Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Jul 7.
Published in final edited form as: Depress Anxiety. 2013 Jun 26;30(10):997–1004. doi: 10.1002/da.22135

DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF CONSISTENT AND INCONSISTENT LONGITUDINAL REPORTERS OF LIFETIME SUICIDE ATTEMPTS IN ADOLESCENCE THROUGH YOUNG ADULTHOOD

Shelley R Hart 1, Rashelle J Musci 1, Nicholas Ialongo 1, Elizabeth D Ballard 2, Holly C Wilcox 1,2,*
PMCID: PMC4083681  NIHMSID: NIHMS581425  PMID: 23804209

Abstract

Background

Within the context of the recent release of the 2012 National Suicide Prevention Strategy, and as the third leading cause of death for individuals 10- to 24-years-old, suicide prevention is a national priority. A consistently reported and robust risk factor for suicide is a prior suicide attempt; however few studies have investigated the consistency of self-reported lifetime suicide attempts. The goal of this study is to describe the prevalence and characteristics of inconsistent reporting of suicide attempt in a longitudinal cohort of participants annually assessed in 12 waves of data collected from middle school (age 12) to early adulthood (age 22).

Methods

Among this cohort (n = 678), we compared those who consistently, inconsistently, and never reported a suicide attempt according to demographic and clinical variables.

Results

Almost 90% (88.5%) of our sample inconsistently reported a lifetime suicide attempt. Consistent and inconsistent reporters of lifetime suicide attempt did not differ on demographic or clinical variables with the exception of higher rates of lifetime suicidal ideation among consistent reporters (P < .001). Significant clinical differences were evident between inconsistent reporters and nonattempters.

Conclusions

Some level of inconsistent reporting of suicide attempt is inevitable when schools or health care systems systematically screen for suicide risk in adolescents. Inconsistent and consistent reporters of suicide attempt differ on few demographic or clinical variables; further prospective research should investigate the reasons for inconsistent reporting as well as the validity and stability of reporting in predicting future suicidal behavior.

Keywords: suicide, attempted, reliability, risk factors, longitudinal studies, research design

Suicide is the third leading cause of death among individuals 10- to 24-years-old in the United States (CDC),[1] with recent increases in deaths by suicide in this age group through 2010.[2, 3] Across all age groups, suicide has surpassed motor vehicle accidents as the leading cause of injury mortality in 2009, with a 15% increase from rates in 2000.[4] As such, suicide prevention is a national priority. Improving the use of data for surveillance, research, and evaluation is one goal of the 2012 National Strategy for Suicide Prevention (NSSP).[5]

Suicide attempt has consistently been shown to be associated with future suicidal behavior.[6] Questions assessing suicide attempt are included on suicide screens and suicide risk assessments (e.g., Columbia Suicide Severity Rating Scale [C-SSRS])[7] for identification and case finding, as well on national surveillance surveys of adolescents and adults (Youth Risk Behavior Survey [YRBS], National Survey on Alcohol and Related Conditions [NESARC])[8, 9] to inform policy, monitor changes, and facilitate in the allocation of resources. However, as suicide attempts are typically investigated using one retrospective, self-report lifetime item, the reliability and validity of these data have been called into question.[10] In adult samples, up to half of respondents who report suicidal ideation later deny having those thoughts.[11, 12] In the only published study examining consistency of suicide attempt reporting with adolescents and young adults, approximately one third of the sample was shown to deny an attempt 3–4 years after initially reporting it,[10] a trend also observed in adults.[13, 14] Although the literature is sparse, the studies with adult samples demonstrate few overall differences between consistent and inconsistent reporters of suicide attempt.[13, 14] In contrast, in adolescent and young adult samples, consistent reporters of suicide attempt are more likely to be females and younger, and to have depressive and somatoform disorder diagnoses and more co-morbidity than those who inconsistently report a suicide attempt.[10]

It would be anticipated that as individuals report a lifetime suicide attempt, they would permanently shift to a different category (i.e., having a history of attempt). Therefore, as more individuals report new attempts over time (incidence) without a large number leaving the prevalence estimates (e.g., due to death), it would be anticipated that the rate of lifetime prevalence should increase. Nevertheless, related diagnoses, such as anxiety and depression, have demonstrated declines in lifetime prevalence rates, calling into question the utility of the concept of lifetime prevalence.[15] Although suicide attempt self-reporting over time has not been similarly evaluated, it stands to reason that lifetime prevalence of suicide attempt is prone to a similar phenomenon that warrants further investigation.

