Table 3. Hierarchical Linear Model Results Predicting C-Reactive Protein From %FAT, Physical Activity, Socioeconomic Status (SES), and Sex Among First-Year College Students (n = 177), 2013.

Fixed Effects	Parameter	Log CRP					P Valuef
		Model Aa	Model Bb	Model Cc	Model Dd	Model E, β (SE)e
Composite model
Intercept	γ00	−0.26	−0.26	−0.26	−0.32	−0.32 (0.07)	<.001
SES	γ01	—	0.04	0.04	0.03	0.03 (0.02)	.05
Sexg	γ10	—	—	0.10	0.18	0.18 (0.13)	.17
%FAT	γ20	—	—	—	2.28	2.15 (0.75)	.006
Physical activityh	γ30	—	—	—	—	−0.01 (0.15)	.48
Goodness of fit, deviance	NA	303.46	299.11	297.83	288.39	287.89	NA

Abbreviations: %FAT, relative adiposity measured by dual x-ray absorptiometry; SE, standard error; —, not included in model; NA, not applicable.

^a Analysis of variance model to detect differences between SES groups.

^b Area-level SES included as a level-2 predictor.

^c Sex included as level-1 predictor, area-level SES included as a level-2 predictor.

^d Sex and %FAT included as level-1 predictors, area-level SES included as a level-2 predictor.

^e Full model with sex, %FAT, and physical activity included as level-1 predictors, area-level SES included as a level-2 predictor.

^f Calculated by using χ² distribution with degrees of freedom equal to the number of parameters estimated.

^g Sex classified as female or male (coded 0 or 1, respectively).

^h Measured via accelerometer and reported as steps per day divided by 1,000 so that all variables are expressed on a similar scale.