Cutoff scores for suicidal ideation using the Japanese interpersonal needs questionnaire-15: a cross-sectional study in a high suicide mortality region

Abstract

Introduction

Suicide prevention is critical in regions with high suicide mortality rates, such as Akita Prefecture, Japan, requiring effective tools for identifying suicidal ideation. The 15-item Interpersonal Needs Questionnaire (INQ-15) measures perceived burdensomeness (PB) and thwarted belongingness (TB). The statistically derived cutoff scores for the Japanese version (INQ-15J) remain unexamined.

Objectives

To determine cutoff scores for the INQ-15J to identify individuals with suicidal ideation and evaluate screening accuracy based on population data.

Methods

A total of 1,650 residents aged 20–79 years (response rate: 33.0%) completed the INQ-15J. Receiver operating characteristic analyses identified cutoff scores for PB and TB for suicidal ideation in the past month and lifetime. Multiple logistic regression assessed the independent associations of PB and TB with suicidal ideation. Screening performance was further examined using parallel (high-PB or high-TB) and serial (high-PB and high-TB) applications.

Results

The statistically derived cutoff scores (PB = 21; TB = 34) showed moderate accuracy for identifying suicidal ideation in the past month. PB and TB were independently associated with suicidal ideation. The parallel application increased sensitivity, whereas the serial application increased specificity.

Conclusion

The INQ-15J offers a practical, indirect approach for identifying individuals with suicidal ideation in community settings. Combining PB and TB cutoff scores allows flexibility depending on whether sensitivity or specificity is prioritized. Further validation in diverse populations is warranted.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-026-08056-3.

Introduction

Suicide prevention is an important global public health issue. Japan has one of the highest suicide mortality rates among the Organization for Economic Cooperation and Development member countries, with suicide countermeasures implemented as national policies.

Gatekeeper (GK) training has become a key strategy adopted by many countries [1, 2], and the importance of GK training is emphasized in the Comprehensive Suicide Prevention Policy in Japan [3]. The role of GK is to recognize and appropriately respond to signs of suicide risk. If a gatekeeper identifies an individual at a high risk of suicide, it is recommended that they ask about the presence of suicidal ideation. However, the most challenging aspect of the role of GKs is the identification of suicidal ideation. Even nurses who are proficient in interviewing patients find it difficult to ask directly about suicidal ideation [4]. Since the introduction of GK training, Japan has focused more on community residents than on professionals, unlike other countries [5]. Indirect and easy measures are needed to assess suicidal ideation.

According to Joiner’s interpersonal theory of suicide, suicidal ideation arises when two interpersonal states become severe and persistent. The first is perceived burdensomeness (PB), defined as the belief that one is a burden to others. The second is thwarted belongingness (TB), a sense of social disconnection or lack of meaningful relationships [6]. Within this theory, PB and TB are conceptualized as proximal causes of suicidal desire, and when they co-occur at elevated levels, individuals are theorized to be more likely to experience suicidal ideation. These constructs therefore play a central role in suicide risk assessment.

To assess PB and TB, the Interpersonal Needs Questionnaire (INQ) was developed as a theory-based instrument [7]. The INQ provides an indirect method of evaluating suicide risk by measuring these interpersonal constructs without directly asking about suicidal ideation. Several versions of the INQ exist, and the 15-item version (INQ-15) is the most widely used [6–9], with validated translations in multiple languages. The INQ-15 has demonstrated strong psychometric properties, including test–retest reliability, content and structural validity, and external validity through associations with loneliness and perceived social support [9–12]. Additionally, previous studies have proposed cutoff scores for identifying suicidal ideation in clinical populations, such as individuals with mental disorders or cancer [13–15].

Although the reliability and validity of the Japanese version of the INQ-15 (INQ-15 J) have been established [10], cutoff scores for identifying individuals with suicidal ideation have not yet been examined. Therefore, this study aimed to determine statistically derived PB and TB cutoff scores for the INQ-15 J and evaluate their screening performance, with particular attention to their accuracy in the general population. Establishing these cutoffs will provide gatekeepers and professionals with a practical and indirect tool for assessing suicide risk and guiding appropriate interventions.

