Table 3.
Change in individual metabolic variables associated with low GFR relative to normal/near-normal GFRa
| Metabolic Variable | Model (β [95% CI])b
|
||||
|---|---|---|---|---|---|
| Ac | Bd | Ce | Df | Eg | |
| Waist circumference | 0.87 (−2.29 to 4.02) | – | – | – | – |
| SBP | 0.22 (−3.83 to 4.27) | – | – | – | – |
| DBP | −0.44 (−2.81 to 1.93) | – | – | – | – |
| Fasting insulin concentrationh | 0.32 (0.15 to 0.48) | 0.30 (0.13 to 0.47) | 0.22 (0.08 to 0.36) | 0.17 (0.04 to 0.30) | 0.17 (0.04 to 0.30) |
| Insulin sensitivity index (Si)h | −0.15 (−0.30 to −0.01) | −0.15 (−0.29 to 0.00) | −0.10 (−0.21 to 0.02) | – | – |
| AIRh | 0.07 (−0.14 to 0.27) | – | – | – | – |
| Proinsulin-to-insulin ratioh | 0.11 (−0.07 to 0.29) | – | – | – | – |
| Fasting leptin concentrationh | 0.27 (0.09 to 0.44) | 0.26 (0.09 to 0.44) | 0.20 (0.07 to 0.33) | 0.17 (0.04 to 0.30) | 0.18 (0.05 to 0.30) |
| CRP concentrationh | 0.24 (−0.06 to 0.54) | – | – | – | – |
Covariates as shown were fit into the respective models in addition to all covariates in the previous models in order to examine to what degree the relation of low GFR to individual metabolic variables was explained by the respective covariates. Additional models were not tested when the regression coefficient was not statistically significant.
The regression coefficient (β) quantified the amount of change in individual metabolic variables associated with low GFR relative to normal/near-normal GFR. Coefficients with 95% CI crossing 0 were considered not statistically significant.
Included the following covariates: Age, gender, race/ethnicity, and clinic location.
Included covariates in model A plus education, diet, smoking, and treatment with nonsteroidal anti-inflammatory drugs.
Included covariates in models A and B plus waist circumference, BP, antihypertensive treatment, and levels of triglycerides, HDL cholesterol, fasting glucose, and 2-h glucose.
Included covariates in models A, B, and C plus Si.
Included covariates in models A, B, C, and D plus BMI.
Natural log transformation of Si, AIR, proinsulin-to-insulin ratio, and levels of insulin, leptin, and CRP were used to correct for skewness and kurtosis.