First suicide attempts in early adolescents: a descriptive outcomes study

Abstract

Objective

In 2021, suicide was the second leading cause of death in early adolescent Americans ages 10–14. Research into suicidal behavior in this age group is limited. We report on prior psychiatric care, attempt method, and attempt outcomes in a cohort of 164 early adolescents accrued by first suicide attempt coming to medical attention.

Methods

Our cohort constitutes a subsample from a previously reported retrospective-prospective study identified through the Rochester Epidemiology Project that recruited individuals making first suicide attempts coming to medical attention (index attempt, IA) during a 22-year period (1/1/1986–31/12/07). Among 1490 all-age index attempters followed until 12/31/2010, 164 (11.0%) were aged 10–14.

Results

3/164 died on IA (1.8% of the cohort; 2 females, 1 male). Nearly half (72/164, 43.9%) had no prior psychiatric history. Females were less likely than males to have seen a mental health provider (p=0.029) or been prescribed psychiatric medications (p<0.001) prior to IA. Medication overdose was the most common attempt method in females (81/128, 63.3%), while cutting or piercing wounds were the most common method in males (13/36, 36.1%). Females were significantly more likely than males to overdose (p=0.001). Of IA survivors, 19.9% (32/161) were initially medically hospitalized, 52.8% (85/161) were psychiatrically hospitalized—initially or in transfer—and 37.2% (60/161) were discharged without hospitalization.

Conclusion

Medication overdoses accounted for over half of all IAs and were significantly more common in females. While IA mortality was low relative to older patients from the all-age-cohort, morbidity was substantial with nearly a fifth of attempts severe enough to warrant medical hospitalization and more than half initial or eventual psychiatric hospitalization. These findings emphasize the importance of both means restriction and identification of early adolescents at risk before they make their first attempt.

Introduction:

In the United States in 2021, 598 10–14 year olds killed themselves, making suicide this age group’s second leading cause of death.¹ Herein we refer to individuals aged 10–14 as early adolescents. Between 2001 and 2021, the U.S. suicide rate for early adolescents increased by 283% in females and 68% in males.² Typically peripubertal, those in this age range are undergoing significant physical, social, and cognitive changes and in the United States are typically attending late elementary to early high school.

Completed suicides are exceedingly uncommon before the age of 10.³ In comparisons of early and late adolescents (ages 15–19), the younger attempters have lower suicidal intent,^4–6 less likelihood of previous psychiatric diagnoses⁷ and more impulsivity in their attempts.^4,7,8 Diagnoses in those who complete suicide differ as well, with children under the age of 12 more likely to have had attention deficit/hyperactivity disorder and those older being more likely to have had a mood disorder.^7–9

Survivors of youth suicide attempts experience sequelae that stretch into adulthood. A longitudinal study comparing survivors of suicide attempts before the age of 25 with non-attempters found that in adulthood, attempters were more likely to have mental and physical health difficulties, violent behaviors, and reliance on social supports manifesting in inability to maintain employment.¹⁰ Despite this knowledge, little is known about the specific epidemiology, antecedents, and management of those making initial suicide attempts as early adolescents. Considering the substantial concerns for suicidal behavior in early adolescence, combined with recent trends showing rising suicide rates and worse adult outcomes, it is vital to understand the antecedents of early suicidal behavior.

We have published two studies from a cohort of patients accrued by first suicide attempt coming to medical attention (index attempt, IA). The first reported on the all-age-cohort (N=1490) and the second on a subsample of youth ages 10–24 (N=813).^11,12 These studies underscored the potential lethality of IA, with 59.3% in the all-age-cohort and 71.4% in the youth subsample dying by suicide on their IA. Firearms figured prominently in completed IAs, and were responsible for 72.9% of IA deaths in the all-age-cohort and 85.0% of IA deaths in the youth subsample.

