Gun and Other Weapon–Involved Intimate Partner Violence and Psychiatric Symptoms Authors: Tyson Whitten, Kimberlie Dean, Alvin Kuowei Tay, Batool Moussa, Fatima Hassoun, Jane Fisher, Melissa J. Green, Susan J. Rees Published: 2026 ABSTRACT Key PointsQuestionIs intimate personal violence (IPV) involving a gun or other weapon associated with worse mental health symptoms, independent of other abuse and prior mental disorder symptoms?FindingsIn this cohort study of 1114 women, gun and other weapon–involved IPV was associated with higher panic disorder and posttraumatic stress disorder symptoms but not depression or separation anxiety symptoms after adjustment. Increases were most often modest but were larger for women with recent experience of IPV involving weapons.MeaningThese results suggest that gun and other weapon–involved IPV signals heightened fear-related harm; its disclosure should prompt safety-focused responses alongside mental health care. Introduction Australia’s gun policy debate has recently intensified after the mass shooting at Bondi Beach in December 2025, which killed 15 people and injured many others. This was followed by federal gun reforms, including a national buyback and tighter licensing and background checks.1 In the US, the frequency of mass shootings and widespread access to firearms, including assault weapons, underscore the pervasive risk faced by its population. In 2020, more than 45 000 firearm-related deaths occurred in the US, more than half from suicide and more than 40% from homicide, with fatalities increasing further in 2021.2,3 Although firearm-related deaths and injuries are often framed around mass casualty events, firearm and other weapon access also has implications for the more prevalent problem of intimate partner violence (IPV), in which guns, knives, and other weapons can be used to inflict serious injury or homicide, facilitate coercive control, and contribute to profound psychological harm.3,4,5,6 Despite extensive evidence linking IPV broadly to adverse mental health outcomes in survivors,7,8,9 comparatively little is known about the psychological consequences of gun and other weapon–involved IPV that does not result in fatal injury.5 Firearms are frequently implicated in the intimate partner homicides of women, yet far less is understood about their role in nonfatal abuse,3,5 despite nonfatal firearm physical injuries being 2 to 3 times as frequent as firearm deaths.10 From a psychological perspective, the inherent lethality of a gun, knife, or other deadly weapon, combined with the minimal effort required to inflict fatal harm, means that even its display or use as a threat can establish a powerful regimen of entrapment and coercive control.11,12 Such dynamics can entrench fear, restrict autonomy, and enable patterns of long-term and escalating abuse.5 In this context, the gun or other weapon can function as a persistent instrument of intimidation, even when physical injury does not occur, by constraining everyday autonomy and heightening anticipatory fear. These dynamics may amplify hypervigilance, physiologic arousal, and intrusive reexperiencing by persistently activating the body’s threat-detection systems. When an individual perceives a credible risk to life, such as the presence or threatened use of a gun or other harmful weapon, the brain’s fear circuitry rapidly prioritizes survival. The sympathetic nervous system mobilizes the classic fight-or-flight response, increasing heart rate, respiratory rate, muscle tension, and stress hormone release to prepare the body for imminent danger.13,14,15,16 Prolonged exposure to life-threatening situations in the context of IPV directly increases the risk of developing both panic disorder and posttraumatic stress disorder (PTSD) because repeated experiences of intense fear of death or harm heighten vigilance and self-protective responses, making individuals hyperaware and fearful of the actual threat and their own physical reactions, whereas the traumatic response becomes amplified with emotional and sensory intensity.15,16,17,18 Nonfatal gun, knife, or other weapon threats may also contribute to depressive symptoms through sustained stress, hopelessness, and