CURRENT STUDY

Unfortunately, longitudinal studies examining consistent suicide attempt reporting over a decade and across developmental periods do not exist in the literature. This gap leaves unanswered questions, such as the prevalence of inconsistent reporting and whether sociodemographic and clinical differences exist between consistent and inconsistent reporters of suicide attempt. In the current study, we examined these issues using annual lifetime suicide attempt reporting over a 12-year period within a community-based, urban sample recruited upon entry to first grade.

MATERIALS AND METHODS

PARTICIPANTS

Data were drawn from a randomized controlled trial and longitudinal follow-up evaluating the effectiveness of two first-grade interventions (i.e., family- or classroom-based) aimed at reducing risk behaviors and improving long-term mental health and psychosocial outcomes (n = 678). For the current analyses, annual in-person interviews were administered from sixth grade (mean age = 12 years) through early adulthood (mean age = 22 years). The interviews averaged 90 min in duration, with an annual incentive ranging from $10 during grade school to $100 for each young adult survey. Written informed consent was obtained for each participant and the study was approved by the Johns Hopkins School of Public Health Institutional Review Board.[16]

ANALYTIC STRATEGY

In order to explore differences in longitudinal self-report of suicidal behaviors, three groups were identified, those who (1) consistently, (2) inconsistently, and (3) never reported a lifetime suicide attempt. To examine stability of reporting, two conceptualizations of lifetime suicide attempt reporting were explored. These conceptualizations were chosen to capitalize on the longitudinal study design and to maximize power and sample size.[17]

  • One-year follow-up (1YR): Investigated responses for the year following initial reported suicide attempt (e.g., an individual would be considered inconsistent if he answered “yes” to the lifetime attempt item in eighth grade followed by “no” in ninth grade). Using the year following first report (regardless of when report occurred) allowed us to examine all individuals who reported an attempt and how they respond 1 year later; that is, all reporters shared the same follow-up time frame.

  • At any time following an initial report (EVER): Investigated responses each year after initial reported attempt (e.g., an individual would be considered inconsistent if she answered “yes” to the lifetime attempt item in sixth grade followed by “no” at any of the data collection waves through young adulthood). This conceptualization resulted in different follow-up periods, in that each reporter was investigated following their first-reported attempt, regardless of when that attempt was reported (i.e., the first attempt could have occurred during young adulthood leaving an individual with only one follow-up wave to report an attempt vs. reporting an attempt in middle school with 10 follow-up waves).

Comparisons were conducted among the groups using a series of analyses of variance (ANOVA) for each conceptualization (i.e., 1YR, EVER). A Bonferroni correction was used to adjust for the large number of tests, resulting in a statistically significant P-value of ≤.004 (i.e., .05/12). Additional comparisons were made between the attempter groups to see if characteristics of the attempt or of the attempter at the time of the attempt distinguished the consistent from the inconsistent reporters (e.g., age at first suicide attempt, plan, medical care for an attempt). After Bonferroni correction, a statistically significant P-value was ≤.01 (i.e., .05/5).

Sociodemographic and clinical characteristics were chosen for comparisons that have demonstrated in community or clinical samples statistically significant differences between adolescent and young adult suicide attempters and nonattempters. A variety of demographic (i.e., gender,[18] race, and socioeconomic status[19]); and clinical variables (depression[20] and anxiety[21] symptoms, suicidal ideation,[22] Major Depressive Disorder (MDD),[23] and Drug and Alcohol Dependence Disorders[24, 25]) were assessed prospectively during childhood, adolescence, and young adulthood. Characteristics of the suicide attempt such as age at the time of first suicide attempt, the presence of suicide plans, and medical care for suicide attempt(s) might also distinguish inconsistent and consistent attempters.[22, 26, 27] All analyses were conducted using SPSS version 19.0.

Missing Data

Participation rates at each of the postintervention yearly assessments averaged 80%. Those who did not consent, either refused to participate, could not be located, failed to respond to our consent requests, or had died. In order to address missing data associated with attrition, we allowed for list-wise deletion as is the default and an acceptable strategy if the reduction in power that results is not detrimental to the results.[28]

MATERIALS

Demographic Information

Race and gender were self-reported. Free and reduced meal status—a proxy for income—was collected from elementary school records.

