried, or upset?” (0 = none of the time; 5 = all of the time). Items were summed to create a continuous “emotional well-being” variable with a minimum score of 0 (low emotional well-being) and a maximum score of 24 (high emotional well-being).
Additional covariates included demographics (categorical variables of gender, grade, and race/ethnicity); family structure (categorized as “mother and father both present” or “mother and/or father absent”); socioeconomic status (dichotomized as receipt of free/reduced-price lunch); special education status (dichotomized as ever having had an Individualized Education Program; not asked of sixth-graders); and active substance use (categorized as use in the past 30 days of tobacco, alcohol, and/or marijuana). All of these have been linked in previous studies to having a chronic condition and/or suicide risk.3,8,12,–,19
χ2 and 2-tailed Student's t tests were used to compare categorical and continuous variables, respectively. A logistic regression model was created to estimate the odds of self-harm, suicidal ideation, and attempted suicide for youth in the 4 categories of chronic health condition (none, physical only, mental only, or both), controlling for the described covariates, stratified by gender and grade level. Because the “emotional well-being” variable was somewhat skewed, a separate posthoc analysis modeled this variable in categorical quartiles, with similar results (data not shown).
Given the large sample size, a significance level of .01 (99% confidence interval [CI]) was arbitrarily set for tests of the primary hypotheses. All analyses were performed with SAS 9.1.3 for Windows (SAS Institute Inc, Cary, NC). The University of Minnesota institutional review board approved all data analysis protocols.
Children without chronic conditions made up 81% (n = 106 967) of the overall sample. Of children with chronic conditions (19%; n = 25 405), 9.5% had a chronic physical condition (n = 12 554), 6.6% had a chronic mental health condition (n = 8752), and 3.0% had both (n = 4099) (Table 1). Overall, 17.3% of the overall sample population had a history of self-harm (n = 22 662), 22.4% had a history of suicidal ideation (n = 29 120), and 5.9% had a history of attempted suicide (n = 7666). The sample was predominantly white (75%), in line with state census data. Additional characteristics of the sample are shown in Table 1.
Rates of self-harm and suicide-related outcomes varied among the 4 groups; youth with co-occurring chronic physical and mental health conditions had the highest rates (Table 2), albeit only somewhat higher than the rates for youth with mental conditions alone. Levels of emotional well-being (range: 4–24; overall mean: 17.80 ± 0.01; median: 19.00) showed a significant decremented gradient according to the type of health condition, such that youth with no chronic conditions had a “high” well-being with a mean score of 18.5 ± 0.1, youth with chronic physical conditions alone had a mean score of 17.2 ± 0.4, youth with chronic mental health conditions alone had a mean score of 13.1 ± 0.4, and youth with both types of conditions had “low” well-being with a mean score of 12.4 ± 0.6 (P < .0001).
The differences in rates of self-harm, suicidal ideation, and attempted suicide were maintained after adjustment for covariates (Table 3). Compared with peers without chronic conditions, youth with chronic physical conditions alone had a slightly higher but significant risk for self-harm (adjusted odds ratio [OR]: 1.21 [99% CI: 1.12–1.30]), suicidal ideation (OR: 1.22 [99% CI: 1.14–1.31]), and attempted suicide (OR: 1.24 [99% CI: 1.09–1.40]; Table 3). Youth with chronic mental health conditions alone also had an increased risk for self-harm (OR: 2.55 [99% CI: 2.37–2.75]), suicidal ideation (OR: 2.67 [99% CI: 2.47–2.89]), and attempted suicide (OR: 2.97 [99% CI: 2.69–3.27]). The presence of co-occurring chronic physical and mental health conditions placed youth at high risk for self-harm (OR: 2.54 [99% CI: 2.28–2.82]) and suicidal ideation (OR: 2.54 [99% CI: 2.27–2.84]) and at a substantially elevated risk for attempted suicide (OR: 3.48 [99% CI: 3.07–3.94]).
Male and female youth did not differ overall in any of the adjusted analyses, except for increased odds of self-harm in female youth with chronic mental health conditions (male adjusted OR: 2.12 [99% CI: 1.87–2.40]; female adjusted OR: 2.75 [99% CI: 2.49–3.03]). The odds of self-harm, suicidal ideation, and attempted suicide generally increased with increasing grade level for youth with any type of chronic condition, such that 12th-graders with co-occurring chronic physical and mental health conditions were at very elevated risk for attempted suicide, regardless of gender (male adjusted OR: 4.02 [99% CI: 2.84–5.70]; female adjusted OR: 5.01 [99% CI: 3.91–6.42]).
Suicidal ideation may be highly prevalent among youth regardless of risk and protective factors but seems to peak at approximately age 14 and then decline.19 Self-injurious behaviors and suicidal ideation are more common in girls than in boys, and certain racial and ethnic groups of youth seem to be at higher risk as well (independent of socioeconomic status).19 In adults with known chronic health problems, previous research showed that suicidal ideation is linked to poor physical functioning,20 chronic pain,21 multiple sclerosis,22 chronic dermatologic conditions,23 and surviving childhood cancer.24 To date, it has been less clear whether similar associations hold true for youth with chronic health conditions.