Methods

Participants

This study was conducted in Akita Prefecture, northern Japan, with a population of 700,000 people aged 18–79 years, in 2022. Akita faces various challenges, including economic hardship, rural depopulation, and a rapidly aging population. Long snowfall and limited daylight during harsh winters exacerbate social isolation and mental health issues, particularly among older adults. Akita is one of the most rapidly aging prefectures in Japan, with the highest suicide mortality rate among all prefectures in Japan (Japan vs. Akita = 17.4 vs. 22.6 per 100,000 of the population in 2022) over the past two decades [16]. However, Akita is also known for its history of active suicide prevention measures and has significantly contributed to the establishment of the Basic Act on Suicide Prevention in Japan [17].

The authors conducted a questionnaire survey among Akita residents in August 2022. The target population was residents aged 20–79 years, excluding inpatients and institutional residents, such as individuals in nursing homes or welfare facilities. A total of 5,000 individuals were randomly selected across the 25 municipalities of Akita Prefecture using stratified random sampling by sex and age. The number of individuals selected from each municipality was proportional to its population size. The required sample size was calculated based on the population composition by sex and age using national demographic data from August 2022.

Participants were provided with a written explanation of the study’s purpose, procedures, and self-administered questionnaires. They anonymously returned the completed questionnaires. Among the 2,500 (50.0%) responses, we analyzed 1,650 participants (33.0%) who completed the INQ-15, including data on lifetime and past-month suicidal ideation. This study was conducted in accordance with the ethical standards of the Declaration of Helsinki and approved by the Ethics Committee of Akita University (approval number: 2920). This study was conducted in accordance with STROBE guidelines.

Measures

INQ-15J

The INQ-15J was used to measure two constructs: PB and TB. It consists of 15 items, six for PB and nine for TB, rated on a 7-point Likert scale (1 = Not at all true for me; 7 = Very true for me). Higher scores indicate higher PB and TB levels. Cronbach’s alphas for PB and TB were 0.953 and 0.722, respectively.

Suicidal ideation

Suicidal ideation was evaluated using two questions. The first was “Have you ever had thought of ending your life at any time?” (Lifetime suicidal ideation), and the second was “Have you had the thought of ending your life in the past month?” (Suicide ideation in the past month). Responses of Yes and No were rated as 1 and 0, respectively. These items have been validated in previous studies involving nonclinical populations [18].

Data analysis

To determine the statistically derived cutoff scores of PB and TB for identifying suicidal ideation in the past month and lifetime, a receiver operating characteristic (ROC) curve was used. Youden’s index was used to determine the thresholds that maximize the balance between sensitivity and specificity.

Additionally, multiple logistic regression analysis was performed to examine the associations of PB and TB with suicidal ideation. We calculated the odds ratios (OR), and 95% confidence intervals (CI) adjusted for potential confounders, including sex, age, and marital status.

Finally, based on these results, we explored the screening performance of PB and TB using two approaches: a parallel application (high-PB or high-TB) and a serial application (high-PB and high-TB). This complementary analysis provides further insights into how these combinations affect the trade-off between sensitivity and specificity, offering additional guidance for optimizing screening strategies. All statistical analyses were performed using the IBM SPSS Statistics version 28 (IBM SPSS Software Group, Chicago, IL, USA). Two-sided tests were applied, with statistical significance set at p < 0.05.

Results

Compared with the overall population of Akita Prefecture [19], respondents showed a similar distribution by sex (male: 47.5% vs. 47.2%; female: 52.5% vs. 52.8%). However, respondents included fewer individuals in their 20s (4.7% vs. 6.9%) and 30s (7.8% vs. 10.1%), and more individuals in their 60s (28.4% vs. 15.2%) and 70s (22.3% vs. 15.9%). Additional details are provided in Supplementary Table 1.

Of the 1,650 participants (53.0% female), 75 (4.4%) reported suicidal ideation in the past month, and 457 (26.9%) reported lifetime suicidal ideation (Table 1).

Table 1.

Basic attributes of respondents (n = 1650)

Variables	Category	n	%
Sex	Male	794	(47.5)
	Female	866	(52.5)
Age	20–29 years	78	(4.7)
	30–39 years	128	(7.8)
	40–49 years	258	(15.6)
	50–59 years	350	(21.2)
	60–69 years	468	(28.4)
	70–79 years	368	(22.3)
Marital Status	Married	1133	(68.7)
	Unmarried	271	(16.4)
	Divorced	105	(6.4)
	Widowed	112	(6.8)
	Others or missing	29	(1.8)

Mean (SD) PB score was 14.2 (0.17), and the mean (SD) TB score was 29.8 (0.17). PB and TB score distributions were examined descriptively to provide an overview of the data

The ROC curve for suicidal ideation in the past month is shown in Fig. 1. The area under the curve (AUC) for PB and TB scores were 0.75 and 0.73 (95% CI 0.69–0.81 and 0.67–0.80), respectively. The statistically derived cutoff scores for the INQ-15J were PB ≥ 21 and TB ≥ 34 for identifying individuals with suicidal ideation in the past month.