Suicidal Ideation and Behaviors

Questions from the Diagnostic Interview Schedule for Children (C-DISC-IV)[29] were asked each year beginning in sixth grade (mean age 12 years). Lifetime suicide attempt was measured by the question: “Have you ever, in your whole life, tried to kill yourself, or made a suicide attempt?” In addition, several questions were asked regarding suicidal ideation (“In the last year did you think seriously about killing yourself?,” “Did you think about killing yourself many times in the last year?”), which were combined and averaged over the 12 years to measure lifetime suicidal ideation. Finally, other suicide-related behaviors, such as a suicide plan and medical care for an attempt were measured for attempters only (i.e., inconsistent or consistent reporters), using the following questions: “In the last year did you have a plan for exactly how you would kill yourself?” (plan) and “Did you go see a doctor, go to an emergency room, or go into a hospital because of trying to kill yourself?” (medical care). Age at the time of each suicide attempt, whether they had also reported ideation, and number of suicide attempts was also recorded. Due to budget constraints in the 11th and 12th grades, all individuals were not asked the suicide questions in the DISC MDD module. Therefore, these data points were used only if an individual reported a first attempt.

Depressive and Anxious Symptoms

Items pertaining to depressive and anxious symptoms over the last 2 weeks were obtained using the Baltimore How I Feel Questionnaire (BHIF).[15] The BHIF contains 45 self-reported items measured on a 4-point Likert-type scale from “Never” to “Most Times,” and were developed to map onto the Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (DSM-IV).[30] Symptom averages were calculated for childhood (i.e., 1–3 grades) and adolescence (i.e., 6–12 grades). Reliability estimates averaged α = .79 over elementary school and α = .81 for the middle school years.

Psychopathology

Several lifetime diagnoses were investigated using data collected during the waves from ages 17–18 through 22–23 and variables were considered endorsed if diagnostic criteria were met at any wave. MDD was assessed using items from the C-DISC-IV,[29] whereas the Drug and Alcohol Dependence Disorder diagnoses were assessed using criteria from the National Survey on Drug Use and Health.[31] These disorders were coded as present if the individual endorsed items in three or more symptom categories. Both measures are consistent with criteria set forth by the DSM-IV.[30]

RESULTS

Descriptive statistics for nonattempters, and inconsistent and consistent reporters of suicide attempt, separated by conceptualization (1YR and EVER), are displayed in Table 1. Overall, 19.3% of the sample reported a suicide attempt at any time during the study, with females significantly more likely to report a lifetime attempt than males (t = 2.89, df = 618, P = .004). The incidence of suicide attempt (i.e., new cases reporting a suicide attempt each year) ranged from 0.40 to 3.5%. Figure 1 displays the annual lifetime prevalence (i.e., any cases who reported a lifetime attempt during each wave of data collection) superimposed on the cumulative incidence of attempt (i.e., what we would expect given the increase in new cases being added each year). It can be seen that although the lifetime prevalence of suicide attempt reporting would be expected to increase over time, the rates appear stable in our sample.

TABLE 1.

Description of the Sample

Total sample
N = 678		 Nonattempters
547 (80.7%)		 Lifetime suicide attempt reporters 131 (19.3%)
Inconsistent reporters
Consistent reporters
1YR
68 (10.0%)		 EVER
116 (17.1%)		 1YR
36 (5.3%)		 EVER
15 (0.02%)
Continuous variables: mean (SD)
 Childhood depression 0.74 (0.25) 0.72 (0.24) 0.77 (0.24) 0.81 (0.27) 0.86 (.030) 0.75 (0.23)
 Childhood anxiety 0.76 (0.28) 0.74 (0.27) 0.81 (0.28) 0.83 (0.30) 0.83 (0.33) 0.80 (0.28)
 Depression 0.61 (0.33) 0.55 (0.30) 0.73 (0.30) 0.82 (0.36) 0.88 (0.36) 0.90 (0.37)
 Anxiety 0.56 (0.31) 0.51 (0.29) 0.72 (0.36) 0.71 (0.35) 0.69 (0.39)
 Suicidal ideation 0.03 (0.10) 0.01 (0.04) 0.10 (0.13) 0.12 (0.15) 0.19 (0.22) 0.24 (0.26)
Dichotomous variables: N (%)
 Female 316 (46.6%) 240 (43.9%) 35 (51.5%) 67 (57.8%) 23 (63.9%) 9 (60.0%)
 Black 585 (84.9%) 476 (87.0%) 58 (85.3%) 99 (85.3%) 27 (75.0%) 10 (66.7%)
 Free/reduced lunch status 463 (68.3%) 368 (67.3%) 47 (69.1%) 83 (71.6%) 26 (72.2%) 12 (80.0%)
 Intervention group 459 (67.7%) 368 (67.3%) 45 (66.2%) 79 (68.1%) 25 (69.4%) 12 (80.0%)
 Major depressive disorder 39 (5.8%) 20 (3.7%) 6 (8.8%) 16 (13.8%) 8 (22.2%) 3 (20.0%)
 Drug dependence disorder 95 (13.8%) 57 (10.4%) 16 (23.5%) 32 (27.6%) 14 (38.9%) 6 (40.0%)
 Alcohol dependence disorder 36 (5.3%) 20 (3.7.%) 7 (10.3%) 14 (12.1%) 7 (19.4%) 2 (13.3%)
Open in a new tab

