Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jan 1.
Published in final edited form as: J Affect Disord. 2024 Sep 17;368:439–447. doi: 10.1016/j.jad.2024.09.119

Heterogeneity in momentary affective experiences related to suicidal urges in a non-clinical sample of adult handgun owners and non-owners recruited from the community

Craig J Bryan 1,2, Melanie L Bozzay 1, Jeffrey V Tabares 1, Samantha E Daruwala 1,2, Jonathan E Butner 3, Stephanie M Gorka 1
PMCID: PMC11560658  NIHMSID: NIHMS2024647  PMID: 39299584

Abstract

Although suicide is complex and heterogeneous, most suicide theories assume that suicidal urges occur primarily in the context of extreme emotional distress. Newer models of suicide based on complex systems theory propose greater heterogeneity in suicidal experiences across individuals and groups, such that some, but not all, suicidal thoughts, urges, and behaviors are associated with extreme negative affect. The present study investigated individual differences in affective states experienced during suicidal urges among 138 adults recruited from the community; 81 (59.1%) owned handguns and 57 (41.6%) did not. Participants self-reported their current affect and urge to kill themselves 6 times per day for 28 consecutive days via ecological momentary assessment. Positive and negative affect ratings varied significantly during suicidal urges. The association of positive and negative affect with suicidal urges significantly varied within and between handgun owners and non-owners. Results suggest suicidal urges are characterized by high affective heterogeneity.

Keywords: firearms, guns, suicide, ecological momentary assessment, complex systems

Suicide is one of the leading causes of death in the United States. From 1999 to 2021, the U.S. suicide rate increased by 33% (Curtin et al., 2023), despite considerable investment to understand and prevent suicide. Efforts to prevent suicide are hampered by the commonplace view of suicidal behavior as a homogeneous outcome caused by a finite set of causes or mechanisms (Franklin, 2019; Franklin et al., 2017; Stanley et al., 2021). Suicidal behavior is heterogeneous and complex, however. It can occur without much forethought or after careful and deliberate planning (Chaudhury et al., 2016; Conner, 2004; Conner et al., 2007; Jeon et al., 2010), in response to a life stressor or the absence of any obvious stressor (Apter et al., 1993; Bryan et al., 2015), or in the presence or the absence of mental illness (Caves Sivaraman et al., 2022; Martin et al., 2020; Stanley et al., 2021). Suicidal behavior is also associated with a wide range of intrapsychic experiences including but not limited to the desire to die (“passive” suicidal ideation), considering ways to end one’s life (“active” suicidal ideation), developing specific plans about where and when to attempt suicide, and/or experiencing urges to kill oneself that can occur in combination or in isolation (Bryan et al., 2022; Millner et al., 2015; Substance Abuse and Mental Health Services Administration, 2023; Wastler et al., 2022). These various manifestations of suicidal states, defined as the experience of suicide-related thoughts and/or behaviors at a particular moment in time (American Psychological Association, 2024), are also marked by considerable heterogeneity.

Research further shows that suicidal states are highly dynamic, with different facets and dimensions following different time courses (Butner et al., 2021). The wish to live and the wish to die, for example, can sometimes fluctuate independent of one another but other times fluctuate together in a coordinated manner (Bryan et al., 2016). The wish to die also tends to be more stable over time and follows a more gradual change process than suicidal intent, which tends to come and go suddenly within very short timeframes (Coppersmith et al., 2022). The multidimensional nature of suicidal states may explain findings that many people who attempt suicide do not experience suicidal thinking in advance (Bryan et al., 2021; Bryan, Allen, et al., 2023; Richards et al., 2019; Simon et al., 2016; Simon et al., 2013; Simon et al., 2001; Substance Abuse and Mental Health Services Administration, 2023; Wastler et al., 2022; Wyder & De Leo, 2007), highlighting the need to better understand suicidal urges—sudden or compelling impulses to enact suicidal behavior that can motivation action without deliberation or forethought (cf. “impulse”; American Psychological Association, 2024).

The multitude of ways in which suicidal behaviors can emerge have led researchers to argue that suicidal states are better understood as complex adaptive systems composed of multiple interacting and interdependent components that can potentially differ with respect to their cognitive, affective, physiological, and behavioral features (Bryan, Butner, et al., 2020). Consistent with this perspective, Rudd (2006) introduced the concept of the suicidal mode as a framework for understanding how the cognitive, emotional, physiological, and behavioral components of suicidal states interact with one another and operate as a unified network. This network is activated by external stressors (e.g., relationship problems, financial strain), giving rise to suicidal urges. The suicidal mode does not require any particular set of cognitive, emotional, physiological, and behavioral combinations; many different possible combinations of these components could lead to suicidal urges. This aligns with Franklin’s (2019) argument that suicidal states, broadly defined, will vary because individual differences in how people interpret their internal affective experiences and their external circumstances relative to their personal conceptualization of suicide.

One implication of these complexity-based models is that suicidal urges can be associated with a much broader range of affective states than is typically assumed, to include the absence of negative affect and possibly even the presence of positive affect (Franklin, 2019). This view departs from most contemporary theories of suicide, which generally assume that suicidal thoughts, urges, and behaviors occur primarily, if not solely, within the context of extreme negative affect (Klonsky & May, 2015; O’Connor & Kirtley, 2018; Van Orden et al., 2010). Consistent with these theories, studies utilizing ecological momentary assessment (EMA) have found that negative affect is elevated whereas positive affect is reduced when people are thinking about suicide or experiencing suicidal urges and the time leading up to these moments (Armey et al., 2020; Humber et al., 2013; Husky et al., 2017; Kleiman et al., 2018; Victor et al., 2021; Victor et al., 2019).

A key limitation of previous studies is their emphasis on averaged effects across entire samples with little to no consideration for how the relationships between affect and suicidal urges might vary within and between individual participants or sample subgroups. For some people, “feeling suicidal” may be associated with extreme negative affect (Armey et al., 2020; Kleiman et al., 2018) but for others, “feeling suicidal” may have little to no association with the affective intensity or be poorly explained by affect. Moreover, although few studies have examined the role of positive affect in momentary suicide risk, there is some evidence that “feeling suicidal” can also be related to reduced positive affect (Schatten et al., 2021; Tian et al., 2017) and the relative (im)balance of negative and positive affect (Armey et al., 2020).

Previous studies have also focused primarily on the valence (i.e., negative-positive) dimension of affect, an approach that likely provides limited insight into the role of affective states in suicide risk. Many prominent structural models of affect also incorporate an arousal dimension of affect (Larsen & Diener, 1992; Posner et al., 2005; Russell, 1980; Watson & Tellegen, 1985; Yik et al., 2011) that captures the degree of energy present in the affective state and is independent of positive or negative valence (see Yik et al., 2011 for a discussion). Capturing both dimensions in suicide research is important, as studies implicate somewhat dissociable neurobiological pathways underlying each dimension (see Posner et al., 2005 for a review) that could be important for understanding neurally mediated risk pathways. Moreover, information about the overlap of arousal and valence may permit a more nuanced and useful interpretation of affective states.