Fig. 1.

ROC curves of perceived burden and thwarted belongingness in suicidal ideation in the past month. Abbreviations: ROC: Receiver operating characteristic, AUC: Area under the curve. 95%CI: 95%Confidence interval

The ROC curve for lifetime suicidal ideation is shown in Fig. 2. The AUC for PB and TB was 0.67 (95% CI 0.64–0.70) and 0.63 (95% CI 0.60–0.66), respectively. The statistically derived cutoff scores for lifetime suicidal ideation were PB ≥ 15 and TB ≥ 33.

Fig. 2.

ROC curves of perceived burdensomeness and thwarted belongingness for lifetime suicidal ideation. Abbreviations: ROC: Receiver operating characteristic, AUC: Area under the curve. 95%CI: 95%Confidence interval

Table 2 shows the results of the multiple logistic regression analyses. Multiple logistic regression models were constructed to examine whether high-PB and high-TB were associated with suicidal ideation in the past month and lifetime, adjusting for sex, age, and marital status. The models demonstrated acceptable fit (past-month: Nagelkerke R² = 0.15; Hosmer–Lemeshow p = 0.55; lifetime: Nagelkerke R² = 0.09; Hosmer–Lemeshow p = 0.39). High-PB and high-TB were significantly associated with suicidal ideation in the past month (PB: OR = 3.42, 95% CI = 1.95–6.02; TB: OR = 3.18, 95% CI = 1.79–5.62). For lifetime suicidal ideation, similar significant associations were also observed (PB: OR = 2.22, 95% CI = 1.72–2.87; TB: OR = 1.73, 95% CI = 1.34–2.24).

Table 2.

Association between INQ-15 cutoff scores and suicidal ideation (n = 1650)

Explanatory variables		Suicidal ideation in the past month		Lifetime suicidal ideation
		OR (95%CI)	p-value	OR (95%CI)	p-value
PB	High	3.42 (1.95–6.02)	< 0.001	2.22 (1.72–2.87)	< 0.001
	Low	1.00 (ref)		1.00 (ref)
TB	High	3.18 (1.79–5.62)	< 0.001	1.73 (1.34–2.24)	0.01
	Low	1.00 (ref)		1.00 (ref)
Sex	Male	1.00 (ref)		1.00 (ref)
	Female	0.60 (0.36–0.99)	0.05	0.74 (0.58–0.93)	0.01
Age	20–29 years	1.00 (ref)		1.00 (ref)
	30–39 years	3.87 (1.29–11.6)	0.02	1.07(0.58–1.98)	0.82
	40–49 years	1.52 (0.56–4.09)	0.41	1.14 (0.71–1.98)	0.58
	50–59 years	1.04 (0.44–2.46)	0.93	1.31 (0.90–1.91)	0.16
	60–69 years	1.11 (0.50–2.47)	0.80	1.09 (0.77–1.56)	0.63
	70–79 years	0.84 (0.38–1.83)	0.65	0.90 (0.64–1.26)	0.54
Marital Status	Married	1.00 (ref)		1.00 (ref)
	Unmarried	0.61 (0.13–2.87)	0.53	1.01 (0.43–2.37)	0.99
	Divorced	0.29 (0.06–1.53)	0.14	1.01 (0.41–2.47)	0.99
	Widowed	1.11 (0.19–6.41)	0.91	1.23 (0.50–3.29)	0.69
	Others or missing	0.77 (0.14–4.31)	0.77	1.29 (0.50–3.29)	0.60

Cutoff definitions: For suicidal ideation in the past month, high-PB = PB ≥ 21 and high-TB = TB ≥ 34

For lifetime suicidal ideation, high-PB = PB ≥ 15 and high-TB = TB ≥ 33

Low-PB and Low-TB were defined as scores below their respective cutoff values

Model specification: Odds ratios were estimated using multiple logistic regression with suicidal ideation as the outcome