diminished perceived control associated with the enormity of the threat posed by such harmful mechanisms of injury and death.19,20 It is also theoretically likely that nonfatal gun and other weapon use could relate to separation anxiety disorder in adulthood because exposure to extreme threat and related uncertainty, particularly when inflicted by a close or intimate partner, may heighten fears of abandonment and harm to or loss of key attachments.21,22,23 Gun and other weapon threats directed at children and other individuals close to the survivor are a common coercive tactic used to exert control, restrict autonomy, and instill profound fear, creating conditions that may contribute to the development of separation anxiety disorder particularly related to the child and the violated parent.21,22 Hence, gun and other weapon–involved IPV may be associated with PTSD, panic, anxiety, and depression symptoms beyond the symptom burden associated with co-occurring nonweapon physical IPV and psychological abuse. However, because gun and other weapon involvement commonly occurs with broader patterns of physical and psychological abuse, its specific association with different psychiatric symptom profiles remains unclear. Few studies4,5 have examined whether nonfatal gun and other weapon–involved IPV is associated with mental health symptoms after accounting for other IPV domains or whether these associations differ across trauma-, fear-, anxiety-, and depression-related outcomes. This distinction has important implications for identifying gun and other weapon–involved IPV, assessing risk, planning appropriate safety interventions, and preventing the serious trauma-related mental harms associated with nonfatal gun and other weapon–involved IPV. Using 6 waves of prospective, longitudinal data from the Women Aware with Their Children (WATCH) cohort, we examined the hypothesis that gun and other weapon–involved IPV, defined as threatened or enacted assault involving a gun, knife, or other weapon, was associated with higher psychiatric symptom counts for major depressive disorder (MDD), panic disorder (PD), PTSD, and adult separation anxiety disorder (ASAD), after accounting for other types of IPV and prior symptoms. Symptom counts were used as indicators of severity, allowing for detection of gradations in symptom burden that would be obscured by binary diagnosis classification. Our conceptual model treated gun and other weapon–involved IPV as a high-threat form of IPV that commonly co-occurs with, but is distinguishable from, nonweapon physical IPV and psychological abuse. As such, the primary analysis estimated the incremental association between gun and other weapon involvement and psychiatric symptoms, conditional on concurrent nonweapon physical IPV, psychological abuse, prior IPV exposure, prior psychiatric symptoms, year of birth, migrant status, and survey wave. This estimand differs from the total association between gun and other weapon–involved IPV and psychiatric symptoms, which may include pathways shared with other forms of IPV and broader abuse severity. To aid interpretation, we estimated both pooled marginal contrasts across observed person-waves and within-person transition contrasts comparing symptom change from nonexposure to exposure. Methods Data Data were drawn from the Women Aware with Their Children (WATCH) study, an ongoing longitudinal mental health study of 1335 women (685 consecutively enrolled from refugee backgrounds and 650 randomly selected Australian-born) recruited during pregnancy from 3 public antenatal clinics in Sydney and Melbourne, Australia. Participants were recruited from January 1, 2015, to March 31, 2016, at their first clinical appointment, which typically occurred between 12 and 20 weeks’ gestation. Eligible participants were pregnant women who were either Australian-born or from 1 of 3 conflict-affected humanitarian source populations representing Australia’s largest humanitarian intakes at the time: Tamil-speaking Sri Lanka, Sudan, or Arabic-speaking countries in the Middle East. A total of 1574 women were eligible for interview and 1135 were interviewed. Most declined due to lack of time. Women with overt