In research, these two dimensions are often depicted using a two-dimensional Cartesian plane, but they can alternatively be represented as a circle, consistent with a circumplex structure (Figure 1), which provides the advantage of representing similar but slightly different facets of affective experience with greater clarity and nuance (Yik et al., 2011). Anxiety and sadness, for example, are two negative affective states differentiated by high versus low arousal, respectively. Likewise, excitement and contentment are two positive affective states similarly distinguished by high versus low arousal. Anxiety and excitement, by comparison, are two high arousal states distinguished by negative versus positive valence. The circumplex model also provides for more subtle distinctions: anxiety and fear are similar negative affective states that differ in degree of arousal (fear is characterized by greater arousal). Despite these advantages of the circumplex model, to our knowledge these two dimensions of affect have yet to be simultaneously examined in studies focused on understanding suicidal urges. Considering both dimensions of affect could advance our understanding of suicidal urges specifically and suicidal states more generally. As an example, high activation is thought to be an important factor that characterizes acute suicidal affective states (Rogers et al., 2023). High arousal could therefore have different implications for suicide risk when present in positively valenced (i.e., excited) versus negatively valenced (i.e., agitated) states or within similarly valenced affective states (i.e., anxiety versus fear, unhappiness versus sadness).

Figure 1.

Figure 1

Open in a new tab

Schematic diagram of the 12-point core affect circumplex (cf. Yik et al., 2011)

Circumplex models further posit that cognitive interpretations of internal states impact how they are subjectively perceived (Posner, 2005). Variability in these subjective interpretations could further contribute to heterogeneity in suicidal states. Indeed, a recent discussion by Franklin (2019) emphasized that cultural and group differences can impact the individual experience of heightened suicide risk because suicide concepts are influenced by sociocultural norms and beliefs. What it means to “feel suicidal” could therefore differ by sex, gender identity, race, sexual orientation, and nationality. Supporting this perspective is research suggesting the suicidal experiences of firearm owners differ markedly from those of non-owners. Firearm suicide decedents are less likely to have a mental illness or prior suicide attempt (Centers for Disease Control and Prevention, 2005). Firearm owners are also less likely than non-owners to report experiencing suicidal ideation in the time surrounding suicidal behavior (Bryan et al., 2022). These preliminary findings suggest that the suicidal experiences of firearm owners differ from those of non-owners, implicating the need for different preventative strategies. Because US firearm owners are at higher risk of suicide than their non-owner peers (Barber et al., 2017), additional comparative research focused on suicidal urges rather than suicidal thinking and heterogeneity in affective experiences among firearm owners and non-owners is warranted.

Present study

This study examined heterogeneity in suicidal urges, assessed multiple times per day for 4 consecutive weeks via EMA, across handgun owners and non-owners. We focused specifically on handgun ownership in this study instead of firearm ownership more broadly because handguns are used approximately twice as often as long guns in U.S. suicides (Hanlon et al., 2019; Planty & Truman, 2013). In contrast to previous studies that focused primarily on affective valence (i.e., positive versus negative affective), we assessed both affective valence and various levels of affective arousal, based on the circumplex model. We hypothesized that (1) affect scores would significantly vary when people were experiencing suicidal urges, (2) the association of affect with suicidal urges would significantly vary, and (3) the association of affect with suicidal urges would significantly differ between handgun owners and non-owners.

Method

Participants and Procedures

This project uses data collected as part of a study aimed at identifying biobehavioral processes that increase the risk of suicide among handgun owners. Participants included 138 adults recruited from the local community who were (1) 18+ years old; (2) able to complete an in-person laboratory visit; and (3) owners of a smartphone (either an Android- or Apple-based operating system) who were willing to receive EMA alerts via smartphone app. Owing to several laboratory-based procedures that are not reported in this analysis (e.g., startle reactivity), the following exclusion criteria were employed: (1) serious medical conditions that could interfere with data interpretation (e.g., deafness, moderate or severe traumatic brain injury, or lifetime mania or psychosis); (2) psychotropic medication use within the past 4 months; (3) acute alcohol intoxication (verified via breath test); and (4) heavy recreational alcohol or cannabis use, defined as 5+ alcohol binges per month and cannabis use more than 5 times per week. Participants first completed informed consent procedures. After providing consent, participants completed an eligibility assessment. If eligible, they completed a self-report assessment battery and laboratory-based tasks. At the end of the laboratory visit (approximately 60–90 minutes in duration), a researcher helped participants download a smartphone app for EMA data collection and provided a brief tutorial. Participants received 6 EMA alerts per day for 28 consecutive days at pseudo-random times between 8:00 AM and 10:00 PM, with a median interval between surveys of 155 minutes (95% CI=143–167 minutes). Each survey took less than 5 minutes to complete. At the end of the lab visit, all participants, regardless of reported level of suicide risk, collaboratively created a crisis response plan (CRP) with a researcher. The CRP is a safety planning-type intervention that has been shown to reduce suicidal behaviors (Bryan et al., 2024; Bryan et al., 2017). Participants reporting severe suicidal urges (scores >7 on a 0–10 scale) were contacted by an on-call member of the study team to complete a risk assessment and review their CRP. Study procedures were approved by The Ohio State University Biomedical Institutional Review Board.

Instruments

Suicidal Urges.

Suicidal urges were assessed 6 times per day via EMA using the following item that has been used previously in EMA studies of suicide risk (Coppersmith et al., 2022; Glenn et al., 2021; Glenn et al., 2020; Kleiman et al., 2017): “Right now, how strong is your urge to kill yourself?” This item was rated on a 0 (not strong at all) to 10 (very strong) scale. The concurrent and predictive validity of single-item measures in EMA research have been demonstrated (Song et al., 2022). The intraclass correlation of this item in this sample was 0.80.

Affect.

Momentary affectivity was assessed 6 times per day via EMA using the Positive and Negative Affect Scale Short Form (PANAS-SF) (Watson et al., 1988). Respondents were directed to “indicate to what extent you feel [affective state] right now” using a Likert rating scale ranging from 1 (very slightly or not at all) to 5 (extremely) for the 10 positive affect (PA) states (interested, excited, strong, enthusiastic, proud, alert, inspired, determined, attentive, active) and 10 negative affect (NA) states (distressed, upset, guilty, scared, hostile, irritable, ashamed, nervous, jittery, afraid). The PANAS-SF’s internal consistency and construct validity are established (Crawford & Henry, 2004).

In this study, we used PANAS-SF items to create the core affect scores based on Yik et al.’s (2011) circumplex model, which provides a circular structure that positions similar affect states closer to each other and divergent affect states farther apart (see Figure 1). Embedded within the circumplex model are the two dimensions of valence (negative on the left, positive on the right) and arousal (low on the bottom, high on the top). The 20 PANAS-SF items correspond to four of Yik et al.’s 12 core affective states: pleasant activation, which includes positive affect states with high arousal (interested, excited, alert, inspired, attentive, active, determined); activated pleasure, which includes positive affect states with moderate arousal (strong, enthusiastic, proud); unpleasant activation, which includes negative affect states with high arousal (nervous, jittery); and activated displeasure, which includes negative affect states with moderate arousal (distressed, upset, guilty, hostile, irritable, ashamed, scared, afraid).

Prior Suicide Risk.

Participants completed the Self-Injurious Thoughts and Behaviors Interview-Revised (Fox et al., 2020) to assess lifetime history of suicide ideation and suicide attempts. Prior suicide attempts include aborted, interrupted, and actual attempts.

