Table 2 Linear and Logistic Regression for Simple and Entire Models and Association with CRP.

From: Psychiatric symptoms are not associated with circulating CRP concentrations after controlling for medical, social, and demographic factors

2 A. Simple Model

PHQ-9

DAST

  

beta

p-value

beta

p-value

Linear Regression

All

0.125

0.0000002

0.012

0.564

 

Female

0.091

0.003

−0.048

0.115

 

Male

0.154

0.0001

0.120

0.003

Logistic Regression

All

0.252

0.0000003

0.033

0.495

 

Female

0.171

0.005

−0.061

0.324

 

Male

0.284

0.0006

0.183

0.027

2B. Entire Model

 

PHQ−9

 

DAST

 
  

beta

p-value

beta

p-value

Linear Regression

All

−0.0040

0.872

−0.027

0.305

 

Female

−0.034

0.248

−0.088

0.004

 

Male

0.009

0.833

0.044

0.324

Logistic

Regression

All

0.055

0.397

−0.067

0.329

 

Female

−0.060

0.497

−0.228

0.014

 

Male

0.073

0.505

0.100

0.382

  1. We performed linear and logistic regression using two different models, where one model included only PHQ-9 and DAST (simple, unadjusted model) and the other included all 36 variables. Beta Coefficients are reported. (A) For the unadjusted dataset, we analyzed the association between CRP and PHQ-9 and between CRP and DAST. PHQ-9, but not DAST, was significantly associated with CRP in unadjusted/simple models in the (All; β = 0.125; β = 0.252) and female-only sample (β = 0.091; β = 0.171). However, for the male-only sample, both PHQ-9 (β = 0.154; β = 0.284) and DAST (β = 0.120; β = 0.183) showed significance in unadjusted models. (B) For the All and male-only samples, PHQ-9 and DAST were not associated with CRP after adjusting for the covariates of interest. In the female-only sample, DAST (β = −0.088; β = −0.228) but not PHQ-9 was significantly associated with CRP in the adjusted/entire model.
Back to article page