psychosis, severe medical illness, and obvious intellectual impairment were excluded. Consecutive sampling was used to systematically recruit women from refugee backgrounds and conflict-affected countries to ensure an adequate sample size. A similar number of Australian-born women were recruited through a parallel sampling strategy that randomly identified a small subset of clinic attendees who were eligible for recruitment. To date, 6 interview waves have been undertaken at 12- to 18-month intervals: the latest wave was completed in July 2024 with 65.7% retention. The current study was limited to the 1114 women with nonmissing exposure and outcome data for at least 2 waves of data collection. A total of 742 women (66.6%) in the retained sample completed 5 or more survey waves. The WATCH study is approved by the Southwestern Sydney Local Health District Human Research Ethics Committee and Monash Health Ethics Committees. Participants gave informed consent to participate in the study before taking part and were remunerated A$50.00 (US$35.54) for their time. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies. Measures Interpersonal Violence We used the World Health Organization Violence Against Women questionnaire (version 10) to measure IPV.24,25 At each wave, participants reported whether they had experienced IPV from a current or former partner in the past 12 months. Items from the 15-item measure were grouped into 3 binary indicators: (1) gun and other weapon–involved IPV, defined as being threatened or attacked using a knife, gun, or other weapon; (2) nonweapon physical IPV, defined as threatened or enacted physical assault without a weapon (eg, pushing, punching, strangulation, and threats of harm); and (3) psychological abuse, defined as coercive and controlling behaviors (eg, accusations of infidelity, restricting contact with friends, financial control, or humiliation). Mental Disorder Symptoms The Mini International Neuropsychiatric Interview based on the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) was used at each survey wave to identify the number of symptoms relating to MDD (9 items with a score range of 0-9; diagnostic threshold score [number of endorsed symptoms necessary for a possible diagnosis] = 4; α = .72), PD (17 items with a score range of 0-17; diagnostic threshold score = 4; α = .84), PTSD (16 items with a score range of 0 to 16; diagnostic threshold score = 8; α = .91), and ASAD (14 items with a score range of 0-14; diagnostic threshold score = 3; α = .82). The measures assessed ASAD and PD symptoms at any time in life, PTSD symptoms in the last month, and MDD symptoms in the past 2 weeks. Scores for each measure were aggregated. Covariates Year of birth was centered at the median and migrant status indicated whether the participant was born in or outside Australia. Survey wave was included to adjust for time-related changes in symptom levels and exposure prevalence across follow-up. The lagged mental disorder symptoms outcome (count score from previous observed wave) was included as a covariate to control for prior and baseline risk and autocorrelation in repeated measures. Ever prior nonweapon IPV and psychological abuse were created, indicating whether a participant had reported IPV at any earlier wave; these measures were used to control for longer-term risk associated with prior violence beyond current-wave exposure. Statistical Analysis We used longitudinal mixed-effects models to analyze 4 bounded symptom count outcomes: MDD, panic disorder, PTSD, and ASAD. All models include a participant-specific random intercept to reduce within-person dependence across repeated observations. For each outcome, we fitted a β-binomial (β̂) mixed-effects model with a logit link, treating the symptom count as the number of successes of the maximum count. The β̂ models were used because the outcomes were bounded counts, representing the number of endorsed symptoms out of a fixed maximum for each disorder. Unlike standard count models, which assume an unbounded count (and can therefore model scores above the scale maximum), the β̂ model accommodates the restricted symptom range while allowing overdispersion.26,27 Fixed effects