Data Analysis

To assess individual differences in affect during suicidal urges (hypothesis 1), we constructed separate mixed effects regression models with momentary pleasant activation, activated pleasure, unpleasant activation, and activated displeasure as the outcomes, momentary suicidal urge as a binary (endorsed vs. not endorsed) fixed effect predictor variable, and random intercepts estimated separately for the endorsed and non-endorsed groups. Mixed effects modeling was used because they are appropriate for repeated measurements within participants, allow for specification of both fixed effects (i.e., overall averages) and random effects (i.e., individual differences), and handling missing data (Enders, 2001). Random effects were estimated using a variance component structure.

Because the random intercept estimate (γ0) represents variance in the outcome around the mean fixed effect (β0), approximately 95% of momentary affect scores during a suicidal urge can be calculated using the equation y = β0 ± 1.96*sqrt(γ0). To determine if the individual differences in mean affect across participants differed when experiencing suicidal urges versus not, we conducted a test of homogeneity, which compares model fit when the random intercept is estimated separately for each group relative to the model with a random intercept aggregated across all groups. A statistically significant test indicates that the mean affect score varies more across participants during one state than the other.

To assess individual differences in the association between momentary affect score and severity of suicidal urges across participants (hypothesis 2), we constructed separate mixed effects regression models with momentary suicidal urge score as the outcome, momentary affect state as a fixed effect predictor variable, and random intercepts estimated for the intercept and slopes (i.e., affect score). Because the random slope estimate (γi) represents variance in the outcome around the mean fixed effect (βi), approximately 95% of momentary affect score slopes can be calculated using the equation y = βi ± 1.96*sqrt(γi).

To compare the strength of association between affect score and suicidal urge between handgun owners and non-owners (hypothesis 3), we ran revised models used for hypothesis 2 that included two additional fixed effect predictors: handgun ownership group and an interaction term for affect * handgun ownership group. To determine if the individual differences in the strength of association between affect and suicidal urge differed in magnitude between handgun owners and non-owners, we conducted a test of homogeneity. A statistically significant test indicates that the strength of association between affect and suicidal urge varies more across participants in one group than the other.

Missing data were handled using maximum likelihood estimation. All analyses were conducted using SAS version 9.4. Owing to the positive skew of momentary suicidal urge scores and the significant sex difference across handgun ownership group, we repeated the analyses for hypotheses 2 and 3 with log-transformed momentary suicidal urge scores as the outcome, with sex entered as a covariate, and with negative binomial mixed effects modeling. The overall pattern of results and conclusions did not differ from the unadjusted models using untransformed scores, however. We therefore report the analyses using untransformed data for ease of interpretation.

To estimate statistical power, we conducted a series of a priori Monte Carlo simulations for a multilevel model with three simultaneous predictors and the inclusion of all possible random effects. Given findings from previous EMA studies (Kleiman et al., 2017), we assumed an intraclass correlation (ICC) of 0.5, an additional 0.1 in the ICC for the estimation of the random effects, and a “worst case” scenario of 50% missing data. Under these conditions, power exceeded 94% to detect small effect sizes. After data for this sample were collected, the resulting ICC was 0.22 and data were missing from only 11.0% of EMA observations. Because statistical power is inversely related to ICC and missingness, statistical power in this sample was therefore better than initially estimated.

Results

Sample descriptives are summarized in Table 1. Participants were predominantly White and non-Latino/Hispanic, and were nearly evenly divided across males and females. Approximately 30% of the sample had a lifetime history of prebaseline suicidal ideation or a suicide attempt, and multiple handguns or pistols in the home (M=2.1, SD=3.7). Suicidal urges (i.e., non-zero values of the urge to kill oneself) were endorsed by 42 (31.8%) participants during 431 of 20,505 (2.1%) EMA observations. The mean momentary suicidal urge score (possible range=0–10) across all EMA observations (i.e., inclusive of observations where the urge to kill oneself was endorsed and not endorsed) was 0.04 (95% confidence interval [CI]=0.02 to 0.07); during suicidal urges (i.e., observations where the urge to kill oneself was endorsed), the mean momentary suicidal urge score was 2.30 (95% CI=1.57–3.03). Longitudinal intercorrelations among the four affect states are summarized in Table 2. Pleasant activation and activated pleasure displayed statistically significant positive concurrent and time-lagged correlations; the same was found for unpleasant activation and activated displeasure. Positive and negative affect scores had small concurrent and time-lagged correlations with each other, however.

Table 1.

Sample characteristics

Full Sample (n=147) Handgun Owners (n=81) Non-Owners (n=57) t / χ2 p
Age, M (SD) 36.9 (15.0) 35.9 (11.8) 36.9 (15.0) 0.4 0.664
No. of Firearms in Home, M (SD)
 Handguns & Pistols 2.1 (3.7) 3.5 (4.2) 0.2 (0.6) 5.9 <.001
 Shotguns 0.7 (1.3) 1.1 (1.5) 0.2 (0.6) 4.2 <.001
Rifles 1.1 (2.7) 1.7 (3.3) 0.2 (0.5) 3.5 <.001
 Sex, n (%) 5.4 0.020
 Male 72 (52.2) 49 (60.5) 23 (40.4)
 Female 66 (47.8) 32 (39.5) 34 (59.6)
 Race, n (%) 1.9 0.750
 White 101 (74.3) 62 (76.5) 39 (70.9)
Black 19 (14.0) 11 (13.6) 8 (14.5)
 Asian 9 (6.6) 4 (4.9) 5 (9.1)
 Native American 1 (0.7) 1 (1.2) 0 (0.0)
Multiracial 6 (4.4) 3 (3.7) 3 (5.5)
 Latino/Hispanic, n (%) 0.1 0.770
 No 121 (88.3) 71 (87.7) 50 (89.3)
 Yes 16 (11.7) 10 (12.3) 6 (10.7)
Prior Suicide Risk, n (%) 0.5 0.798
 Prior Suicide Attempt 14 (10.1) 9 (11.1) 5 (8.8)
 Prior Suicidal Ideation 29 (21.0) 18 (22.2) 11 (19.3)
 None 95 (68.8) 54 (66.7) 41 (71.9)
Open in a new tab

M=mean, SD=standard deviation

Table 2.

Regression coefficients from mixed effects modeling of concurrent and time-lagged intercorrelations among affective states across repeated ecological momentary assessment (EMA) observations

Affective State (Time Point) 1 2 3 4
1 Pleasant Activation (t) --
2 Activated Pleasure (t) 0.69*** --
3 Unpleasant Activation (t) 0.15*** 0.03** --
4 Activated Displeasure (t) −0.11*** −0.28*** 0.68*** --
5 Pleasant Activation (t+1) 0.38*** 0.35*** 0.04*** −0.02
6 Activated Pleasure (t+1) 0.32*** 0.37*** 0.02** −0.03**
7 Unpleasant Activation (t+1) 0.05*** 0.02 0.31*** 0.19***
8 Activated Displeasure (t+1) −0.06*** −0.12*** 0.33*** 0.39***
Open in a new tab

Note:

*

p<.05,

**

p<.01,

***

p<.001;

t=score at the current time point; t+1=score at the next (i.e., time-lagged) time point. Because all variables used a common metric, with scores ranging from 1 to 5, regression coefficients range from −1 to +1 and can be interpreted like other correlation coefficients.

