Skip to main content
. 2015 Nov 4;41(6):1447–1456. doi: 10.1038/npp.2015.309

Table 3. Regression Models for the Relationships Between Suicidal Behavior and Total Cortisol Output With and Without Controlling for Demographics and Hour of the Day When the TSST was Conducted.

  Total Cortisol Outputa
  Unstandb β 95% CI t p Unstandb β 95% CI t p
SA −0.47 −0.83, −0.11 −2.58 0.01 −0.47 −0.89, −0.05 −2.2 0.03
SRB −0.13 −0.49, 0.23 −0.70 0.49 −0.14 −0.58, 0.29 −0.65 0.52
NS −0.15 −0.32, 0.03 −1.61 0.11 −0.11 −0.42, 0.20 −0.69 0.49
Age         0.002 −0.02, 0.02 0.27 0.79
Sex, femalec         −0.22 −0.40, −0.04 −2.43 0.02
Race, Caucasianc         0.21 0.04, 0.38 2.45 0.02
Ethnicity, Hispanicc         0.18 −0.23, 0.59 0.88 0.38
Site, Pittsburghc         −0.21 −0.49, 0.08 −1.4 0.16
SESd         0.001 −0.003, 0.005 0.51 0.61
Time of day         −2.53 −4.03, −1.03 −3.34 0.001

Abbreviations: NS, non-suicidal; SA, suicide attempt; SRB, suicide-related behavior.

a

Total cortisol output=ln(AUCG).

b

Unstandardized coefficients.

c

Compared with reference categories males, non-Caucasians, non-Hispanic, and NY.

d

SES, socio-economic status. There were 208 high-risk offspring from 134 families with an average number of 1.6 (SD=0.7, range 1–5) offspring per family participating in the study. Regression models take into account the correlation between subjects.

d

Bold values indicate p<0.05.