The table is organized to provide descriptive statistics for each group investigated. Therefore, values displayed exist for the total sample, individuals who never reported an attempt during the duration of the study, and those who did report an attempt at some point during the study. The suicide attempt reporters are further broken down into inconsistent and consistent reporters by study design or conceptualization. Due to missing data (i.e., individuals who reported their first suicide attempt at the last data collection wave could not have a 1-year follow-up), the number of suicide attempt reporters differs based on conceptualization. SD, standard deviation; 1YR, the 1-year follow-up study design; investigating consistency 1 year following first reported attempt. EVER, study design investigating consistency at each data collection wave following first attempt reporting, regardless of how many follow-up collection points occurred.

Figure 1.

Figure 1

Open in a new tab

Lifetime suicide attempt reporting based on observed versus expected reporting. Note. Due to anomalies related to budgeting, the years 2004 and 2005 are not represented in the table. OBSERVED, percentage of the sample per year that reported lifetime suicide attempt; EXPECTED, what would be expected based on the cumulative incidence.

DEMOGRAPHIC AND CLINICAL CHARACTERISTIC COMPARISONS

Comparisons between the three groups (i.e., consistent, inconsistent, and nonattempt reporters) based on the different conceptualizations are displayed in Table 2. Overall, the pattern of statistically significant differences is quite similar for both conceptualizations, with omnibus tests indicating no significant differences on demographic variables and significant differences evident for clinical variables particularly in adolescence. There were significant differences in childhood self-report of depression, rather than anxiety, across the groups.

TABLE 2.

Comparisons between inconsistent (In) and consistent (Cs) suicide attempt reporters and those who never report an attempt (non) based on different conceptualizations of consistent reporting

1YR
EVER
F P-Value Post hoc F P-Value Post hoc
Continuous variables
 Childhood depression 5.51 .004 [(Cs/non (P = .006)] 4.97 .007 (In/non, P = .005)
 Childhood anxiety 3.00 .051 4.40 .013
 Depression 26.68 <.000 Cs/non (P < .000)
In/non (P < .000)		 38.99 <.000 Cs/non (P < .000)
In/non (P < .000)
 Anxiety 11.86 Cs/non (P = .001)
In/non (P = .002)		 23.18 <.000 In/non (P < .000)
 Suicidal ideation 124.59 <.000 Cs/non (P < .000)
In/non (P < .000)
Cs/In (P < .000)		 124.33 <.000 Cs/non (P < .000)
In/non (P < .000)
Cs/In (P < .000)
Dichotomous variables
 Female 3.24 .40 4.29 .014
 Black 1.58 .40 2.07 .127
 Free/reduced Lunch status 0.38 .40 1.00 .369
 Intervention status 0.62 .40 1.26 .286
 Major depressive disorder 13.21 <.000 Cs/non (P < .000) 12.32 <.000 In/non (P < .000)
 Drug dependence disorder 15.88 <.000 Cs/non (P < .000)
(In/Non [P = .006])		 16.72 <.000 Cs/non (P = .003)
In/non (P < .000)
 Alcohol dependence disorder 10.10 <.000 Cs/non (P < .000) 7.86 <.000 In/non (P = .001)
Open in a new tab

F, F statistic; post hoc, post hoc test comparisons, group with higher mean listed first. Degrees of freedom (df), 2 for all results. Only significant post hoc test (and those approaching significance; P ≤ .006) results from ANOVAs are reported here. Those comparisons approaching significance are presented in parentheses.

a

P ≤ .004.

Regardless of conceptualization, post hoc analyses among the groups indicated a statistically significant difference between consistent and inconsistent reporters on only one variable-suicidal ideation (1YR: F [2,624] = 124.59, P < .001; EVER: F [2,651] = 124.33, P < .001). Consistent reporters reported the highest mean levels of suicidal ideation, followed by inconsistent reporters, then nonattempters. All other differences were accounted for by comparisons between attempter (i.e., either consistent or inconsistent reporters of suicide attempt) and nonattempter groups. Indeed, when statistically significant differences did exist, both attempt reporter groups (i.e., inconsistent and consistent) demonstrated significant differences from the nonattempters. Although there were more differences between the inconsistent reporters and nonattempters using the EVER conceptualization, it is important to note that small numbers of consistent reporters may have made it difficult to detect any differences.

