. 2023 Jun 1;131(6):067003. doi: 10.1289/EHP11651

Table 3.

Circulating extracellular inflammatory markers and change in plasma volume at hour 9 of the simulated heat wave.

	Control [mean (SD)]^a	Cooling [mean (SD)]^a	Difference [mean (95% CI)]^b	$p$ -Value^c
Interleukin-6 (pg/mL)
End of heat wave simulation (hour 9)	6.19 (2.91)	5.90 (3.39)	$negative 0.52$ ( $negative 1.31$ , 0.27)	0.19
Change from pre-exposure to end-simulation	0.19 (1.08)	$negative 0.06$ (0.29)
Tumor necrosis $factor- α$ (pg/mL)
End of heat wave simulation (hour 9)	11.90 (9.85)	13.12 (11.98)	$negative 0.34$ ( $negative 1.94$ , 1.26)	0.67
Change from pre-exposure to end-simulation	0.30 (1.63)	0.00 (2.40)
C-reactive protein (mg/L)
End of heat wave simulation (hour 9)	0.35 (0.20)	0.35 (0.19)	$negative 0.03$ ( $negative 0.07$ , 0.01)	0.16
Change from pre-exposure to end-simulation	0.00 (0.04)	$negative 0.03$ (0.07)
Plasma volume (%)
Change from pre-exposure to end-simulation	$negative 2.90$ (4.0)	$negative 2.2$ (2.8)	0.4 ( $negative 1.8$ , 2.6)	0.71

Note: CI, confidence interval; SD, standard deviation.

^{^a}

Mean and SD of raw data for each inflammatory marker were measured at the end of the heat wave simulation (hour 9), as well as the change in plasma volume from preexposure (baseline) levels for the control ( $lowercase italic n equals 13$ for interleukin-6 and tumor necrosis $factor- α$ ; $lowercase italic n equals 17$ for C-reactive protein and plasma volume) and cooling groups ( $lowercase italic n equals 17$ for interleukin-6 and tumor necrosis $factor- α$ ; $lowercase italic n equals 19$ for C-reactive protein and plasma volume). Missing data were not imputed.

^{^b}

Estimated marginal mean difference and 95% CI were derived from a linear mixed-effects model including fixed effects of group (control vs. cooling) and covariates for sex and baseline values (i.e., value of each inflammatory marker measured prior to heat exposure; inflammatory markers) or sex only (change in plasma volume). Note that the mean difference, which is adjusted for included covariates, may not correspond to the mathematical difference between the summary data presented for the control and cooling groups (see footnote a).

^{^c}

$p$ -Values for between-group comparisons were derived from the linear mixed-effects models.