Given the dearth of published information on IAs in early adolescents, we focus in this descriptive study on the 164 early adolescents ages 10–14 who comprised 11.0% (164/1490) of the all-age cohort. Our goals are to characterize early life index attempts by sex differences, means of attempt, pre-attempt psychiatric care and post-attempt acute psychiatric management. In addition, we have queried the National Death Index to determine whether any individuals died in the years following the IA, including death by suicide.

Methods:

Our cohort is drawn from residents of Olmsted County, Minnesota, USA. Located in Southeastern Minnesota, Olmsted County is geographically isolated, resulting in most residents receiving medical care at one of two facilities: Mayo Clinic and Olmsted Medical Center. Taking advantage of this geographical isolation, the Rochester Epidemiology Project (REP) has accumulated all the county’s medical records, inclusive of these two facilities, since the early 20^th century.¹³ Hundreds of studies using this database have yielded findings generalizable to the nation as a whole.¹⁴

Using the REP, we conducted a retrospective-prospective study of Olmsted County residents followed from an IA. We identified this cohort by searching the REP using Hospital International Classification of Diseases, Adapted and International Classification of Disease, Ninth Edition codes¹⁵ for suicidal ideation, suicidal behavior or self-injurious behavior requiring medical or psychiatric care occurring between January 1, 1986 and December 21, 2007. In this 22-year time span we found 17,288 coded events connected with 8,352 individuals of which 5,283 were county residents. These charts were screened to ensure the coded event was in fact an actual suicide attempt and that it was an IA made during the study period. 1490 subjects met inclusion criteria, of which 164 were early adolescents.

The charts of these early adolescent index attempters were then examined for data regarding method of attempt, psychiatric history, post-attempt acute management and psychiatric follow-up. The National Death Index was queried until January 1, 2011 to see whether any of the 164 subjects perished. Recruited subjects therefore had a variable follow-up period of between 3 and 25 years.

Patients’ demographic, psychological, and IA characteristics were compared by sex using chi-square or Fisher’s exact tests for categorical variables, and a Wilcoxon rank sum test for age. Hospitalization categories of medical, psychiatric, medical and psychiatric post-attempt were examined as a predictor of follow-up psychiatric care using a logistic regression model. Analyses were conducted using SAS (version 9.4; Cary, NC). The study was approved by the Institutional Review Boards at Mayo Clinic and Olmsted Medical Center.

Results:

In early adolescents ages 10–14, 164 IAs occurred during the study period. The mean age of index attempt was 13.7 (SD=1.1) with a trend toward males being younger on IA than females (males =13.2±1.5; females=13.8±0.9 p=0.057).

Females accounted for the majority of IAs at 78.0% (128/164). Proportionately, the ratio of female to male IAs was 3.6:1. All three who died by suicide – two females and a male – did so on IA. None of the surviving 161 died by suicide during the study follow-up period although one subject perished in a non-suicide related death. <Table 1>

Table 1.

Index and subsequent death rates in males and females ages 10–14

	Index Death Rate							Subsequent Death Rate
	Attempt		Suicide		Survival			Suicide		Other Death		Any Death
	N	% of total	N	%	N	%	p-value*	N	%	N	%	N	%
Males	36	22.0	1	2.8	35	97.2	0.527	0	0	0	0	0	0
Females	128	78.0	2	1.6	126	98.4		0	0	1	0.8	1	0.8
Total	164		3	1.8	161	98.2		0	0	1	0.6	1	0.6

^*Fisher’s exact test

Of the three who died by suicide on the IA, two were by firearm and one by hanging. Medication overdose accounted for the preponderance of IAs: 56.7% (93/164). The most common method for females was medication overdose (63.3%, 81/128), while the most common method for males was cutting/piercing (36.1%. 13/36), closely followed by medication overdose (33.3%, 12/36). Females were significantly more likely than males to use medication overdose as their index method of attempt (63.3% vs. 33.3%; p=0.001). <Table 2> In neither males nor females was the most common method ever fatal.

Table 2.

