Racial disparities in adult all-cause and cause-specific mortality among us adults: mediating and moderating factors

Table 2 Cox PH of race/ethnicity vs. all-cause mortality: Crude (Model 0)^a and socio-demographic factor-adjusted (Model 1)^a NHANES III

	Model 0: Crude^b			Model 1: sociodemographic factor-adjusted^c
	Log_e(HR)	95 % CI	P	Log_e(HR)	95 % CI	P
<50y, women, PIR ≥ 125 % ^d	(N = 3105)			(N = 3105)
NHB vs. NHW	+0.12	(-0.33;+0.58)	0.60	−0.07	(-0.59;+0.46)	0.81
MA vs. NHW	−0.26	(-0.84;+0.32)	0.37	−0.21	(-0.84;+0.42)	0.43
<50y, men, PIR ≥ 125 % ^d	(N = 2860)			(N = 2860)
NHB vs. NHW	+0.87	(+0.50;+1.24)	<0.001	+0.95	(+0.55;+1.34)	<0.001
MA vs. NHW	+0.22	(-0.26;+0.70)	0.37	+0.49	(-0.00;+0.99)	<0.001
<50y, women, PIR < 125 % ^d	(N = 1728)			(N = 1728)
NHB vs. NHW	+0.34	(-0.55;+1.24)	0.44	+0.45	(-0.43;+1.32)	0.31
MA vs. NHW	−0.20	(-1.11;+0.70)	0.66	−0.18	(-1.12;+0.76)	0.71
<50y, men, PIR < 125 % ^d	(N = 1269)			(N = 1269)
NHB vs. NHW	−0.05	(-0.57;+0.46)	0.55	−0.07	(-0.61;+0.47)	0.80
MA vs. NHW	−0.83	(-1.37;-0.28)	0.003	−0.41	(-0.98;+0.15)	0.15
≥50y, women, PIR ≥ 125 % ^d	(N = 2420)			(N = 2420)
NHB vs. NHW	−0.03	(-0.22;+0.17)	0.80	+0.23	(+0.01;+0.46)	0.045
MA vs. NHW	−0.66	(-0.95;-0.38)	<0.001	−0.42	(-0.83;-0.02)	0.040
≥50y, men, PIR ≥ 125 % ^d	(N = 2443)			(N = 2443)
NHB vs. NHW	+0.14	(-0.04;+0.32)	0.13	+0.27	(+0.08;+0.46)	0.005
MA vs. NHW	−0.15	(-0.39;+0.08)	0.20	+0.04	(-0.21;+0.29)	0.75
≥50y, women, PIR < 125 % ^d	(N = 1101)			(N = 1101)
NHB vs. NHW	−0.10	(-0.32;+0.12)	0.38	−0.00	(-0.27;+0.28)	0.99
MA vs. NHW	−0.61	(-0.88;-0.34)	<0.001	−0.25	(-0.61;+0.10)	0.16
≥50y, men, PIR < 125 % ^d	(N = 861)			(N = 861)
NHB vs. NHW	+0.09	(-0.18;0.36)	0.51	+0.09	(-0.16;+0.34)	0.49
MA vs. NHW	−0.56	(-0.87;-0.26)	<0.001	−0.29	(-0.60;+0.03)	0.07

Abbreviation: CI Confidence Interval, exp exponent, HR Hazard Ratio, LCL Lower confidence limit, Log_e Natural logarithm, MA Mexican-American, NHANES National Health and Nutrition Examination Surveys, NHB Non-Hispanic Black, NHW Non-Hispanic White, PIR Poverty Income Ratio, SE Standard error, UCL Upper confidence limit
^aValues are the natural log of hazard ratios (HR) and 95 % CI with p-values, taking into account unequal probability of selection or sampling weights. Statistical significance is inferred from a 95 % CI not crossing the value of zero
^bModel 0 is crude unadjusted HR
^cModel 1 adjusted the HR for age, sex, marital status, urban-rural area of residence, and household size, within each age group/sex/poverty status stratum. Note that the point estimate of the HR can be computed as exp(β) where β = Log_e(HR). The 95 % CI for the HR is computed as exp^{(β±1.96*SE(β))}, whereby SE(β) = (UCL_β -LCL_β)/3.92

ISSN: 1471-2458