. 2011 Jun 9;6(6):e21002. doi: 10.1371/journal.pone.0021002

Table 5. Multivariate linear regression using log-transformed values of cytokines.

	IL-1β	IL-6	TNF-α	CRP
Men(yes vs. no)	−0.0638(0.0495)	0.0756(0.0366)^*	0.0517(0.0228)^*	−0.2210(0.0261)^***
Age	−0.0077(0.0023) ^***	0.0029(0.0017)	0.0071(0.0011)^***	0.0128(0.0012)^***
BMI	−0.0074(0.0055)	0.0224(0.0041)^***	0.0163(0.0026)^***	0.0969(0.0029)^***
Former smoker(yes vs. no)	0.0080(0.0572)	0.0965(0.0425)^*	0.0297(0.0264)	−0.0030(0.0301)
Current smoker(yes vs. no)	0.1118(0.0605)	0.2524(0.0450)^***	0.1165(0.0280)^***	0.2327(0.0321)^***
Leisure-time PA(yes vs. no)	−0.0035(0.0512)	−0.0276(0.0378)	−0.0063(0.0236)	−0.1520(0.0270)^***

Results are expressed as slope and (standard error). BMI, body mass index; hs-CRP, high sensitive C reactive protein; IL-1β, interleukin-1β; IL-6, interleukin-6; PA, physical activity; TNF-α, tumor necrosis factor-α. Statistical analysis by linear regression on log-transformed cytokine values:

*, p<0.05;

***, p<0.001.