COMPARISONS OF CHARACTERISTICS OF THE ATTEMPT OR ATTEMPTERS

To further understand specific suicidal thoughts and behaviors among individuals who had reported an attempt at any wave of data collection, additional comparisons were made regarding characteristics of the attempts. As indicated in Table 3, the only significant difference after Bonferroni correction occurred in the EVER conceptualization with age at first reported attempt. Inconsistent reporters were significantly more likely to be younger (mean = 9.73, SD = 3.11) than consistent reporters (mean = 13.6, SD = 3.48). Interestingly, age at first attempt was approaching significance for the 1YR conceptualization, and the results were in the opposite direction for the EVER conceptualization, with inconsistent reporters older (mean = 10.3, SD = 3.47) than consistent reporters (mean = 8.83, SD = 2.89). No other significant differences (P > .01) were evident between consistent and inconsistent reporters, regardless of reporting conceptualization.

TABLE 3.

Characteristics of suicide attempts among inconsistent and consistent reporters

1YR
EVER
t df P t df P
Age first attempt reported 2.29 84 .03 −4.10 17 .001
Ideation at time of reported attempt − 1.74 62 .09 −0.56 17 .580
Medical care for an attempt − 0.38 30 .71 −2.59 8 .032
Plan for attempt − 0.99 27 .33 −1.09 6 .322
Multiple attempts 0.71 27 .49 −0.40 22 .700
Open in a new tab

t, two-tailed-test statistic; df, degrees of freedom; P, P-value. After Bonferroni correction for multiple tests, a statistically significant P-value is ≤.01.

DISCUSSION

Using prospective data from 12 annual waves of data collection from adolescence to young adulthood, we estimated the lifetime prevalence of suicide attempt in our sample to be 19.3%. Although rates may vary (e.g., 4.1–23.5% in European samples[32]), these rates appear higher than in nationally representative samples (i.e., approximately 3.0% with U.S. adolescents[18]). However, if data were analyzed cross-sectionally, as is common, the rate of lifetime suicide attempt in our sample would be commensurate with established rates (i.e., yearly rates ranged from 3.5 to 5.9% with an average of 4.8%). Further, as is highlighted by our analyses and similar results from related disorders,[15, 33] current estimates of lifetime prevalence of suicide attempt may in fact represent an underestimate of these behaviors. Although there should be an increase over time as new cases are added (i.e., cumulative lifetime prevalence), there instead remains a fairly constant yearly rate of approximately 5% of the sample. As would be expected, inconsistent reporting of suicide attempts was quite common, with the large majority (i.e., between 65.3 and 88.5% based on one wave or over 12 waves, respectively) inconsistently reporting a suicide attempt.

Another goal of the current study was to compare characteristics of those who consistently, inconsistently, and never reported a suicide attempt. There did not appear to be any demographic differences among the groups, while statistically significant clinical differences did exist between attempters and nonattempters (i.e., either inconsistent reporters and nonattempters or consistent reporters and nonattempters). Importantly, few differences existed between those who consistently and inconsistently reported a lifetime suicide attempt. In fact, the only way in which these reporters differed was on measures of lifetime suicidal ideation. Consistent reporters had a higher prevalence of suicidal ideation over time, followed by inconsistent reporters and finally nonattempters. These findings are similar to comparable studies with adult samples,[13, 14] yet diverge from adolescent and young adult research in this area, which has found that consistent reporters are more likely to be female and to have increased mental health diagnoses.[10] Additionally, Christl et al.[10] observed age differences within their sample, with younger participants more likely to be consistent reporters. Our findings indicate that age differences might be a function of follow-up, in that when the follow-up extended over the length of the study (EVER), inconsistent reporters were significantly more likely to be younger, yet when follow-up was 1 year (1YR), inconsistent reporters tended to be older. No further differences were evident between consistent and inconsistent reporters on suicide attempt characteristics such as age at first reported attempt, medical care for an attempt, plan for the attempt, or lifetime multiple attempts.

Previous research has suggested that inconsistent reporting of suicide attempt may be related to forgetting and that such forgetting might have a protective effect[10, 11] or might be related to less depression at the time of follow-up assessment.[14] It is important to note that in this sample, consistent and inconsistent reporters were more similar than different. Several important differences should be noted between our study and those demonstrating statistically significant differences between consistent and inconsistent reporters of a suicide attempt. For example, many of the independent variables used in our study were collected prospectively and averaged over distinct developmental periods (e.g., depression over the adolescent years) rather than retrospective measurement at one time point, our comparisons included nonattempters, and the number of consistent reporters was small in several comparisons, perhaps influencing the lack of significant results for consistent reporters in our EVER conceptualization. Our results point to the need to further investigate the stability of suicide attempt reporting. It may be that the consistent and inconsistent reporters represent different subtypes of suicide attempters (e.g., depressive and impulsive, respectively) and each are associated with different pathways and trajectories over time.