(1 = present, 0 = absent), including PB and TB as explanatory variables and adjusting for sex, age, and marital status

Abbreviations: PB, perceived burdensomeness; TB, thwarted belongingness; OR, odds ratio; CI, confidence interval; Ref, reference category

The sensitivity and specificity of any combination of PB and TB cutoff scores were calculated for suicidal ideation in the past month and lifetime suicidal ideation (Table 3). For past-month suicidal ideation, high-PB showed a sensitivity of 58.7% and a specificity of 80.8%, while high-TB showed a sensitivity of 66.7% and a specificity of 71.8%. The parallel application yielded a sensitivity of 78.7% and a specificity of 66.2%. The serial application increased specificity to 86.4%, while sensitivity decreased to 46.7%.

Table 3.

Sensitivity and specificity of INQ-15 cutoff scores for suicidal ideation (n = 1650)

Explanatory Variables	Suicidal ideation in the past month		Lifetime suicidal ideation
	Sensitivity (%)	Specificity (%)	Sensitivity (%)	Specificity (%)
High-PB	58.7	80.8	55.6	70.0
High-TB	66.7	71.8	49.9	71.5
Parallel application	78.7	66.2	66.3	59.1
Serial application	46.7	86.4	39.2	82.4

Cutoff definitions: For suicidal ideation in the past month, high-PB = PB ≥ 21 and high-TB = TB ≥ 34

For lifetime suicidal ideation, high-PB = PB ≥ 15 and high-TB = TB ≥ 33

Screening approaches: Parallel application = high-PB OR high-TB

Serial application = high-PB AND high-TB

Abbreviations: PB, perceived burdensomeness; TB, thwarted belongingness; SI, suicidal ideation

For lifetime suicidal ideation, the sensitivity and specificity of high-PB were 55.6% and 70.0%, respectively, and those of high-TB were 49.9% and 71.5%, respectively. Under the parallel application, sensitivity and specificity were 66.3% and 59.1%, respectively, whereas the serial application increased specificity to 82.4% but reduced sensitivity to 39.2%.

Discussion

This study examined the screening utility of statistically derived cutoff scores for the INQ-15J in identifying individuals with suicidal ideation in a high-suicide-mortality region. High-PB and high-TB were significantly associated with suicidal ideation in both the past month and lifetime. The cutoff scores demonstrated moderate accuracy for detecting past-month suicidal ideation, whereas the discriminatory ability for lifetime suicidal ideation was limited. These findings suggest that PB and TB may serve as practical indirect indicators for identifying individuals with recent suicidal ideation in community settings.

The distribution of PB and TB scores in the INQ-15 J has not been reported in previous studies [11]. In this study, PB and TB score distributions were examined descriptively to provide an overview of the data, but they were not used to guide cutoff determination.

Compared with the AUC for suicidal ideation in the past month, the AUC for lifetime suicidal ideation was lower. This suggests that the INQ-15 J may be more sensitive to recent suicidal ideation than to ideation reported over longer periods. One possible explanation is that PB and TB may capture more immediate or situation-dependent interpersonal distress, which may be more closely aligned with recent suicidal ideation than with ideation recalled across extended time frames. In addition, recall bias may contribute to reduced accuracy in lifetime reports of suicidal ideation [20, 21].

Our findings supported the usefulness of the INQ-15 J in identifying individuals with past-month and lifetime suicidal ideation in the general population. These results are consistent with the interpersonal theory of suicide [6], which posits that PB and TB are key interpersonal constructs associated with suicidal ideation. Consistent with previous studies [6, 7, 13, 14], high-PB and high-TB each showed independent associations with suicidal ideation.

The cutoff values identified in the present study for suicidal ideation were broadly consistent with those reported in previous studies. Prior studies have reported specific PB and TB cutoff values across different clinical populations. For example, among women with chronic, invisible illnesses, cutoff values of PB ≥ 17 and TB ≥ 37 have been suggested. In psychiatric populations, PB ≥ 12 has been reported across both inpatients and outpatients, while TB cutoff values differ by setting, with TB ≥ 36 in outpatients and TB ≥ 32 in inpatients. Additionally, other studies have reported PB and TB cutoff values of approximately 22 and 31, respectively, for identifying suicidal ideation [13–15]. In the present study, the thresholds were PB ≥ 21 and TB ≥ 34 for past-month suicidal ideation, and PB ≥ 15 and TB ≥ 33 for lifetime suicidal ideation, which fall within these previously reported ranges. This consistency suggests that PB and TB thresholds associated with suicidal ideation may be relatively consistent across different populations. Although this is the first study to examine INQ-15 J cutoff scores in a general-population sample in Japan, the observed similarities with international studies suggest that elevated PB and TB may reflect interpersonal difficulties associated with suicidal ideation across different populations. However, further research in other regions and cultural contexts will be necessary to evaluate the generalizability and reproducibility of these cutoff scores.