When split by group, there were 81 handgun owners and 57 non-owners. Handgun owners were significantly more likely than non-owners to be male (60.5% vs. 40.4%, χ2(1)=5.4, p=.020). As expected, handgun owners reported keeping a significantly larger number of all types of firearms in their homes than non-owners: handguns and pistols (M=3.5, SD=4.2 vs. M=0.2, SD=0.6, t(136)=5.9, p<.001), shotguns (M=1.1, SD=1.5 vs. M=0.2, SD=0.6, t(136)=4.2, p<.001), and rifles (M=1.7, SD=3.3 vs. M=0.2, SD=0.5, t(136)=3.5, p<.001). Handgun owners and non-owners did not differ with respect to prebaseline suicidal ideation and suicide attempts; approximately one in five (21.0%) participants reported prebaseline suicidal ideation and one in ten (10.1%) reported a prebaseline suicide attempt. Momentary suicidal urges were reported via EMA by a similar percentage of handgun owners and non-owners (29.1% vs. 36.5%; χ2(1)=0.8, p=.373). Handgun owners and non-owners also did not differ with respect to mean affect states and momentary suicidal urge scores from EMA ratings: pleasant activation (F(134)=2.0, p=.162), activated pleasure (F(134)=0.1, p=.713), unpleasant activation (F(1,134)=0.3, p=.596), activated displeasure (F(1,134)=2.2, p=.137), and suicidal urge (F(1,134)=2.3, p=.133).

Variability in affect scores during suicidal urges

Mean affect state scores during suicidal urges significantly differed from mean affect state scores when suicidal urges were denied (Table 3). Specifically, scores for pleasant activation (Suicidal: M=2.14, 95% CI=1.91–2.37, vs. Nonsuicidal: M=2.58, 95% CI=2.45–2.72; ΔM=−0.44, SE=0.14, p=.001) and activated pleasure (Suicidal: M=1.96, 95% CI=1.68–2.25, vs. Nonsuicidal: M=2.53, 95% CI=2.38–2.67; ΔM=−0.56, SE=0.16, p<.001) were greater in nonsuicidal versus suicidal urges. In contrast, scores for unpleasant activation (Suicidal: M=1.65, 95% CI=1.44–1.86, vs. Nonsuicidal: M=1.28, 95% CI=1.21–1.34; ΔM=0.37, SE=0.11, p<.001) and activated displeasure (Suicidal: M=1.82, 95% CI=1.59–2.05, vs. Nonsuicidal: M=1.23, 95% CI=1.18–1.28; ΔM=0.59, SE=0.12, p<.001) were reduced when denying suicidal urges.

Table 3.

Fixed and random effect estimates for affect states, with estimated 95% score ranges, when denying and experiencing suicidal urges

Nonsuicidal Suicidal
Affective State Fixed Random 95% Score Range Fixed Random 95% Score Range
Pleasant Activation 2.584 0.631 1.027–4.141 2.142 0.491 0.769–3.514
Activated Pleasure 2.527 0.733 0.848–4.205 1.965 0.749 0.268–3.661
Unpleasant Activation 1.279 0.144 0.535–2.023 1.650 0.419 0.381–2.919
Activated Displeasure 1.227 0.087 0.650–1.804 1.822 0.583 0.325–3.318
Open in a new tab

The test of homogeneity was not statistically significant for pleasant activation (χ2(1)=0.7, p=.412) or activated pleasure χ2(1)=0.0, p=.941) but was statistically significant for unpleasant activation (χ2(1)=15.4, p<.001) and activated displeasure (χ2(1)=62.5, p<.001). Variability in unpleasant activation and activated displeasure was greater when denying suicidal urges versus during suicidal urges (unpleasant activation: Suicidal: γ0=0.14, SE=0.02, vs. Nonsuicidal: γ0=0.42, SE=0.11; activated displeasure: Suicidal: γ0=0.09, SE=0.01, vs. Nonsuicidal: γ0=0.58, SE=0.13).

Variability in associations among affect and suicidal ideation

In the full sample, all four fixed effects were significantly associated with severity of suicidal urge score (Table 4): pleasant activation (F(1,136)=10.8, p=.001), activated pleasure (F(1,136)=14.9, p<.001), unpleasant activation (F(1,128)=8.4, p=.005), and activated displeasure (F(1,134)=8.7, p=.004). All four random effects were also statistically significant: pleasant activation (γi=0.010, SE=0.002, p=.004), activated pleasure (γi=0.011, SE=0.002, p=.003), unpleasant activation (γi=0.007, SE=0.001, p=.005), and activated displeasure (γi=0.286, SE=0.037, p<.001). On average, pleasant activation and activated pleasure were negatively correlated with suicidal urge scores whereas unpleasant activation and activated displeasure were positively correlated with suicidal urge scores. Activated displeasure showed the relative strongest average association with suicidal urge scores and varied the most, ranging from a moderately negative association (i.e., more intense activated displeasure correlated with more severe suicidal urges) to a moderately positive association with suicidal urge scores (i.e., more intense activated displeasure correlated with less severe suicidal urges).

Table 4.

Results of mixed effects modeling predicting severity of suicidal urge, with estimated 95% slope ranges in the full sample and by handgun ownership group

Fixed Effect Random Effect 95% Slope Range
Affect Score βi (SE) p γi (SE) p Lower Upper
Full Sample
 Pleasant Activation −0.033 (0.010) .001 0.010 (0.002) .004 −0.053 −0.013
 Activated Pleasure −0.038 (0.010) <.001 0.011 (0.002) .003 −0.059 −0.018
 Unpleasant Activation 0.031 (0.011) .005 0.007 (0.001) .005 0.017 0.045
 Activated Displeasure 0.144 (0.049) .004 0.286 (0.010) <.001 −0.418 0.705
Handgun Owners
 Pleasant Activation −0.014 (0.005) .005 0a n/a −0.014 −0.014
 Activated Pleasure −0.016 (0.004) .001 0a n/a −0.016 −0.016
 Unpleasant Activation 0.024 (0.010) .013 0.002 (0.0003) <.001 −0.064 0.112
 Activated Displeasure 0.068 (0.017) .006 0.010 (0.003) .089 −0.517 0.653
Non-Owners
 Pleasant Activation −0.063 (0.027) .020 0.036 (0.008) .022 −0.435 0.309
 Activated Pleasure −0.078 (0.028) .006 0.040 (0.009) .024 −0.470 0.314
 Unpleasant Activation 0.035 (0.019) .071 0.011 (0.003) .061 −0.171 0.241
 Activated Displeasure 0.262 (0.025) .025 0.688 (0.135) .009 −1.364 1.888
Open in a new tab
a

Zero values can result when there is insufficient variance in the parameter being estimated. βi=fixed effect coefficient, γi=random effect coefficient, SE=standard error.

Differences between handgun owners and non-owners

Results of mixed effects modeling for handgun owners and non-owners are summarized in Table 4. Overall, the fixed effect coefficients were smaller among handgun owners than non-owners but only the activated pleasure slopes significantly differed between groups: pleasant activation (F(1,135)=3.2, p=.074), activated pleasure (F(1,135)=4.7, p=.031), unpleasant activation (F(1,135)=0.3, p=.607), and activated displeasure (F(1,135)=2.7, p=.100). Among handgun owners, only the random effect of unpleasant activation was statistically significant (γi=0.002, SE=0.0003, p<.001) but among non-owners, the random effects of pleasant activation (γi=0.036, SE=0.008, p=.022), activated pleasure (γi=0.040, SE=0.009, p=.024), and activated displeasure (γi=0.688, SE=0.135, p=.009) were statistically significant.