Our results have important implications for both clinical care and research. From a clinical perspective, our results demonstrate that inconsistent reporting is common; the prevalence and the lack of consistent association with clinical or demographic variables suggests that inconsistent reporting is likely not “less severe” than consistent reporting. As a result, every positive response on questions related to suicide attempts requires appropriate follow-up examination and discussion. In addition, due to inconsistent reporting, repeated screening, and assessment, in settings such as outpatient and inpatient mental health services, primary care, and schools, is likely needed. From a research perspective, inconsistent reports of suicide attempt will be an inevitable part of longitudinal suicide research. Due to our results, we propose that exclusion of inconsistent reporters from analyses is not necessary and may in fact be detrimental to the science.

As with any study, there are important limitations to discuss. The current study was not designed for the primary purpose of investigating suicidal thoughts and behaviors. However, since one of the goals of the original study was to examine depression over the lifespan, suicidal ideation and behaviors were assessed from sixth grade through adulthood, providing a unique longitudinal data set in the field of suicide research and an opportunity to study these behaviors over the lifespan. Additionally, consistency of reporting was investigated over the transition from adolescence to young adulthood in an urban community sample. It is not clear whether the results would generalize to other populations of interest such as adults and clinical populations. Important risk and protective factors for suicide attempt such as childhood abuse, medical lethality or suicide intent, social and academic competence were also not assessed.

Furthermore, a large number of tests were conducted on the data, resulting in a very small P-value identifying statistical differences between the groups. Despite this, there were few variables for which a more lenient P-value would have changed the determination of significance. Additionally, as the study was primarily descriptive and exploratory in nature, there can be no discussion regarding etiology. Future directions will include examining developmental trajectories of depressive, anxious, aggressive, and impulsive symptoms over time in order to investigate the possibility of multiple pathways to suicidal behaviors. This study was an important first step in this exploration, particularly in the investigation of consistency in reporting.

Finally, no confirmation of suicide attempt was available. As data were acquired starting in sixth grade, it may be that participants did not fully understand the meaning of suicide attempt (and reported that an aborted or interrupted attempt was an actual suicide attempt). It is possible that more serious thoughts about suicide resulted in a participant affirming an attempt when no action was taken. However, there were significant differences between attempters and nonattempters on most variables (including suicide ideation). Additionally, differences related to age of first attempt reported were not consistent, and therefore offer little clarity to the hypothesis that consistent reporters were older and more likely to understand the concept of suicide attempt.

CONCLUSIONS

The recently updated NSSP has called for more nationally representative data collection on suicidal behaviors and improved data quality in suicide research. The issue of inconsistent reporting will likely become more prevalent as longitudinal screening and assessment is conducted as part of routine clinical care or patient-centered outcomes’ initiatives. As demonstrated in the current study, inconsistency in reporting suicidal behaviors over time is frequent. Inconsistent reporters add to our understanding of suicidal thoughts and behaviors. Further research on motivation, memory, and perception of suicide attempts should be conducted to improve the quality of research in this important area.

Within the context of the recent proliferation of routine screening for depression and suicide risk by systems of care (primary care, emergency departments, inpatient and outpatient psychiatry settings), it is inevitable that inconsistency in reporting of suicide attempt will occur. This may leave those responsible for the care of these individuals wondering what to do and how to make sense of these inconsistencies. Nonetheless, our results imply that inconsistent reporters require additional assessment and follow-up when identified.

Acknowledgments

This research was supported by grants from the National Institute of Mental Health (R01MH057005 and T-32MH018834 to Nicholas S. Ialongo; T-32MH014592 to Peter Zandi; R21MH090480 to Holly C. Wilcox) and the National Institute on Drug Abuse (R37DA11796 to Nicholas S. Ialongo). We thank the Baltimore City Public Schools for the collaborative efforts and the parents, children, teachers, principals, school psychologists, and social workers who participated.

Footnotes

Conflict of Interest. The authors have declared no conflict of interest.