included wave-specific gun and other weapon–involved IPV exposure, nonweapon physical IPV exposure, psychological abuse, lagged mental disorder symptoms outcome, ever prior nonweapon physical IPV exposure, ever prior psychological abuse, refugee status, year of birth, and survey wave (wave 1 observations excluded due to adjustment for lagged outcome). All models were based on complete cases at the person-wave level. Adjusted marginal means on the response scale (expected symptom count) were estimated across waves and exposure states using marginal means, contrasts, and model-based estimation. We estimated 2 contrasts from the fitted β̂ models. First, a pooled marginal (independent) contrast compared expected symptom counts under counterfactual gun and other weapon–involved IPV exposure vs no exposure for all observed person-waves, holding all other covariates constant. Second, to focus interpretation on within-person exposure changes, we identified observed gun and other weapon–involved IPV transitions (eg, nonexposure at wave 3 to exposure at wave 4) and calculated the mean contrast using pretransition covariate states. Uncertainty was quantified using parametric simulation (50 000) of fixed effects: regression coefficients were drawn from a multivariate normal distribution defined by the estimated fixed effects and their variance-covariance matrix, converted to expected counts through the inverse-logit and outcome denominator, and summarized as pooled ridgeling density plots across all waves. A 2-sided P < .05 was considered statistically significant. All analyses were conducted in R, version 4.5.0 (R Foundation for Statistical Computing) using glmmTMB,28 emmeans,29 and MASS.30 Data analysis was conducted in February 2025. Results Among the 1114 women surveyed (mean [SD] age at wave 1, 29.48 [5.42] years), 585 (52.5%) were migrants, and the median year of birth was 1987 (range, 1969-1999). Descriptive statistics by study wave are presented in Table 1. Participant counts vary across waves due to item- and wave-level nonresponse. Gun and other weapon–involved IPV was uncommon (43 [3.9%] across any wave), although 13 exposed women (30.2%) reported gun and other weapon–involved IPV at multiple waves. Forty-one women (3.7%) transitioned from gun and other weapon–involved IPV nonexposure to exposure, and 23 (2.1%) transitioned from exposure to nonexposure. Among women exposed to nonweapon physical IPV (171 [15.4%]) or psychological abuse (726 [65.2%]), most reported exposure across multiple waves (88 [51.5%] and 479 [66.0%], respectively). Most women who ever reported gun or other weapon-involved IPV also reported both nonweapon physical IPV and psychological abuse (38 [88.4%]) at some time during the observation period. Across all waves, mean symptom counts were low for MDD (mean [SD], 0.88 [1.47]), PD (mean [SD], 1.18 [2.48]), PTSD (mean [SD], 0.82 [1.71]), and ASAD (mean [SD], 1.82 [2.53]). The proportions meeting the diagnostic threshold at any survey wave were 49.6% (552 of 1114 ) for ASAD, 33.6% (374 of 1114) for MDD, 22.8% (254 of 1114) for PD, and 11.0% (123 of 1114) for PTSD. Table 1. Descriptive Statistics Across Study Waves in the 1114 Study ParticipantsMeasureStudy wave1 (n = 895)2 (n = 1086)3 (n = 946)4 (n = 842)5 (n = 751)6 (n = 791)Domestic violence, No. (%) Weapon IPV8 (0.9)9 (0.8)15 (1.6)10 (1.2)12 (1.6)10 (1.3) Nonweapon IPV40 (4.5)55 (5.1)71 (7.5)55 (6.5)57 (7.6)45 (5.7) Psychological abuse362 (40.4)384 (35.4)338 (35.7)274 (32.5)243 (32.4)170 (21.5)Mental health MDD symptoms, mean (SD)1.08 (2.22)0.65 (1.77)0.72 (1.89)0.80 (2.03)1.04 (2.32)1.04 (2.35) MDD threshold (≥4), No. (%) 149 (16.6)104 (9.6)94 (10.0)97 (11.5)118 (15.8)82 (14.9) PD symptoms, mean (SD)0.91 (2.89)0.92 (2.89)1.03 (3.07)1.38 (3.61)1.38 (3.55)1.99 (4.38) PD threshold (≥4), No. (%)64 (7.2)81 (7.5)74 (7.9)93 (11.1)84 (11.2)86 (15.6)PTSD symptoms, mean (SD)0.64 (2.11)0.64 (2.06)0.95 (2.59)0.87 (2.44)1.04 (2.75)0.92 (2.52) PTSD threshold (≥8), No. (%)23 (2.6)26 (2.4)37 (3.9)30 (3.6)37 (4.9)24 (4.4) ASAD symptoms, mean (SD)2.20 (3.90)1.33 (3.18)1.99 (3.92)1.68 (3.59)1.93 (4.02)1.98 (4.03) ASAD threshold (≥3), No. (%)234 (26.1)176 (16.3)213 (22.6)169 (20.1)152 (20.3)120 (21.7)Demographics Migrant, No. (%)518 (57.9)570 (52.5)483 (51.1)416 (49.4)391 (52.1)403 (50.9) Year of birth, mean (SD)1987.0 (5.4)1986.5 (5.4)1986.5 (5.4)1986.4 (5.4)1986.4 (5.4)1986.4 (5.3)Abbreviations: ASAD, adult separation anxiety disorder; IPV, intimate partner violence; MDD, major depressive disorder; PD, panic disorder; PTSD, posttraumatic stress disorder. Descriptive Statistics Across Study Waves in the 1114 Study Participants Abbreviations: ASAD, adult separation anxiety disorder; IPV, intimate partner violence; MDD, major depressive disorder; PD, panic disorder; PTSD, posttraumatic stress disorder. Preliminary β̂ mixed-effects models, adjusted only for survey wave, indicate that gun and other weapon–involved IPV was associated with higher symptom counts for MDD (β̂ = 1.14; 95% CI, 0.43-1.85), PD (β̂ = 1.65; 95% CI, 0.92-2.38), PTSD (β̂ = 1.74; 95% CI, 1.03-2.45), and ASAD (β̂ = 0.78; 95% CI, 0.09-1.47). The pooled marginal person-wave difference in expected symptom count was 0.59 (95% CI, 0.15-1.39) for MDD, 1.04 (95% CI, 0.37-2.32) for PD, 1.17 (95% CI, 0.47-2.40) for PTSD, and 0.76 (95% CI, 0.05-2.00) for ASAD. Likewise, the estimated within-person change in expected symptom count from no exposure to exposure was 1.07 (95% CI, 0.13-1.33) for MDD, 2.39 (95% CI, 0.31-2.04) for PD, 2.00 (95% CI, 0.43-2.33) for PTSD, and 1.33 (95% CI, 0.03-1.97) for ASAD. As detailed in Table 2, adjustment for additional covariates attenuated these associations. Collinearity was not present according to correlation analysis (eTable 1 in Supplement 1) and variance inflation factor and tolerance statistics (eTable 2 in Supplement 1). Gun and other weapon–involved IPV remained significantly associated with PD (β̂ = 0.79; 95% CI, 0.08-1.50) and PTSD (β̂ = 1.07; 95% CI, 0.38-1.76) but not with MDD (β̂ = 0.12; 95% CI, −0.55 to 0.79) or ASAD (β̂ = −0.12; 95% CI, −0.81 to 0.57). The adjusted pooled marginal person-wave difference in expected symptom count was 0.57 (95% CI, 0.04-1.44) for PD and 0.69 (95% CI, 0.17–1.59) for PTSD, whereas the adjusted within-person change from no exposure to exposure was 1.12 (95% CI, 0.08-2.24) for PD and 1.14 (95% CI, 0.26-2.02) for PTSD. Corresponding estimates were null for MDD (adjusted pooled marginal person-wave difference: mean, 0.06; 95% CI, −0.22 to 0.49; adjusted within-person change from no exposure to exposure: mean, 0.11; 95% CI, −0.43 to 0.84) and ASAD (adjusted pooled marginal person-wave difference: mean, −0.10; 95% CI, −0.53 to 0.62; adjusted within-person change from no exposure to exposure: mean, −0.21; 95% CI, −1.04 to 1.01). Figure 1 displays the model-predicted distributions of expected symptom counts (pooled across waves) by gun and other weapon–involved IPV exposure, with uncertainty propagated from the model estimates. Figure 2 shows the corresponding distributions of model-estimated within-person changes in expected symptom count associated with transitioning from no exposure to exposure between consecutive waves, again incorporating fixed-effect uncertainty. Table 2. β̂ Mixed-Effects Models Estimating Association Between Weapon-Involved IPV and Mental Health Symptom CountsCovariateModel 1 (MDD)aModel 2 (PD)bModel 3 (PTSD)cModel 4 (ASAD)dβ̂ (SE)P valueβ̂ (SE)P valueβ̂ (SE)P valueβ̂ (SE)P valueIntercept−3.96 (0.16)<.001−3.89 (0.16)<.001−4.05 (0.16)<.001−3.12 (0.13)<.001Weapon IPV0.12 (0.34).730.79 (0.36).031.07 (0.35).002−0.12 (0.35).72Nonweapon IPV0.55 (0.18).0020.65 (0.19)<.0010.46 (0.19).0170.90 (0.17)<.001Psychological abuse0.57 (0.12)<.0010.66 (0.12)<.0010.04 (0.12).740.24 (0.11).03Lagged mental health0.18 (0.03)<.0010.11 (0.02)<.0010.11 (0.02)<.0010.08 (0.01)<.001Prior nonweapon IPV0.45 (0.17)<.0010.17 (0.12).390.50 (0.19).0080.18 (0.18).31Prior psychological abuse0.05 (0.12).660.07 (0.13).57−0.23 (0.12).070.07 (0.11).57Migrant0.49 (0.11)<.001−0.85 (0.14)<.001−0.11 (0.13).370.06 (0.11).56Year of birth0.01 (0.01).550.01 (0.01).64−0.04 (0.01).002−0.02 (0.01).09Survey wave 21.00 [Reference]NA1.00 [Reference]NA1.00 [Reference]NA1.00 [Reference]NA 30.14 (0.14).300.03 (0.13).810.37 (0.13).0030.21 (0.11).06 40.24 (0.14.080.23 (0.14).090.23 (0.13).09−0.08 (0.12).54 50.44 (0.14).0020.25 (0.14).070.37 (0.13).006−0.01 (0.13).99 60.40 (0.15).0090.58 (0.15)<.0010.18 (0.15).220.01 (0.14).96Abbreviation: ASAD, adult separation anxiety disorder; β̂, β-binomial; IPV, intimate partner violence; MDD, major depressive disorder; NA, not applicable; PD, panic disorder; PTSD, posttraumatic stress disorder. a Score range, 0-9; n = 3939 observations and 1111 participants. Dispersion parameter = 0.91; mean random intercept variance (SD) = 0.78 (0.88); Akaike information criterion = 6251.7; bayesian information criterion = 6345.3; and log likelihood = −3110.60. b Score range, 0-17; n = 3926 observations and 1110 participants. Dispersion parameter = 1.43; mean random intercept variance (SD) = 2.04 (1.43); Akaike information criterion = 7080.0; bayesian information criterion = 7174.1; and log likelihood = −3535.00. c Score range, 0-16) (n = 3923 observations and 1108 participants). Dispersion parameter = 2.16; mean random intercept variance (SD) = 1.71 (1.31); Akaike information criterion = 6906.3; bayesian information criterion = 7000.4; and log likelihood = −3438.1. d Score range, 0-14; n = 3945 observations and 1111 participants. Dispersion parameter = 0.74; mean random intercept variance (SD) = 1.19 (1.09); Akaike information criterion = 8820.0; bayesian information criterion = 8914.2; and log likelihood = −4395.0. β̂ Mixed-Effects Models Estimating Association Between Weapon-Involved IPV and Mental Health Symptom Counts Abbreviation: ASAD, adult separation anxiety disorder; β̂, β-binomial; IPV, intimate partner violence; MDD, major depressive disorder; NA, not applicable; PD, panic disorder; PTSD, posttraumatic stress disorder. Score range, 0-9; n = 3939 observations and 1111 participants. Dispersion parameter = 0.91; mean random intercept variance (SD) = 0.78 (0.88); Akaike information criterion = 6251.7; bayesian information criterion = 6345.3; and log likelihood = −3110.60. Score range, 0-17; n = 3926 observations and 1110 participants. Dispersion parameter = 1.43; mean random intercept variance (SD) = 2.04 (1.43); Akaike information criterion = 7080.0; bayesian information criterion = 7174.1; and log likelihood = −3535.00. Score range, 0-16) (n = 3923 observations and 1108 participants). Dispersion parameter = 2.16; mean random intercept variance (SD) = 1.71 (1.31); Akaike information criterion = 6906.3; bayesian information criterion = 7000.4; and log likelihood = −3438.1. Score range, 0-14; n = 3945 observations and 1111 participants. Dispersion parameter = 0.74; mean random intercept variance (SD) = 1.19 (1.09); Akaike information criterion = 8820.0; bayesian information criterion = 8914.2; and log likelihood = −4395.0. Graphs Showing Pooled (All-Wave) Adjusted Marginal Difference in Expected Symptom Distribution by Outcome ASAD indicates adult separation anxiety disorder; DV, domestic violence; MDD, major depressive disorder; PD, panic disorder; PTSD, posttraumatic stress disorder. Graphs Showing Pooled (All-Wave) Within-Person Adjusted Change in Expected Symptom Distributions at Exposure by Outcome ASAD indicates adult separation anxiety disorder; MDD, major depressive disorder; PD, panic disorder; PTSD, posttraumatic stress disorder. Sensitivity analyses assessed potential overadjustment, collider bias, model specification, and influential exposed participants. The exclusion of lagged mental health or concurrent nonweapon IPV and psychological abuse produced somewhat larger estimates, but gun and other weapon–involved IPV remained significantly associated only with PD and PTSD (eTables 3-4 in Supplement 1). Negative binomial mixed-effects models produced results consistent with the primary β̂ models, suggesting that conclusions were not sensitive to the distributional specification (eTable 5 in Supplement 1). Leave-one-exposed-out analysis, which involved repeating the fully adjusted models after sequentially excluding each exposed participant, indicates that results were not driven by any single exposed participant: PD and PTSD estimates remained consistently positive and similar to the original estimates, whereas MDD and ASAD estimates remained small and close to null (eTable 6 in Supplement 1). Discussion In this study of 1114 women and 6 waves of longitudinal