The test of homogeneity was statistically significant for pleasant activation (χ2(2)=280.0, p<.001), activated pleasure (χ2(2)=270.7, p<.001), unpleasant activation (χ2(2)=73.1, p<.001), and activated displeasure (χ2(2)=237.1, p<.001). The strength of association between all four affect states and suicidal state score was less variable among handgun owners than non-owners. Greater pleasant activation and activated pleasure were associated with less severe suicidal urge score on average among both handgun owners and non-owners but the strength of these associations varied significantly for non-owners, ranging from a moderate negative to a moderate positive correlation. Greater unpleasant activation was associated with more severe suicidal urge score on average among handgun owners but not non-owners, with the strength of association among handgun owners varying significantly and ranging from a small negative to a small positive correlation. Finally, greater activated displeasure was associated with more severe suicidal urge score on average among both handgun owners and non-owners, but the strength of this association varied significantly for non-owners, ranging from a large negative to a large positive correlation.

Discussion

Suicidal thoughts, urges, and behaviors have historically been viewed as an outcome of extreme negative cognitive-affective states (Klonsky & May, 2015; O’Connor & Kirtley, 2018; Van Orden et al., 2010) but newer conceptual models based on complexity theory implicate heterogeneity of suicidal experiences within and between people and groups in suicide risk (Bryan, Butner, et al., 2020; Franklin, 2019). Consistent with this perspective, participants in the current study reported considerable heterogeneity in momentary affective experiences, assessed using a circumplex model of affect, when they were experiencing suicidal urges. Like previous studies, positive affect was lower and negative affect was higher on average when participants in this study experienced suicidal urges relative to when they were not (Armey et al., 2020; Humber et al., 2013; Husky et al., 2017; Kleiman et al., 2018; Victor et al., 2021; Victor et al., 2019). Moreover, these effects were relatively similar in size within each type of affective valence (negative or positive), regardless of the level of affective arousal (moderate or high).

Interestingly, there was greater variability in negative affect regardless of arousal level when participants experienced suicidal urges, but there were similar degrees of variability in positive affect when experiencing and not experiencing suicidal urges. These results suggest that the experience of suicide urges was not limited to heightened negative affect and reduced positive affect; in this sample, the intensity of negative affect varied considerably and to a greater degree when people were experiencing suicidal urges versus when they were not. As compared to when people were experiencing suicidal urges, moments without suicidal urges were characterized by more intense positive affect combined with less intense and less variable negative affect. These patterns emerged even though the intensity of the affective states examined (regardless of valence or arousal) ranged from low (i.e., very slightly or not at all) to moderate. It is also notable that participants endorsed a wider range of positive affect than negative affect and rarely endorsed the most severe levels of negative affect during suicidal urges, two patterns hypothesized by Franklin (2019).

Also like previous studies, suicidal urge severity was inversely correlated with positive affect but positively correlated with negative affect on average. These effects were similar in size for most of the affect scores examined, with one exception: we found that activated displeasure showed the relative strongest average association with suicidal urge severity and varied more than the affective states. This pattern was much more pronounced among non-owners than handgun owners. That these effects were so much larger in this moderate arousal state (versus the high arousal state) may indicate that suicidal urges may be differentially related to certain levels of negatively valenced arousal. Additional studies that more comprehensively assess suicidal urges and other manifestations of suicidal states (e.g., suicidal ideation, suicidal planning, suicidal behaviors) across a wider range of arousal (i.e., including low arousal affect) are needed to further elucidate the nature of this relationship. Nevertheless, we did find a wide range in variability in associations between activated displeasure and suicide urge severity (ranging from a moderately negative association to a moderately positive association), suggesting a high degree of heterogeneity in associations for this particular affective state. Collectively, these findings suggest the importance of considering multiple dimensions of affect, including arousal, in characterizing suicide risk states.

We also found that associations between affective states and suicidal urges varied considerably across participant subgroups. For example, unique associations of affective states were observed during suicidal urges for handgun owners (e.g., weak activated pleasure, highly variable unpleasant activation) and non-owners (e.g., highly variable pleasant activation and activated pleasure). Suicidal urges were therefore sometimes more severe when negative affect was elevated but other times were more severe when negative affect was low. Our results implicate a similar conclusion for positive affect. Taken together, these findings support complexity-based models that allow for heterogeneity in suicidal states and the possible existence of multiple trajectories to suicide (Bryan, Butner, et al., 2020; Franklin, 2019; Huang et al., 2020; Rudd et al., 2006).

Our findings may provide some context for historic difficulties in understanding and predicting suicidal behaviors. Traditional approaches to suicide risk prediction and screening are typically based on averaged group effects that are unable to fully capture the heterogeneity of suicidal thoughts and urges. Because these averaged effects tend to be small in magnitude, they provide negligible advantage beyond random guessing for prediction purposes (Franklin et al., 2017). Our results suggest these small effects may be explained by heterogeneity of suicidal experiences. Because the association of affective states with severity of suicidal urges ranged from negative to positive, larger individual effects in opposing directions are “cancelled out,” leaving only modest averaged associations. These individual differences implicate the potential value of complexity-based approaches within clinical and non-clinical settings. For example, clinical suicide risk detection and screening efforts could be improved by applying novel computational methods based on complex dynamical systems theory, which conceptualizes suicide risk as multiple discrete stable states, with the emergence of suicidal behavior reflecting a transition from one stable state (i.e., not attempting suicide) to another (i.e., attempting suicide). Because of heterogeneity, there is no limit to the number of pathways that someone can theoretically follow when transitioning from one state to another, but some pathways will be more probable than others and some pathways will be so improbable that they can be practically considered impossible. Dynamical systems modeling of heterogeneity could reveal and characterize these various pathways, thereby leading to improved risk prediction and monitoring methods.

Our results also highlight the potential value of a precision medicine framework for suicide prevention that better accounts for individual variability and heterogeneity of personal background, life circumstances, and internal experiences; what works to prevent or reduce suicidal urges for one person or group may not work for another person or group. In this study, for example, the suicidal urges of handgun owners were much less strongly associated with activated displeasure, mirroring previous research finding that firearm owners with recent suicidal behavior were less likely than non-owners to report suicidal ideation (Bryan et al., 2022). Taken together, these patterns suggest at least two possibilities. First, suicidal urges among firearm owners are less likely to be characterized by aversive emotions and thoughts about death or suicide. Second, something(s) other than aversive emotions and thoughts about death or suicide are more central to the concept of “feeling suicidal” for firearm owners relative to non-owners. Previous research has shown that firearm owners view the world as more dangerous and threatening than non-owners (Anestis & Bryan, 2021; Bryan, Bryan, & Anestis, 2020b; Buttrick, 2020). Preliminary findings suggest heightened threat perceptions are, in turn, associated with disrupted cognitive-affective processes among firearm owners (Bryan, Bryan, & Anestis, 2020a; Bryan, Daruwala, et al., 2023; Manzler et al., 2024). Of note, firearm owners display heightened reactivity when in the presence of firearms (Bryan, Daruwala, et al., 2023), increasing their vulnerability to experiencing rapid, sudden shifts in cognitive-affective states. Heighted reactivity is, in turn, associated with increased suicidal ideation (Lieberman et al., 2020). Although research is needed to further elucidate and test these possibilities, the present findings suggest that suicidal urges among firearm owners may differ from the suicidal urges of non-owners, implicating the need to identify screening, assessment, and prevention strategies that are more applicable to this group.