References

  • 1.Center for Disease Control and Prevention. Morbidity and Mortality Weekly Report. SS6. Vol. 61. Atlanta, GA: U.S. Department of Health and Human Services, CDC; 2012a. Surveillance for violent deaths: national violent death reporting system, 16 states, 2009. Available at: http://www.cdc.gov/mmwr/pdf/ss/ss6106.pdf. [PubMed] [Google Scholar]
  • 2.Lubel KM, Kegler SR, Crosby AE, Karch D. Suicide trends among youths and young adults aged 10–24 years—United States, 1990–2004. Morb Mortal Wkly Rep. 2007;56:905–908. Available at www.cdc.gov/mmwr/ [PubMed] [Google Scholar]
  • 3.Center for Disease Control and Prevention. National suicide statistics at a glance: trends in suicide rates among persons ages 10 years and older, by sex, United States, 1991–2009. 2012b Available at: http://www.cdc.gov/ViolencePrevention/suicide/statistics/trends01.html.
  • 4.Rockett IRH, Regier MD, Kapusta ND, et al. Leading causes of unintentional and intentional injury mortality: United States, 2000–2009. Am J Public Health Advance online publication. 2012;102:e84–e92. doi: 10.2105/AJPH.2012.300960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.U.S. Department of Health and Human Services (HHS), Office of the Surgeon General and National Action Alliance for Suicide Prevention. National Strategy for Suicide Prevention: Goals and Objectives for Action. Washington, DC: HHS; 2012. Available at: http://www.surgeongeneral.gov/library/reports/national-strategy-suicide-prevention/ [PubMed] [Google Scholar]
  • 6.Groholt B, Ekeberg O, Haldorsen T. Adolescent suicide attempters: what predicts future suicidal acts? Suicide Life Threat Behav. 2006;36:638–650. doi: 10.1521/suli.2006.36.6.638. [DOI] [PubMed] [Google Scholar]
  • 7.Posner K, Brown GK, Stanley B, et al. The Columbia-Suicide Severity Rating Scale: initial validity and internal consistency findings from three multisite studies with adolescents and adults. Am J Psychiatry. 2011;168:1266–1277. doi: 10.1176/appi.ajp.2011.10111704. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Center for Disease Control and Prevention. Youth Risk Behavior Survey. 2012c Available at: www.cdc.gov/yrbs.
  • 9.National Institute on Alcohol Abuse and Alcoholism. National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) 2013 Available at: www.psc.isr.umich.edu/dis/data/catalog/detail/1179.
  • 10.Christl B, Wittchen H-U, Pfister H, Lieb R, Bronisch T. The accuracy of prevalence estimations for suicide attempts. How reliably do adolescents and young adults report their suicide attempts? Arch Suicide Res. 2006;10:253–263. doi: 10.1080/13811110600582539. [DOI] [PubMed] [Google Scholar]
  • 11.Goldney RD, Smith S, Winefield AH, Tiggemann M, Winefield HR. Suicidal ideation: its enduring nature and associated morbidity. Acta Psychiatr Scand. 1991;83:115–120. doi: 10.1111/j.1600-0447.1991.tb07375.x. [DOI] [PubMed] [Google Scholar]
  • 12.Statham DJ, Heath AC, Madden PAF, et al. Suicidal behavior: an epidemiological and genetic study. Psychol Med. 1998;28:839–855. doi: 10.1017/S0033291798006916. [DOI] [PubMed] [Google Scholar]
  • 13.Barber ME, Marzuk PM, Leon AC, Portera L. Gate questions in psychiatric interviewing: the case of suicide assessment. J Psychiatr Res. 2001;35:67–69. doi: 10.1016/S0022-3956(01)00004-8. [DOI] [PubMed] [Google Scholar]
  • 14.Palmetto NP. How unreliable are adult-reported suicide attempts? An examination of correlates and underlying causal mechanisms of discordant reporting over time. Sci Eng. 2012;73(2-B):847. (UMI No. 3479435). Retrieved April 13, 2013 from Dissertations and Theses database. [Google Scholar]
  • 15.Streiner DL, Patten SB, Anthony JC, Cairney J. Has ‘lifetime prevalence’ reached the end of its life? An examination of the concept. Int J Methods Psychiatr Res. 2009;18:221–228. doi: 10.1002/mpr.296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Ialongo NS, Kellam SG, Poduska J. A developmental epidemiological framework for clinical child and pediatric psychology research. In: Drotar D, editor. Handbook of Research in Pediatric and Clinical Child Psychology: Practical Strategies and Methods. Dordrecht, the Netherlands: Kluwer Academic Publishers; 1999. pp. 3–19. [Google Scholar]