Our study reveals that chronic conditions are, in fact, associated with an elevated risk for self-harm, suicidal thoughts, and suicidal behaviors in young people in Minnesota. Overall, 5.9% of our sample, regardless of risk factors, reported ever attempting suicide; although this proportion may seem high, it is in line with estimates from other contemporary studies,8 as are the overall rates of suicidal ideation19 and self-harm25 that we report (Table 1).
Even after controlling for risk factors that are common both to having a chronic condition and to suicidality, youth with chronic physical conditions alone have an ∼20% increase over their healthy peers in their odds of self-harm, suicidal thinking, and attempted suicide; youth with chronic mental health conditions alone have a two- to threefold increase for each of these 3 outcomes. Significantly, youth with co-occurring chronic mental and physical health conditions have 2.5 to 3.5 times the odds of such outcomes. The risks for self-harm and suicidal ideation seem similarly elevated for those with mental health conditions alone and for those with both physical and mental health conditions, yet the risk for actually attempting suicide seems quite elevated for youth with co-occurring chronic physical and mental health conditions, even beyond that expected of youth with mental health conditions alone. For all outcomes, the odds of suicidality seem to increase across age levels from sixth through 12th grades, implicating developmental factors. Female youths have a small increase in these odds that may not be statistically significant.
Our results suggest that suicide and self-harm may be more prevalent than previously thought in children with chronic health conditions, extending the findings of previous studies that revealed links between health status and emotional-behavioral problems in this population. Our results are especially intriguing given recent research highlighting the longitudinal associations between chronic physical problems (eg, asthma, recurrent infections) with chronic mental health problems (eg, anxious, depressive, or disruptive disorders) in children and adolescents.26,27 The links between co-occurring chronic conditions and severe psychopathology could be unidirectional or bidirectional or could vary according to the contexts in which an individual experiences a given chronic condition28; for example, an adolescent with conduct disorder may engage in risk-taking behaviors and have recurrent sexually transmitted infections, whereas a preteen with diabetes may have adjustment problems that lead to anxiety and major depressive disorder. An important contribution of this study is that, whereas most youth with chronic conditions enjoy normal levels of emotional well-being and do not have suicidal thoughts or self-injurious behaviors, there may be a developmental trajectory among certain children with chronic conditions, not yet fully elucidated, that contributes to excess emotional distress and suicide risk.
Because this survey was anonymous and not linked to individuals, these cross-sectional data preclude inferences of causality. Our method was limited because we could not ask students about all putative negative life events, risk factors, or covariates; neither could we control for recall bias (suicidal youth may more likely perceive that they have chronic conditions). Our results should be generalized with caution, given the restriction of the sample to the population of school-aged youth in an upper Midwestern US state. Finally, an unknown proportion of youth with severe impairments or disabilities might have been excluded from this study because of absence from school.
Because the presence or absence of a chronic condition was self-reported, our data may systematically overestimate or underestimate the prevalence of chronic conditions. This is attributable in part to the unknown reliability of youth self-report for chronic physical and mental health conditions. Our method did not specify which youth also had elevated health care needs, as is done in caregiver-reported data sets such as the National Survey of Children's Health and the National Survey of Children with Special Health Care Needs. Those surveys used the CSHCN Screener,29 which requires endorsement of at least 1 of 5 health care needs above and beyond those expected for children in general; these needs must be attributed to a medical, behavioral, or other health condition that has lasted (or is expected to last) at least 12 months. We used only this final criterion of the CSHCN Screener to define our population, creating a broad and perhaps overly inclusive definition of chronic health problems,30 yet although not all youth who perceive themselves as having chronic health conditions also have “special health care needs,” it is interesting to note that the overall prevalence of self-reported chronic health conditions in our sample (19%) was similar to the prevalence of CYSHCN in national surveys that used caregiver-reported criteria.5
This study has numerous strengths that should be highlighted despite these caveats. Many of the topics addressed by self-report, in addition to chronic health problems, were anonymity-sensitive and perhaps would have been difficult for youth to disclose under other circumstances. This may be especially true of the emotional distress and self-harm items. In fact, youth self-report has been shown in numerous studies to be more accurate than caregiver report for internalizing symptoms and suicidal behaviors.31,–,33 Children as young as 6 can accurately report suicidality, and parents of younger adolescents may actually be more likely than those of older adolescents to be unaware of their children's suicide attempt(s).31 Another strength lies in our “dimensional” approach, which created 4 categories of self-reported health status along a “mental–physical” axis. Assessing these aspects of individuals' experience of their health has been suggested as worthy of additional study34; for example, youth who have conditions that are classified as either “visible” or “invisible” (eg, epilepsy, diabetes, asthma, cerebral palsy, arthritis, scoliosis) have similar levels of emotional distress, worries, and other socioemotional problems.7 Youth with chronic health conditions thus have needs that transcend specific diseases; indeed, those who need elevated health services are more likely than their healthy peers to have mental health needs that go unmet, and racial/ethnic and economic disparities further widen such gaps.35 As such, children with chronic conditions seem best served by a dimensional paradigm instead of a disease-specific (diagnostic) paradigm.36,–,39
Minnesota youth with self-reported chronic, co-occurring mental and physical health conditions during late childhood and early, middle, and late adolescence are at two- to fivefold higher risk for attempted suicide, compared with their generally healthy peers; those in the highest grades are at the most risk, and there is little difference in this risk between boys and girls. Furthermore, having such co-occurring chronic mental and physical conditions does not seem to alter the risk for self-harm or suicidal ideation beyond the risk of having a mental health condition alone.