Conclusions based on these findings should be made with consideration for several limitations. First, the participants in this study were recruited from only a single state in the U.S. Because beliefs about firearms and firearm ownership patterns can vary across states, regions, and nations, the present results may not extend to firearm owners in other parts of the U.S. or other nations (additional information about firearm laws in Ohio and other U.S. states can be found at www. https://everytownresearch.org). Second, our handgun group had a larger percentage of male participants than the non-owner group, mirroring national trends in firearm ownership (Parker et al., 2020). To assess the impact of this group difference, we repeated our models with sex entered as a covariate and found no change in the pattern of results. Although this reduces concern about the influence of sex on our findings, we cannot completely rule out the possibility of confounding effects. Third, we assessed suicidal urges using only a single item. Although this approach has been shown to be reliable and valid for study designs like EMA with repeated measurements (Song et al., 2022), newer research suggests suicidal experiences are multidimensional, with different dimensions of suicidal experiences following different time courses (Coppersmith et al., 2022). The present findings may therefore be specific to the urge to kill oneself versus other facets of suicidal states (e.g., desire for death, thinking about suicide attempt methods or locations, suicidal behaviors). Fourth, due to limitations of the existing data, we were only able to examine a limited range of arousal conditions (i.e., moderate and high). Additional research that measures low-arousal affect is needed to further elucidate the patterns observed in this study. Finally, as would be expected in a nonclinical community sample, suicidal urges were infrequently endorsed infrequently and was low in intensity. Conclusions based on the present study therefore may not generalize to higher risk clinical samples (e.g., psychiatric inpatients) with more persistent and higher severity suicidal experiences.

Despite these limitations, the present study provides novel information about variability in suicidal urges across individuals and subgroups. Our results suggest that suicidal urges are marked by considerable affective heterogeneity, can occur in the absence of extreme negative affect and/or the presence of positive affect, and differ between handgun owners and non-owners. Future research should aim to extend these findings to clinical settings and identify other similarities and differences in the suicidal urges of various population subgroups.

Highlights.

  • Little is known about the range of affect experienced during suicidal states

  • Affect and suicidal urges were assessed repeatedly using ecological momentary assessment

  • Affect during suicidal urges ranged considerably across individuals

  • The relationship of affect and suicidal urges also ranged considerably

  • The relationship of affect and suicidal urges was weaker among handgun owners

Funding/Support

Research reported in this publication was supported in part by the National Institute of Mental Health of the National Institutes of Health under Award Number R61MH125759 (PI: Bryan). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The study sponsor had no role in the design or conduct of the study; the collection, management, analysis, and interpretation of the data; the preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Declaration of Interest

