Table 1

Joinpoint analyses for total suicide rates and rates for specific methods among men by age group, Quebec, 1981–2006

	Joinpoint (years)	Segment	APC^†	95% CI
Men aged 15–34 years
All methods: total rate		1981–1999	1.1^*	(0.5 to 1.7)
	1999	1999–2006	−8.4^*	(−11.0 to −5.7)
Hanging		1981–1999	5.5^*	(4.5 to 6.4)
	1999	1999–2006	−7.1^*	(−10.4 to −3.6)
Firearms		1981–1996	−2.7^*	(−4.1 to −1.3)
	1996	1996–2006	−11.1^*	(−14.8 to −7.2)
Gas poisoning		1981–1995	4.6^*	(1.9 to 7.3)
	1995	1995–2006	−13.0^*	(−17.3 to −8.4)
Solid or liquid poisoning		1981–2006	−5.2^*	(−6.5 to −4.0)
Other methods		1981–2006	−1.0	(−2.1 to 0.2)
Men aged 35–64 years
All methods: total rate		1981–1999	−0.8	(−3.1 to 1.4)
	1989	1989–1999	3.2^*	(1.5 to 4.9)
	1999	1999–2006	−3.5^*	(−5.6 to −1.4)
Hanging		1981–1991	1.0	(−1.6 to 3.8)
	1991	1991–1999	10.4^*	(6.7 to 14.1)
	1999	1999–2006	−3.9^*	(−6.5 to −1.3)
Firearms		1981–1997	−1.2^*	(−2.0 to −0.4)
	1997	1997–2006	−5.6^*	(−7.6 to −3.6)
Gas poisoning		1981–1995	6.0^*	(3.7 to 8.4)
	1995	1995–2006	−6.3^*	(−8.8 to −3.7)
Solid or liquid poisoning		1981–2006	−0.8	(−1.9 to 0.3)
Other methods		1981–2006	−0.5	(−1.4 to 0.3)
Men aged 65 years and over
All methods: total rate		1981–2006	−0.6	(−1.3 to 0.1)
Hanging		1981–2006	0.2	(−0.9 to 1.4)
Firearms		1981–2006	−1.0	(−2.0 to 0.1)
Gas poisoning^‡		1981–2006	–	–
Solid or liquid poisoning^‡		1981–2006	–	–
Other methods		1981–2006	−2.8^*	(−3.9 to −1.7)
Men: all ages^§
All methods: total rate		1981–1999	1.3^*	(0.7 to 1.9)
	1999	1999–2006	−4.2^*	(−6.3 to −2.0)
Hanging		1981–1999	5.0^*	(4.2 to 5.8)
	1999	1999–2006	−3.2^*	(−5.7 to −0.7)
Firearms		1981–1997	−1.7^*	(−2.3 to −1.1)
	1997	1997–2006	−6.8^*	(−8.3 to −5.2)
Gas poisoning		1981–1995	5.4^*	(3.7 to 7.1)
	1995	1995–2006	−8.1^*	(−10.2 to −6.0)
Solid or liquid poisoning		1981–2006	−1.5^*	(−2.5 to −0.5)
Other methods		1981–2006	−1.0^*	(−1.8 to −0.2)

Sources: MSSS, Quebec death database from 1981 to 2006.
MSSS, demographic outlook.
↵* Annual percentage change (APC) is statistically significantly different from zero (two-sided p<0.05, calculated using a t test).
↵† APC, that is, the slope of the line segment.
↵‡ Dropped from the analysis because the log-linear model does not deal with a rate equal at zero.
↵§ Age-adjusted rates, Quebec population in 2001.