Table 2. The relationship between adolescent suicide rates and socioeconomic indicators.

	β	Wald 95%CI	Wald χ2 (df)	p-value
Total suicide rate, age 10-19
GDP	1.90	-5.97 to 9.76	0.22(1)	0.637
Unemployment	-0.03	-0.19 to 0.14	0.12(1)	0.733
Gini	10.69	2.32 to 19.05	6.27(1)	0.012*
Male suicide rate, age 10-19
GDP	4.04	-5.84 to 0.00	0.64(1)	0.423
Unemployment	-0.03	-0.21 to 0.16	0.07(1)	0.788
Gini	7.30	-2.31 to 16.90	2.22(1)	0.136
Female suicide rate, age 10-19
GDP	-5.82	-8.41 to -3.23	19.37(1)	< 0.001†
Unemployment	0.06	0.02 to 0.10	7.10(1)	0.008*
Gini	9.99	8.23 to 11.75	122.69(1)	< 0.001†
Total suicide rate, age 10-14
GDP	0.00	0.00 to 3.47	3.66(1)	0.056
Unemployment	0.12	-0.02 to 0.26	2.96(1)	0.085
Gini	-0.62	-12.40 to 11.16	0.01(1)	0.918
Total suicide rate, age 15-19
GDP	-3.34	0.00 to 5.34	0.57(1)	0.122
Unemployment	-0.02	-0.17 to 0.13	0.07(1)	0.796
Gini	9.63	2.31 to 16.96	6.64(1)	0.010*
Male suicide rate, age 15-19
GDP	5.19	-5.29 to 0.00	0.94(1)	0.332
Unemployment	0.01	-0.13 to 0.15	0.03(1)	0.854
Gini	4.93	-1.60 to 11.45	2.19(1)	0.139
Female suicide rate, age 15-19
GDP	3.38	-9.21 to 0.00	0.57(1)	0.122
Unemployment	0.02	-0.09 to 0.12	0.07(1)	0.796
Gini	13.20	9.23 to 17.14	6.64(1)	0.010*

95%CI = 95% confidence interval; df = degrees of freedom; GDP = gross domestic product; QIC = quasi-likelihood under independence model criterion.

Generalized estimating equation model results. Seven separate models were tested. The six large cities were used as the subject variable and the year was used as the repeated measure variable. The dependent variable was modeled with gamma distribution. The structure that provided the model with the lowest QIC was selected, and this was independent structure in all models, except the suicide rate among 15- to 19-year-old females, for which the unstructured model provided the lowest QIC.

^*p < 0.05;

^†p < 0.01.