Skip to main content

Table 4 Multivariate linear regression for log10(HOMA-IR)

From: Klotho and fibroblast growth factors 19 and 21 serum concentrations in children and adolescents with normal body weight and obesity and their associations with metabolic parameters

 

Multivariate linear regression for log10(HOMA-IR)

R2 = 0.30, adj. R2 = 0.28

Parameter

95% CI

p-value

Eta2 [%]

Commentary

Intercept

0.278

(0.099 to 0.457)

0.0025

5%

 

Gender - male

0.006

(−0.037 to 0.038)

0.9731

< 0.1%

Associated with 1.4% higher HOMA-IR compared with girls

Tanner - stage III

0.079

(0.018 to 0.139)

0.0112

3.8%

Associated with 19.9% higher HOMA-IR than other stages of puberty

Age [years]

0.009

(−0.003 to 0.021)

0.1509

1.2%

Associated with 2.1% higher HOMA-IR for each year

BMI Z-score

[standard deviations]

0.121

(0.086 to 0.156)

< 0.0001

21.3%

Associated with 32.1% higher HOMA-IR for each unit increase in BMI Z-score

FGF19 concentrations

[100 pg/ml]

−0.037

(−0.067 to − 0.008)

0.0124

3.7%

Associated with 8.2% drop in HOMA-IR for each 100 pg/ml increase in FGF19

  1. The constructed model explains a minor fraction (~ 30%) of HOMA-IR variability among the patients, which demonstrates that individual insulin resistance is highly variable and might depend on factors other than those investigated in this study
  2. HOMA-IR Homeostatic Model Assessment of Insulin Resistance, BMI body mass index, FGF19 fibroblast growth factor19
  3. R2 – proportion of variance in log10 (HOMA-IR) explained by the model
  4. Eta2 – proportion of variance in log10 (HOMA-IR) explained by each factor