Method of index attempt

Method of Index Attempt	Overall (%)	Males (%)	Females (%)	p-value
Hanging/Asphyxiation	11 (6.7)	7 (19.4)	4 (3.1)
Firearms	3 (1.8)	1 (2.8)	2 (1.6)
Medication Overdose	93 (56.7)	12 (33.3)	81 (63.3)	0.0014*
Non-Medication Overdose/Poisoning	5 (3.0)	2 (5.6)	3 (2.3)
Cutting/Piercing	46 (28.0)	13 (36.1)	33 (25.8)
Other	6 (3.7)	1 (2.8)	5 (3.9)
TOTAL	164	36	128

^*Chi-square test; medication overdose vs all other categories

Males were more likely than females to have had prior psychiatric histories. Male index attempters were significantly more likely to have met with a mental health professional than females {66.7% (24/36) vs. 46.1% (59/128), p=0.029}. Males were also significantly more likely than females to have been treated with a psychotropic medication {55.6% (20/36) vs. 24.2% (31/128), p<0.001}. There was also a trend towards males having been more likely to have been diagnosed with a documented psychiatric condition {61.1% vs. 43.0%, p=0.054}. Overall, 43.9% of individuals making IAs had no recorded past psychiatric history. <Table 3>

Table 3.

Psychiatric history prior to index attempt

History	Overall (%)	Males (%)	Females (%)	p-value*
Past psychiatric history	83 (50.6)	24 (66.7)	59 (46.1)	0.029
Past psychiatric medication	51 (31.1)	20 (55.6)	31 (24.2)	<0.001
Past psychiatric diagnosis	77 (47.0)	22 (61.1)	55 (43.0)	0.054
No past psych record	72 (43.9)	11 (30.6)	61 (47.7)	0.068

^*Chi-square test

IAs in early adolescents were frequently serious enough to require medical hospitalization. Of those surviving their IA, 19.9% (32/161) required inpatient medical treatment with 50% (16/32) of these subsequently transferred for psychiatric hospitalization and the remainder discharged without acute psychiatric hospital care following medical stabilization. Psychiatric hospitalization was the initial management for 38.1% (69/161) of IAs with a total of 52.7% (85/161) being psychiatrically hospitalized as part of treatment following IA. 37.3% (60/161) of survivors were discharged without any inpatient medical or psychiatric hospitalization following their index attempt.<Figure 1>

Figure 1.

Acute Post-Index Attempt Management

Overall, 79.5% (128/161) of early adolescent survivors had a follow up appointment made following their IA. Those neither medically nor psychiatrically hospitalized following their IA were the least likely to have a follow up appointment, with only 60.0% (36/60) having a documented scheduled follow-up appointment. Overall, the odds for those receiving any type of hospitalization to have psychiatric follow up care made was 6.8 times higher than the odds for those who were not hospitalized (95% CI = 2.89–16.06). There was no difference by sex. <Table 4>

Table 4.

The effects of post-attempt hospitalization categories on the likelihood of follow-up psychiatric care.

Effect	OR	95% OR CI	Model p-value
Med hosp. vs. No hosp.	10.00	1.24–80.78	<0.001
Med and Psych hosp. vs No hosp.	10.00	1.24–80.78
Psych hosp. vs No hosp.	5.91	2.31–15.07
Any hosp. vs. No hosp.	6.81	2.89–16.06

Discussion:

To the best of our knowledge, this is the first study characterizing some aspects of the post-IA course of early adolescents ages 10–14. With only three deaths (1.8%, 3/164), the rate in the early adolescent subsample was much lower than in the all-age-cohort (5.4%) and the youth subsample (3.6%). But like their elders, the three who died did so by violent means – two by gunshot, one by hanging –emphasizing the need for firearms restriction.^11,12,16

Beyond mortality, however, substantial medical morbidity was incurred with approximately 20% of attempts necessitating inpatient medical care. Medication overdose was responsible for almost 80% of these medical hospitalizations. Though girls most commonly attempted suicide by overdose and were significantly more likely than boys to overdose on IA, it still ranked as the second most common attempt method in boys. Given the prominence of medication overdoses and their connection to medical hospitalization, household restriction of medication access is an important part of reducing the morbidity of suicide attempts.