data, we examined whether gun and other weapon–involved IPV was associated with women’s mental disorder symptom burden beyond other forms of IPV and baseline risk. Gun and other weapon–involved IPV usually co-occurred with nonweapon physical IPV and psychological abuse. In models adjusted only for survey wave, gun and other weapon–involved IPV was associated with higher symptom counts across all mental disorder outcomes. After full adjustment for concurrent and prior exposure to nonweapon physical IPV, psychological abuse, and covariates, associations attenuated markedly: gun and other weapon–involved IPV remained independently associated with higher PD and PTSD symptom counts but not for MDD or ASAD. The pooled marginal contrasts and within-person transition estimates indicate that gun and other weapon–involved IPV was associated with a mean of 0.57 additional PD symptoms and 0.69 additional PTSD symptoms per person-wave and with 1.12 additional symptoms on both outcomes when women transitioned from nonexposure to exposure. These findings are consistent with the premise that gun and other weapon involvement marks a distinct threat context characterized by perceived lethality, unpredictability, and loss of control.4,5,6,13,15,16 Threats with guns and other weapons can be experienced as imminently life-threatening, plausibly intensifying hyperarousal and leading to intrusive reexperiencing and physiologic fear responses that align closely with PTSD13,14 and panic symptoms,15,16,31 both of which often also occur together due to their similarity.16 It is crucial to recognize that nonfatal gun and other weapon use as a form of IPV represents an extremely serious risk to a person’s life, given the inherent lethality of the weapon and the documented patterns in which such threats are used as tools of coercive control, often preceding or facilitating subsequent homicide.4,5,6,11 The within-person transition estimates are consistent with this interpretation because they focus on symptom changes around observed exposure transitions within the same individuals, reducing, but not eliminating, concern about stable between-person differences. Furthermore, because we adjusted for other IPV domains and prior symptoms, the PD and PTSD findings are best interpreted as incremental associations with weapon involvement conditional on broader IPV exposure and prior symptom burden. Clinically, these findings support treating gun and other weapon involvement as an important marker during IPV assessments, not only in relation to heightened risk of serious physical injury but also in terms of mental harm. Routine inquiry should explicitly include not only firearms but also knives and other weapons because the mechanisms appear to operate through perceived lethality and credible threat.11 When gun and other weapon involvement is disclosed or suspected, clinicians should anticipate elevated risk of PD and PTSD symptoms and interpret short-term increases in hypervigilance, sleep disturbances, panic attacks, somatic anxiety, and trauma reexperiencing as potentially linked to IPV. Importantly, clinical response should integrate symptom management with rapid safety-focused strategies to alleviate biophysiologic symptoms of short-term fear for life and the related PTSD and PD symptoms.32 Similar support ought also to be extended to the children of survivors, who appear to be at risk for trauma-related disorders associated with weapon use against their parents.33 As such, trauma-informed care, stabilization strategies, and referral to specialist family violence services should be coordinated with safety planning, including discussion of escalation risk, help-seeking options, and supports that reduce isolation and improve protective capacity. Mental health assessment and safety planning should be routine for all patients disclosing IPV. Our findings suggest that disclosure of gun and other weapon involvement provides additional clinical information, identifying a subgroup at particularly high risk for PTSD and panic symptoms who may warrant heightened vigilance and more urgent safety-focused responses. With regard to prevention, the results suggest that weapon access is relevant to