The screening patterns observed in this study indicate that the parallel application of PB and TB yields higher sensitivity, whereas the serial application yields higher specificity. These trade-offs illustrate how PB and TB can be applied flexibly depending on whether the priority is to minimize missed cases or reduce false positives. In community settings where directly assessing suicidal ideation can be challenging, the parallel application may offer a practical approach to increasing the likelihood of recognizing individuals who may be experiencing suicidal ideation in the past month.

In practice, PB and TB may provide useful indirect indicators that can assist gatekeepers, practitioners, and community-based professionals in recognizing individuals who may be experiencing recent suicidal ideation. Integrating these constructs into routine assessments may help facilitate earlier engagement or referral, particularly in situations where directly asking about suicidal ideation is challenging.

Limitations

This study has several limitations that should be considered when interpreting the findings. First, it is not possible to evaluate the ability of the INQ-15J to identify changes in suicidal ideation or subsequent suicidal behaviors over time due to the cross-sectional design of this study. While the cutoff scores identified here are useful for detecting individuals with recent suicidal ideation, longitudinal studies are required to determine whether these scores are informative for understanding longer-term patterns of suicidal ideation or future suicidal behaviors.

Second, this study did not include longitudinal data to track changes in PB and TB scores over time. Consequently, it remains unclear whether variations in these scores reflect clinically meaningful improvements or deterioration. Future research should focus on examining the evolution of PB and TB scores over time, particularly in response to interventions designed to reduce suicide risk, to better understand their roles in suicidal ideation.

Third, the measurement of suicidal ideation relied on self-reported data, which are inherently subjective and may be influenced by social desirability bias or inaccurate self-perception. These factors could limit the reliability of the findings and should be addressed in future research by incorporating objective measures or multi-method approaches.

Fourth, the generalizability of the PB and TB cutoff scores to populations with different cultural, demographic, or socio-economic backgrounds remains unclear. Further validation studies across diverse populations are required to confirm their broad applicability.

Conclusions

This study demonstrated that the INQ-15J is useful for identifying individuals with both past-month and lifetime suicidal ideation in a general population sample. The statistically derived cutoff scores for PB and TB provide a practical indirect approach for recognizing individuals who may be experiencing suicidal ideation in the past month, particularly in high-risk regions such as Akita, Japan. The findings supported the interpersonal theory of suicide and highlighted the independent associations of PB and TB with suicidal ideation.

Because the INQ-15J appears to be more sensitive to suicidal ideation in the past month than to ideation reported over longer periods, longitudinal studies are needed to examine how PB and TB relate to temporal fluctuations in suicidal ideation and to evaluate the generalizability of these cutoff scores across diverse populations. Overall, the INQ-15J represents a promising indirect tool that may support broader suicide-prevention efforts in community settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

GK

Gatekeeper

PB

Perceived burdensomeness

TB

Thwarted belongingness

INQ

Interpersonal Needs Questionnaire

INQ-15

15-item Interpersonal Needs Questionnaire

INQ-15J

Japanese version of the Interpersonal Needs Questionnaire-15

ROC

Receiver operating characteristic

OR

Odds ratio

CI

Confidence interval

AUC

Area under the curve

Author contributions

Syohei Miyamoto: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resource, visualization, writing—original draft and writing—review and editing and guarantor; Atsushi Iwasawa and Takayuki Kageyama.

Funding

This study was conducted with the support of the “Regional project to improve the prevention of suicide” provided by Prefecture and JSPS KAKENHI Grant Numbers .

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the ethical standards of the Declaration of Helsinki. Ethics approval was obtained from the Institutional Review Board of the Akita University Medical School (Approval No. 2920).

Consent to participate

All participants provided written informed consent prior to participation in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.