Dr. C. J. Bryan reported grants from the Department of Defense, the National Institute of Mental Health, the New Jersey Gun Violence Research Center, the Bob Woodruff Foundation, the USAA Foundation, the Boeing Company, and the American Foundation for Suicide Prevention during the conduct of the study; personal fees from Oui Therapeutics outside the submitted work; and ownership of Anduril, LLC, outside the submitted work. Dr. M. L. Bozzay reported grants from the National Institute of Mental Health and the Department of Defense. Dr. Tabares reported grants from the New Jersey Gun Violence Research Center and the National Institute of Mental Health during the conduct of the study. Dr. S. E. Daruwala reported grants from the New Jersey Gun Violence Research Center. Dr. J. E. Butner reported grant funding from the National Institute of Mental Health during the conduct of the study. Dr. S. M. Gorka reported grants from the National Institute of Mental Health and the National Institute on Alcohol Abuse and Alcoholism during the conduct of this study.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. American Psychological Association. (2024). APA Dictionary of Psychology. American Psychological Association. Retrieved June 21, 2024 from https://dictionary.apa.org/ [Google Scholar]
  2. Anestis MD, & Bryan CJ (2021). Threat perceptions and the intention to acquire firearms. Journal of Psychiatric Research, 133, 113–118. [DOI] [PubMed] [Google Scholar]
  3. Apter A, Bleich A, King RA, Kron S, Fluch A, Kotler M, & Cohen DJ (1993). Death without warning?: a clinical postmortem study of suicide in 43 Israeli adolescent males. Archives of general psychiatry, 50(2), 138–142. [DOI] [PubMed] [Google Scholar]
  4. Armey MF, Brick L, Schatten HT, Nugent NR, & Miller IW (2020). Ecologically assessed affect and suicidal ideation following psychiatric inpatient hospitalization. Gen Hosp Psychiatry, 63, 89–96. 10.1016/j.genhosppsych.2018.09.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barber C, Frank E, & Demicco R (2017). Reducing Suicides Through Partnerships Between Health Professionals and Gun Owner Groups—Beyond Docs vs Glocks. JAMA Internal Medicine, 177(1), 5–6. 10.1001/jamainternmed.2016.6712 [DOI] [PubMed] [Google Scholar]
  6. Bryan CJ, Allen MH, Thomsen CJ, May AM, Baker JC, Bryan AO, Harris JA, Cunningham CA, Taylor KB, & Wine MD (2021). Improving suicide risk screening to identify the highest risk patients: results from the PRImary care Screening Methods (PRISM) study. The Annals of Family Medicine, 19(6), 492–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bryan CJ, Allen MH, Wastler HM, Bryan AO, Baker JC, May AM, & Thomsen CJ (2023). Rapid intensification of suicide risk preceding suicidal behavior among primary care patients. Suicide and Life-Threatening Behavior. [DOI] [PubMed] [Google Scholar]
  8. Bryan CJ, Bryan AO, & Anestis MA (2020a). Positive and negative affective processes associated with firearm acquisition and ownership. Journal of Social and Clinical Psychology, 39(10), 861–875. [Google Scholar]
  9. Bryan CJ, Bryan AO, & Anestis MD (2020b). Associations among exaggerated threat perceptions, suicidal thoughts, and suicidal behaviors in US firearm owners. Journal of Psychiatric Research, 131, 94–101. [DOI] [PubMed] [Google Scholar]
  10. Bryan CJ, Bryan AO, Khazem LR, Aase DM, Moreno JL, Ammendola E, Bauder CR, Hiser J, Daruwala SE, & Baker JC (2024). Crisis response planning rapidly reduces suicidal ideation among US military veterans receiving massed cognitive processing therapy for PTSD. Journal of anxiety disorders, 102, 102824. [DOI] [PubMed] [Google Scholar]
  11. Bryan CJ, Bryan AO, Wastler HM, Khazem LR, Ammendola E, Baker JC, Szeto E, Tabares J, & Bauder CR (2022). Assessment of Latent Subgroups With Suicidal Ideation and Suicidal Behavior Among Gun Owners and Non–Gun Owners in the US. JAMA network open, 5(5), e2211510–e2211510. 10.1001/jamanetworkopen.2022.11510 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bryan CJ, Butner JE, May AM, Rugo KF, Harris JA, Oakey DN, Rozek DC, & Bryan AO (2020). Nonlinear change processes and the emergence of suicidal behavior: A conceptual model based on the fluid vulnerability theory of suicide. New ideas in psychology, 57, 100758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Bryan CJ, Clemans TA, Leeson B, & Rudd MD (2015). Acute vs. chronic stressors, multiple suicide attempts, and persistent suicide ideation in US soldiers. The Journal of nervous and mental disease, 203(1), 48–53. [DOI] [PubMed] [Google Scholar]
  14. Bryan CJ, Daruwala SE, Tabares JV, Butner JE, Coccaro EF, & Gorka SM (2023). Heightened threat perceptions and reduced stability in anxiety and fear among US adults who carry handguns. Journal of anxiety disorders, 99, 102764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Bryan CJ, Mintz J, Clemans TA, Leeson B, Burch TS, Williams SR, Maney E, & Rudd MD (2017). Effect of crisis response planning vs. contracts for safety on suicide risk in US Army soldiers: a randomized clinical trial. Journal of affective disorders, 212, 64–72. [DOI] [PubMed] [Google Scholar]
  16. Bryan CJ, Rudd MD, Peterson AL, Young-McCaughan S, & Wertenberger EG (2016). The ebb and flow of the wish to live and the wish to die among suicidal military personnel. Journal of affective disorders, 202, 58–66. [DOI] [PubMed] [Google Scholar]
  17. Butner JE, Bryan CJ, Tabares JV, Brown LA, Young-McCaughan S, Hale WJ, Mintz J, Litz BT, Yarvis JS, & Fina BA (2021). Temporal-dimensional examination of the Scale for Suicidal Ideation in a cohort of service members in treatment for PTSD. Psychological trauma: theory, research, practice, and policy, 13(7), 793. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Buttrick N (2020). Protective gun ownership as a coping mechanism. Perspectives on psychological science, 15(4), 835–855. [DOI] [PubMed] [Google Scholar]
  19. Caves Sivaraman JJ, Ranapurwala SI, Proescholdbell S, Naumann RB, Greene SB, & Marshall SW (2022). Suicide typologies among Medicaid beneficiaries, North Carolina 2014–2017. BMC psychiatry, 22(1), 104. 10.1186/s12888-022-03741-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Centers for Disease Control and Prevention. (2005). Web-based Injury Statistics Query and Reporting System (WISQARS) www.cdc.gov/injury/wisqars
  21. Chaudhury SR, Singh T, Burke A, Stanley B, Mann JJ, Grunebaum M, Sublette ME, & Oquendo MA (2016). Clinical correlates of planned and unplanned suicide attempts. The Journal of nervous and mental disease, 204(11), 806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Conner KR (2004). A call for research on planned vs. unplanned suicidal behavior. Suicide and Life-Threatening Behavior, 34(2), 89–98. [DOI] [PubMed] [Google Scholar]
  23. Conner KR, Hesselbrock VM, Meldrum SC, Schuckit MA, Bucholz KK, Gamble SA, Wines JD, & Kramer J (2007). Transitions to, and correlates of, suicidal ideation, plans, and unplanned and planned suicide attempts among 3,729 men and women with alcohol dependence. Journal of studies on alcohol and drugs, 68(5), 654–662. [DOI] [PubMed] [Google Scholar]
  24. Coppersmith DD, Ryan O, Fortgang R, Millner A, Kleiman E, & Nock M (2022). Mapping the Timescale of Suicidal Thinking. Proceedings of the National Academy of Sciences, 120, e2215434120. 10.1073/pnas.2215434120 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Crawford JR, & Henry JD (2004). The positive and negative affect schedule (PANAS): construct validity, measurement properties and normative data in a large non-clinical sample. Br J Clin Psychol, 43(Pt 3), 245–265. 10.1348/0144665031752934 [DOI] [PubMed] [Google Scholar]
  26. Curtin SC, Garnett MF, & Ahmad FB (2023). Provisional Estimates of Suicide by Demographic Characteristics: United States, 2022. [Google Scholar]
  27. Enders CK (2001). A primer on maximum likelihood algorithms available for use with missing data. Structural Equation Modeling, 8(1), 128–141. [Google Scholar]
  28. Franklin JC (2019). Psychological primitives can make sense of biopsychosocial factor complexity in psychopathology. BMC medicine, 17(1), 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Franklin JC, Ribeiro JD, Fox KR, Bentley KH, Kleiman EM, Huang X, Musacchio KM, Jaroszewski AC, Chang BP, & Nock MK (2017). Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research. Psychological bulletin, 143(2), 187–232. [DOI] [PubMed] [Google Scholar]
  30. Glenn CR, Kleiman EM, Kearns JC, Boatman AE, Conwell Y, Alpert-Gillis LJ, & Pigeon W (2021). Sleep problems predict next-day suicidal thinking among adolescents: A multimodal real-time monitoring study following discharge from acute psychiatric care. Development and Psychopathology, 33(5), 1701–1721. [Google Scholar]
  31. Glenn CR, Kleiman EM, Kearns JC, Santee AC, Esposito EC, Conwell Y, & Alpert-Gillis LJ (2020). Feasibility and acceptability of ecological momentary assessment with high-risk suicidal adolescents following acute psychiatric care. Journal of Clinical Child & Adolescent Psychology, 1–17. [DOI] [PubMed] [Google Scholar]