IPV not only because of fatal outcomes but also because of nonfatal, psychologically consequential coercive control. Although women are murdered by intimate partners using a range of methods, including strangulation, blunt force trauma, fire, and other forms of physical violence, the involvement of guns or other weapons signals a credible lethal threat that requires particular attention. Weapon threats may function as force multipliers within abusive relationships by increasing credibility of harm and constraining autonomy, which can amplify trauma and panic symptoms. This supports a prevention framing in which gun and other weapon access restrictions and enforcement mechanisms are understood as potential levers for reducing both physical danger and fear-based psychological burden. System responses across health, police, and community agencies that identify weapon involvement early may be particularly valuable for triage and coordination, given that weapon involvement is relatively infrequent but highly informative and strongly comorbid with other IPV types. At a minimum, the findings reinforce the importance of integrated approaches that link firearm and weapon regulation, IPV risk management, and mental health supports instead of treating these as separate policy domains. Strengths and Limitations This study has several strengths, including the prospective multiwave design of systematically collected data, adjustment for prior symptom levels, and estimation of both pooled and within-person exposure contrasts. The β̂ mixed-effects models accounted for bounded, overdispersed symptom counts and within-person dependence, whereas marginal and transition contrasts improved interpretability. The study also has some limitations, including the rarity of reported gun and other weapon–involved IPV, which reduced precision and may have increased sensitivity to model specification and residual confounding, particularly for transition-based estimates. Underreporting is also possible because gun and other weapon use in IPV is explicitly life-threatening and criminal, which may reduce disclosure relative to other IPV forms. Exposures and outcomes were self-reported and may be affected by recall or disclosure bias, and differing symptom reference periods across disorders may weaken temporal alignment with past-year IPV exposure. Lastly, estimates should not be interpreted as definitive causal effects. Gun and other weapon–involved IPV frequently co-occurred with other IPV types and may be embedded within broader trajectories of escalating abuse. Although the transition analyses reduce concern about stable between-person differences, random-intercept models do not fully isolate within-person effects and may still be biased if unmeasured individual characteristics are correlated with weapon-involved IPV exposure. Despite sensitivity analyses supporting the main findings, residual confounding, overadjustment, and time-varying confounding remain possible. Conclusions In this longitudinal cohort study of women, gun and other weapon–involved IPV was associated with higher PD and PTSD symptom counts after adjustment for co-occurring and prior IPV experiences, prior psychiatric symptoms, and demographic factors. These findings suggest that gun and other weapon involvement may mark a distinct high-threat IPV context linked particularly to fear and trauma-related symptoms rather than to psychiatric symptom burden generally. Routine assessment of IPV should explicitly inquire about threats or assaults involving guns, knives, and other weapons because their disclosure may identify women requiring urgent safety-focused and clinical intervention. The findings support public health and policy approaches that recognize gun and other weapon access as relevant not only to homicide prevention but also to the substantial psychological burden associated with nonfatal weapon-related coercive control. Greater integration of firearm and weapon regulation, family violence prevention, and trauma-informed mental health care is required, along with coordinated safety responses to reduce the immediate threat to life and the psychological consequences experienced by survivors and their children.