Comparisons of adipocyte diameter between timepoints using a negative binomial regression model.

ADIPOCYTE_SIZE_STATS

Format

A data.frame with 40 rows and 11 variables:

contrast

factor; the comparison being made ("4W / SED" or "8W / SED").

diameter_bin

factor; adipocyte diameter binned in approximately 5 micron intervals.

sex

factor; the sex of the rat with levels "Female" and "Male".

ratio

numeric; ratio of the trained adipocyte rate to the sedentary adipocyte rate. See details.

SE

numeric; standard error for the ratio.

df

integer; degrees of freedom for the test.

asymp.LCL

numeric; lower bounds of the 95% confidence interval.

asymp.UCL

numeric; upper bounds of the 95% confidence interval.

null

integer; value for the null hypothesis: the ratio specified by contrast is 1.

z.ratio

numeric; Z-statistic.

p.value

numeric; Dunnett p-value.

Details

Adipocytes are binned in approximately 5 micron intervals according to their diameter (diameter_bin). The number of adipocytes in each bin per animal, as well as the total number of adipocytes are tallied per animal. The log of the total adipocytes per animal is included as an offset term in a negative binomial regression model that contains all main effects and interactions between sex, timepoint, and diameter_bin as predictors of the number of binned adipocytes. The offset term allows for modeling of the number of binned adipocytes per total number of adipocytes by animal (modeled as a rate).

Within each combination of sex and diameter bin, the Dunnett procedure is applied to test for differences between the trained timepoints (4W, 8W) and the sedentary group (SED).

See vignette("ANALYTES_STATS") for details.