Preventive interventions for chronically ill youth should be developed and evaluated, perhaps targeting multilevel risk and protective factors among youth with co-occurring mental and physical conditions. Complementary efforts, such as surveillance for emotional distress and suicidality, might be targeted more generally at children who have multiple risk factors for suicide along with ≥1 chronic health problem. Educational strategies might begin at the practitioner, caregiver, and teacher levels to increase awareness of the risk for suicide in some children with chronic conditions.
Additional studies, ideally multiple-informant and longitudinal in design, will help in understanding the links between chronic health conditions, socioemotional functioning, psychopathology, and self-destructive behavior. Such work could extend our findings and elucidate developmental pathways, effect mediators and moderators, and help in the development of preventive interventions that target the heightened risk for suicidality among certain children with chronic health conditions who seem, for 1 reason or another, to be more vulnerable to negative outcomes. In the meantime, clinicians and therapists should inquire—sensitively but frequently—about emotional well-being and suicidality in their young patients with chronic conditions while fostering resiliency and hope.40,41
We thank the peer reviewers for useful insights; those at the Minnesota Departments of Health, Human Services, Education, and Public Safety who took part in administering the 2007 MSS; Allison Anfinson, NCLB-Safe and Health Learners at the MN Department of Education, for facilitating distribution of the MSS data set; and the students and teachers who took part in the survey.
- Accepted January 4, 2010.
- Address correspondence to Andrew J. Barnes, MD, MPH, University of Minnesota, Department of Pediatrics, Division of Academic General Pediatrics, 717 Delaware St SE, Suite 353-01, Minneapolis, MN 55414. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- CYSHCN =
- children and youth with special health care needs •
- MSS =
- Minnesota Student Survey •
- MDH =
- Minnesota Department of Health •
- CI =
- confidence interval •
- OR =
- odds ratio
- McPherson M,
- Arango P,
- Fox H,
- et al
- Sneed RC,
- May WL,
- Stencel CS
- Newacheck PW,
- Strickland B,
- Shonkoff JP,
- et al
- 4.↵US Department of Health and Human Services, Health Resources Administration, Maternal and Child Health Bureau. The National Survey of Children with Special Health Care Needs Chartbook 2005–2006. Rockville, MD: US Department of Health and Human Services; 2007
- Borowsky IW,
- Ireland M,
- Resnick MD
- 9.↵Minnesota Student Survey Interagency Team. 2007 Minnesota Student Survey statewide tables. Roseville, MN: Minnesota Department of Education; Fall2007. Available at: http://education.state.mn.us/mdeprod/groups/SafeHealthy/documents/Report/033569.pdf. Accessed May 30, 2009
- Newacheck PW,
- Kim SE,
- Blumberg SJ,
- Rising JP
- Recklitis CJ,
- Lockwood RA,
- Rothwell MA,
- Diller LR
- Whitlock J,
- Eckenrode J,
- Silverman D
- Porges SW,
- Matthews KA,
- Pauls DL
- Perrin EC,
- Newacheck P,
- Pless IB,
- et al
- 36.↵CAHMI—Child and Adolescent Health Measurement Initiative. Fact Sheet: Who Are Children With Special Health Care Needs?Portland, OR: Data Resource Center for Child and Adolescent Health; 2007
- 37.↵American Academy of Pediatrics: Committee on Child Abuse and Neglect and Committee on Children With Disabilities. Assessment of maltreatment of children with disabilities. Pediatrics. 2001;108(2):508–512
- 38.↵American Academy of Pediatrics: Committee on Children With Disabilities and Committee on Psychosocial Aspects of Child and Family Health. Psychosocial risks of chronic health conditions in childhood and adolescence. Pediatrics. 1993;92(6):876–878
- 39.↵US Department of Health and Human Services. Healthy People 2020. Available at: www.healthypeople.gov/hp2020. Accessed April 15, 2009
Redefining the Workforce: Women Take Over : Women currently outnumber men on the US workplace. According to an article in The New York Times (Rampell C, February 5, 2010), a possible reason for the turnaround is the current economy where men are losing their jobs faster than women, a situation referred to as a “man-cession”.
Noted by JFL, MD
- Copyright © 2010 by the American Academy of Pediatrics