Table 2 Multivariate logistic regression models of the association between living in a household with a firearm and known suicidal risk factors/suicidal behaviours among respondents of the National Comorbidity Survey Replication (n = 5554)

Psychiatric disorders and suicidal behaviours in past 12 months	Home with firearms
Anxiety disorders
Panic disorder	1.1 (0.7 to 1.7)
Generalised anxiety disorder	1.0 (0.8 to 1.2)
Post-traumatic stress disorder	0.9 (0.6 to 1.2)
Any of the above anxiety disorders	1.0 (0.8 to 1.2)
Mood disorders
Major depressive disorder	0.9 (0.7 to 1.2)
Bipolar disorder	0.7 (0.5 to 1.1)
Any of the above mood disorders	0.9 (0.7 to 1.1)
Substance use disorders
Alcohol abuse	0.9 (0.6 to 1.3)
Alcohol dependence	0.5 (0.3 to 1.0)
Drug abuse	0.8 (0.5 to 1.5)
Drug dependence	0.9 (0.3 to 2.7)
Any of above substance use disorders	0.9 (0.6 to 1.3)
Any mood/anxiety/substance disorder	0.9 (0.8 to 1.2)
Suicidality
Suicidal ideation	0.8 (0.5 to 1.3)
Suicidal plan	0.5 (0.2 to 1.4)
Serious suicidal attempt	0.0†
Treatment-related characteristics
Saw professional for mental health	0.8 (0.7 to 1.1)
Hospitalised for psychiatric disorder	0.4 (0.2 to 0.9)*

Values are odds ratio (95% CI). Each model includes the following covariates: age, race, education, poverty/income ratio, region of country, childhood urbanicity, and sex. The Taylor series estimation method was used to estimate sampling errors, accounting for the stratified, clustered sampling design; weights were applied to all models to adjust for differential probabilities of selection, non-response and differences between the sample and 2000 US census population estimates.
*Significant difference at p<0.05.
†95% CI could not be estimated because of no serious suicide attempts in households with firearms.