Compared to the larger youth cohort of 10–24 year olds with IAs,¹² rates of initial medical hospitalization were lower in the 10–14 year old cohort (25.7% [204/793] vs. 19.9% [32/161]). This is consistent with prior literature positing lower suicidal intent among early adolescents compared to late adolescents.^4–6 However, psychiatric hospitalization without a preceding medical hospitalization was more likely among the younger cohort (44.3% [351/793] vs. 52.7% [85/161]). This discrepancy could reflect differences in the age of consent for hospitalization between the two groups, differences in medical severity of attempts between the two groups, the seriousness with which early adolescent attempts are viewed, or some other explanation which could be elucidated in a future study.43.9% of early adolescents making an IA had no psychiatric history. General surveys of adolescents 13–18 years old have reported that only 36.2% with mental health problems receive services and only half with severely impairing conditions are in care.¹⁷ In a survey of parents of children ages 4–17, only 15% reported that they had shared mental health concerns with school staff or health care providers.¹⁸ While subjects in our study were more likely to have received psychiatric care than these survey subjects, nearly half had had no previous psychiatric exposure.

Additionally, recognition of psychiatric concerns in early adolescents prior to IA differed between sexes. We found that males were more likely than females to have had psychiatric visits (p=0.029) or have been prescribed psychotropic medication (p<0.001). We also observed a trend for boys to have been more likely to have received a psychiatric diagnosis (p=0.054). This concurs with prior reports on deliberate self-harm (DSH) in children under 15 that showed that girls were less likely than boys to have had psychiatric treatment prior to the first DSH incident.^17,19 Boys are more likely than girls to be diagnosed and treated for mental health conditions,²⁰ especially disruptive behaviors and ADHD, which are more common in suicide deaths among the early adolescent age group.⁸”

Although approximately 80% of early adolescents IA survivors were given psychiatric follow-up appointments, we found a sharp discrepancy between those hospitalized and those without hospitalization. Post-IA follow-up appointments were 6.8 times more likely to be scheduled in those receiving any type of hospitalization versus those who were not hospitalized. This means that IA survivors who were deemed not to require an inpatient level of care were also less likely to have had follow-up appointments scheduled. In light of a history of a suicide attempt being both the strongest predictor of repeated suicide attempt²¹ and death by suicide,^7,22 this was surprising. Future investigation of follow-up care IA survivors receive could illuminate how initial management impacts long-term morbidity and mortality and inform what the standard of care following IA should be.

Limitations of our study include the relatively small size of the cohort and the low numbers of IA deaths, making it difficult to generalize these findings. In addition, while findings from the REP have been shown to be generalizable to the nation as a whole,¹⁴ it is important to note that data came from one county in Minnesota and may not generalize to areas with different cultural or sociodemographic backgrounds. Our cohort includes only those early adolescents who came to medical attention, thus failing to include those who did not receive care and likely underestimating the actual rate.¹² As we state in the introduction, suicide attempts dramatically increased in the United States in the 10–14 age group between 2001 to 2021.² As our study enrollment period straddles three decades, it is conceivable that subjects recruited in the 80s and 90s for our study differ from the 2000s; however, the sample size did not permit time trend comparisons.

Our findings have several public health implications. Means restriction – both firearms and pills – could affect both mortality and morbidity in early adolescents, just as it does in older people. Two of three IA deaths occurred by gunshot. Approximately 80% of medical hospitalizations stemmed from medication overdoses. Making it more difficult for early adolescents to access guns and pills could reduce mortality and morbidity respectively. For nearly half of this cohort, their IA was their first psychiatric presentation. Suicide prevention efforts should focus on identifying at risk youth and getting them the psychiatric care they need before they make a first attempt.