  32. Hanlon TJ, Barber C, Azrael D, & Miller M (2019). Type of Firearm Used in Suicides: Findings From 13 States in the National Violent Death Reporting System, 2005–2015. Journal of Adolescent Health, 65(3), 366–370. 10.1016/j.jadohealth.2019.03.015 [DOI] [PubMed] [Google Scholar]
  33. Huang X, Ribeiro JD, & Franklin JC (2020). The differences between suicide ideators and suicide attempters: Simple, complicated, or complex? Journal of consulting and clinical psychology, 88(6), 554. [DOI] [PubMed] [Google Scholar]
  34. Humber N, Emsley R, Pratt D, & Tarrier N (2013). Anger as a predictor of psychological distress and self-harm ideation in inmates: A structured self-assessment diary study. Psychiatry Research, 210(1), 166–173. [DOI] [PubMed] [Google Scholar]
  35. Husky M, Swendsen J, Ionita A, Jaussent I, Genty C, & Courtet P (2017). Predictors of daily life suicidal ideation in adults recently discharged after a serious suicide attempt: A pilot study. Psychiatry Research, 256, 79–84. [DOI] [PubMed] [Google Scholar]
  36. Jeon HJ, Lee J-Y, Lee YM, Hong JP, Won S-H, Cho S-J, Kim J-Y, Chang SM, Lee HW, & Cho MJ (2010). Unplanned versus planned suicide attempters, precipitants, methods, and an association with mental disorders in a Korea-based community sample. Journal of affective disorders, 127(1–3), 274–280. [DOI] [PubMed] [Google Scholar]
  37. Kleiman EM, Coppersmith DDL, Millner AJ, Franz PJ, Fox KR, & Nock MK (2018). Are suicidal thoughts reinforcing? A preliminary real-time monitoring study on the potential affect regulation function of suicidal thinking. J Affect Disord, 232, 122–126. 10.1016/j.jad.2018.02.033 [DOI] [PubMed] [Google Scholar]
  38. Kleiman EM, Turner BJ, Fedor S, Beale EE, Huffman JC, & Nock MK (2017). Examination of real-time fluctuations in suicidal ideation and its risk factors: Results from two ecological momentary assessment studies. Journal of abnormal psychology, 126(6), 726. [DOI] [PubMed] [Google Scholar]
  39. Klonsky ED, & May AM (2015). The three-step theory (3ST): A new theory of suicide rooted in the “ideation-to-action” framework. International Journal of Cognitive Therapy, 8(2), 114–129. [Google Scholar]
  40. Larsen RJ, & Diener E (1992). Promises and problems with the circumplex model of emotion. In Clark MS (Ed.), Emotion (pp. 25–29). Sage Publications. [Google Scholar]
  41. Lieberman L, Petrey K, Shankman SA, Phan KL, & Gorka SM (2020). Heightened reactivity to uncertain threat as a neurobehavioral marker of suicidal ideation in individuals with depression and anxiety. International Journal of Psychophysiology, 155, 99–104. 10.1016/j.ijpsycho.2020.06.003 [DOI] [PubMed] [Google Scholar]
  42. Manzler CA, Gorka SM, Tabares JV, & Bryan CJ (2024). Impact of handgun ownership and biological sex on startle reactivity to predictable and unpredictable threats. International Journal of Psychophysiology, 197, 112297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Martin J, LaCroix JM, Novak LA, & Ghahramanlou-Holloway M (2020). Typologies of Suicide: A Critical Literature Review. Arch Suicide Res, 24(sup1), 25–40. 10.1080/13811118.2018.1564100 [DOI] [PubMed] [Google Scholar]
  44. Millner AJ, Lee MD, & Nock MK (2015). Single-item measurement of suicidal behaviors: Validity and consequences of misclassification. PloS one, 10(10), e0141606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. O’Connor RC, & Kirtley OJ (2018). The integrated motivational–volitional model of suicidal behaviour. Philosophical Transactions of the Royal Society B: Biological Sciences, 373(1754), 20170268. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Parker K, Horowitz J, Igielnik R, Oliphant J, & Brown A (2020). America’s complex relationship with guns: an in-depth look at attitudes and experiences of US adults. Pew Research Center. 2017. In. [Google Scholar]
  47. Planty M, & Truman JL (2013). Firearm Violence, 1993–2011. US Department of Justice, Office of Justice Programs, Bureau of Justice …. [Google Scholar]
  48. Posner J, Russell JA, & Peterson BS (2005). The circumplex model of affect: An integrative approach to affective neuroscience, cognitive development, and psychopathology. Development and Psychopathology, 17(3), 715–734. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Richards JE, Whiteside U, Ludman EJ, Pabiniak C, Kirlin B, Hidalgo R, & Simon G (2019). Understanding why patients may not report suicidal ideation at a health care visit prior to a suicide attempt: a qualitative study. Psychiatric Services, 70(1), 40–45. [DOI] [PubMed] [Google Scholar]
  50. Rogers ML, Jeon ME, Zheng S, Richards JA, Joiner TE, & Galynker I (2023). Two sides of the same coin? Empirical examination of two proposed characterizations of acute suicidal crises: Suicide crisis syndrome and acute suicidal affective disturbance. Journal of Psychiatric Research, 162, 123–131. [DOI] [PubMed] [Google Scholar]
  51. Rudd MD (2006). Fluid vulnerability theory: A cognitive approach to understanding the process of acute and chronic suicide risk. [Google Scholar]
  52. Rudd MD, Mandrusiak M, & Joiner TE (2006). The case against no-suicide contracts: the commitment to treatment statement as a practice alternative. J Clin Psychol, 62(2), 243–251. 10.1002/jclp.20227 [DOI] [PubMed] [Google Scholar]
  53. Russell JA (1980). A circumplex model of affect. Journal of personality and social psychology, 39(6), 1161. [DOI] [PubMed] [Google Scholar]
  54. Schatten HT, Brick LA, Holman CS, & Czyz E (2021). Differential time varying associations among affective states and suicidal ideation among adolescents following hospital discharge. Psychiatry Research, 305, 114174. [DOI] [PubMed] [Google Scholar]
  55. Simon GE, Coleman KJ, Rossom RC, Beck A, Oliver M, Johnson E, Whiteside U, Operskalski B, Penfold RB, & Shortreed SM (2016). Risk of suicide attempt and suicide death following completion of the Patient Health Questionnaire depression module in community practice. The Journal of clinical psychiatry, 77(2), 20461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Simon GE, Rutter CM, Peterson D, Oliver M, Whiteside U, Operskalski B, & Ludman EJ (2013). Does response on the PHQ-9 Depression Questionnaire predict subsequent suicide attempt or suicide death? Psychiatric Services, 64(12), 1195–1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Simon TR, Swann AC, Powell KE, Potter LB, Kresnow M. j., & O’Carroll PW (2001). Characteristics of impulsive suicide attempts and attempters. Suicide and Life-Threatening Behavior, 32(Supplement to Issue 1), 49–59. [DOI] [PubMed] [Google Scholar]
  58. Song J, Howe E, Oltmanns JR, & Fisher AJ (2022). Examining the concurrent and predictive validity of single items in ecological momentary assessments. Assessment, 10731911221113563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Stanley B, Itzhaky L, & Oquendo MA (2021). Identifying Neurobiological Underpinnings of Two Suicidal Subtypes. Journal of Psychiatry and Brain Science, 6(4), e210016, Article e210016. 10.20900/jpbs.20210016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Substance Abuse and Mental Health Services Administration. (2023). Key substance use and mental health indicators in the United States: results from the 2022 National Survey on Drug Use and Health (HHS Publication No. PEP23-07-01-006, NSDUH Series H-58). https://www.samhsa.gov/data/report/2022-nsduh-annual-national-report
  61. Tian L, Yang Y, Yang H, & Huebner ES (2017). Prevalence of suicidal ideation and its association with positive affect in working women: A day reconstruction study. Frontiers in psychology, 8, 285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby EA, & Joiner TE Jr (2010). The interpersonal theory of suicide. Psychological review, 117(2), 575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Victor SE, Brown SL, & Scott LN (2021). Prospective and concurrent affective dynamics in self-injurious thoughts and behaviors: An examination in young adult women. Behavior therapy, 52(5), 1158–1170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Victor SE, Scott LN, Stepp SD, & Goldstein TR (2019). I want you to want me: Interpersonal stress and affective experiences as within-person predictors of nonsuicidal self-injury and suicide urges in daily life. Suicide and Life-Threatening Behavior, 49(4), 1157–1177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Wastler HM, Bryan AO, & Bryan CJ (2022). Suicide attempts among adults denying active suicidal ideation: An examination of the relationship between suicidal thought content and suicidal behavior. J Clin Psychol, 78(6), 1103–1117. 10.1002/jclp.23268 [DOI] [PubMed] [Google Scholar]
  66. Watson D, Clark LA, & Tellegen A (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol, 54(6), 1063–1070. 10.1037//0022-3514.54.6.1063 [DOI] [PubMed] [Google Scholar]
  67. Watson D, & Tellegen A (1985). Toward a consensual structure of mood. Psychological bulletin, 98(2), 219–235. [DOI] [PubMed] [Google Scholar]
  68. Wyder M, & De Leo D (2007). Behind impulsive suicide attempts: Indications from a community study. Journal of affective disorders, 104(1–3), 167–173. [DOI] [PubMed] [Google Scholar]
  69. Yik M, Russell JA, & Steiger JH (2011). A 12-point circumplex structure of core affect. Emotion, 11(4), 705. [DOI] [PubMed] [Google Scholar]

RESOURCES