Predicting mortality with biomarkers: a population-based prospective cohort study for elderly Costa Ricans

Table 3 Death rate ratios of normalized biomarkers estimated with several specifications of hazard regression models (effect of 1 SD on the death rate)

Biomarkers and control variables	Demo-graphic model	Partial models	Full model 1	Full model 2
Metabolic hormones
Glycosylated hemoglobin (HbA1c)		1.23**	1.23**	1.24**
Cardiovascular biomarkers
Systolic blood pressure (SBP)		0.85**	0.89*	0.90+
quared SBP		1.05*	1.03	1.03
SBP * Male		1.15+	1.13	1.15+
Metabolic – lipids
Total cholesterol		0.73**	0.83**	0.85*
Squared total chol.		1.08**	1.04+	1.03
HDL cholesterol		1.26**	1.19*	1.16*
Total/HDL cholesterol ratio		1.24**	1.19**	1.16*
Stress hormones
DHEAS		0.79**	0.83**	0.84*
Squared DHEAS		1.08**	1.05*	1.04
Inflammation, immune system
CRP		1.43**	1.27**	1.24**
Squared CRP		0.96**	0.97*	0.97+
Organ-specific functional reserve
Handgrip strength		0.55**	0.59**	0.65**
Handgrip * Male		1.62**	1.60**	1.53**
Distance in 10 seconds		0.75**	0.79**	0.90+
Pulmonary peak flow		0.71**	0.72**	0.73**
Nutrition, body size
Waist circumference		1.04	0.99	1.00
Squared waist circumference		1.06**	1.06*	1.06*
BMI * age 80+		0.84*	0.94	0.92
Baseline health controls
Normalized disability scale		1.59**		1.27**
Cancer diagnosis		1.46**		1.51**
Involuntary weight loss		1.24*		1.17
Current smoker		1.43+		1.50*
Bad self-reported health		1.22		1.12
Demographic controls
Male	1.65*		4.31**	3.76**
Male * Age	0.99		0.99	0.99
Nicoya	0.71*		0.66**	0.73*
Age (1 extra year)	1.09**		1.05**	1.05**

Significance assessed with robust estimates of standard errors: + means significant at *p < 0.10; at **p < 0.05; and at p < 0.01.
Partial models include the biomarkers of the respective group AND demographic controls. N = 2,313 individuals, 11,433 observed age segments, and 564 deaths.

ISSN: 1478-7954