Table 2

 Individual level studies of the link between firearm availability and suicide

AuthorsNumber of suicide victims (cases)*Comparison groupOutcome assessed/exposureOR (CI)†/p value
*Suicide victims in all studies cited in this table were community dwellers. †Odds ratio (OR) =  95% confidence interval (CI). RS, risk of suicide; RFS, risk of firearm suicide; SMR, standardised mortality ratio.
(A) Adult individual level studies
Bailey et al, 19977120 females120 matched community dwelling controlsRS among women/⩾ one gun in home4.6 (1.2 to 17.5)
Beautrais et al., 1996341971028 community dwelling controlsRS/access to a firearm at home1.4 (0.96 to 1.99)
Conwell et al, 20023786, ages 50 and over86 matched community dwelling controlsRS/⩾ one gun in home3.23 (1.15 to 11.20)
Cummings et al, 1997303531756 matched community dwelling controlsRS/history of family handgun purchase1.9 (1.4 to 2.5)
Grassel et al, 200332887 suicide victims and 1546 gun suicide victims208738 non-injury deaths-RS6.8 (5.7 to 8.1)
-RFS/handgun purchase12.5 (10.4 to 15.0)
Kellermann et al, 199236438438 matched community dwelling controlsRS/⩾ one gun in home4.8 (2.7 to 8.5)
Kung et al, 20033522772448710 decedents with natural causes of deathRS/any firearms in or around decedent’s home during last year of lifeMen living with others 3.53 (2.42 to 5.15)
Women living with others 2.99 (1.58 to 5.65)
Wiebe, 20038195913535 community dwelling controls-RS3.44 (3.06 to 3.86)
-RFS/having a gun in home16.89 (13.26 to 21.52)
Wintemute et al, 199931238,292 handgun purchasers in California—observed 1991–1996General adult population of California-RS in first year after handgun purchaseSMR 4.31
-RFS in first year after handgun purchaseSMR 7.12
(B) Adolescent individual level studies
Brent et al, 1988422756 suicidal psychiatric inpatientsRS/firearms available in home2.7 (1.1 to 6.4)
Brent et al, 19914147(a) 47 suicide attemptersRS/gun available in homea: 2.1 (1.2 to 3.7)
(b) 47 never suicidal psychiatric controlsb: 2.2 (1.4 to 3.5)
Brent et al, 1993387, all with no apparent psychopathology(a) 60 adolescent suicide victims with psychiatric disorderRS/gun available in homea: none reported
(b) 38 community controls with no psychiatric disorderb: p = 0.04
Brent et al, 1993436767 matched community dwelling controlsRS/any gun available in the home3.3 (1.4 to 7.7)
Brent et al, 19944463, all with a history of affective illness23 community dwelling controls with a history of affective illnessRS/handgun available in the homep<0.01
Brent et al, 199940140131 community dwelling controlsRS/gun in home4.00 (1.30 to 14.70)
Shah et al, 20003936 firearm suicide victims36 matched community dwelling controlsRFS/household access to firearms3.91 (1.11 to 13.80)