  • 17.Kessler RC, Borges G, Walters EE. Prevalence of and risk factors for lifetime suicide attempts in the National Comorbidity Study. Arch Gen Psychiatry. 1999;56:617–626. doi: 10.1001/archpsyc.56.7.617. [DOI] [PubMed] [Google Scholar]
  • 18.Wolitzky-Taylor KB, Ruggiero KJ, McCart MR, et al. Has adolescent suicidality decreased in the United States? Data from two national samples of adolescents interviewed in 1995 and 2005. J Clin Child Adolesc Psychol. 2010;39:64–76. doi: 10.1080/153744109034011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kubrin CE, Wadsworth T. Explaining suicide among blacks and whites: how socioeconomic factors and gun availability affect race-specific suicide rates. Soc Sci Q. 2009;90:1203–1227. doi: 10.1111/j.1540-6237.2009.00654.x. [DOI] [Google Scholar]
  • 20.Ialongo NS, Koenig-McNaught AL, Wagner BM, et al. African American children’s reports of depressed mood, hopelessness, and suicidal ideation and later suicide attempts. Suicide Life Threat Behav. 2004;34:395–407. doi: 10.1521/suli.34.4.395.53743. [DOI] [PubMed] [Google Scholar]
  • 21.Hill RM, Castellanos D, Pettit JW. Suicide-related behaviors and anxiety in children and adolescents: a review. Clin Psychol Rev. 2011;31:1133–1144. doi: 10.1016/j.cpr.2011.07.008. [DOI] [PubMed] [Google Scholar]
  • 22.Borges G, Nock MK, Abad JMH, et al. Twelve-month prevalence of and risk factors for suicide attempts in the World Health Organization World Mental Health Surveys. J Clin Psychiatry. 2010;71:1617–1628. doi: 10.4088/JCP.08m04967blu. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Wilcox HC, Arria AM, Caldeira KM, Vincent KB, Pinchevsky GM, O’Grady KE. Prevalence and predictors of persistent suicide ideation, plans, and attempts during college. J Affect Disord. 2010;127:287–294. doi: 10.1016/j.jad.2010.04.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.MacDonald R, Taylor J, Clarke D. The relationship between early suicide behaviors and mental health: results from a nine-year panel study. J Adolesc. 2009;32:1159–1172. doi: 10.1016/j.adolescence.2009.01.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Nock MK, Hwang I, Sampson NA, Kessler RC. Mental disorders, comorbidity and suicidal behavior: results from the national comorbidity survey replication. Mol Psychiatry. 2010;15:868–876. doi: 10.1038/mp.2009.29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kessler RC, Borges G, Walters EE. Prevalence of and risk factors for lifetime suicide attempts in the National Comorbidity Study. Arch Gen Psychiatry. 1999;56:617–626. doi: 10.1001/archpsyc.56.7.617. [DOI] [PubMed] [Google Scholar]
  • 27.Sapyta J, Goldston DB, Erkanli A, et al. Evaluating the predictive validity of suicidal intent and medical lethality in youth. J Consult Clin Psychol. 2012;80(2):222–231. doi: 10.1037/a0026870. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Graham JW. Missing data analysis: making it work in the real world. Annu Rev Psychol. 2009;60:549–576. doi: 10.1146/annurev.psych.58.110405.085530. [DOI] [PubMed] [Google Scholar]
  • 29.Shaffer D, Fisher P, Lucas C, Dulcan M, Schwab-Stone M. NIMH diagnostic interview schedule for children version IV (NIMH DISC-IV): description, differences from previous versions, and reliability of some common diagnoses. J Am Acad Child Adolesc Psychiatry. 2000;39:28–38. doi: 10.1097/00004583-2000001000-00014. [DOI] [PubMed] [Google Scholar]
  • 30.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4. Washington, DC: Author; 1994. (DSM-IV) [Google Scholar]
  • 31.Substance Abuse and Mental Health Services Administration. Results from the 2010 National Survey on Drug Use and Health: Summary of National Findings. Rockville, MD: Author; 2011. NSDUH Series H-41, HHS Publication No. (SMA) 11-4658. [Google Scholar]
  • 32.Kokkevi A, Rotsika V, Arapaki A, Richardson C. Adolescents’ self-reported suicide attempts, self-harm thoughts and their correlates across 17 European countries. J Child Psychol Psychiatry. 2012;53:381–389. doi: 10.1111/j.1469-7610.2011.02457.x. [DOI] [PubMed] [Google Scholar]
  • 33.Patten SB. Major depression prevalence is very high, but the syndrome is a poor proxy for community populations’ clinical treatment needs. Can J Psychiatry. 2008;53:411–419. doi: 10.1177/070674370805300702. Available at: http://publications.cpa-apc.org/browse/sections/0. [DOI] [PubMed] [Google Scholar]

RESOURCES