Table 2.
Overall percent changes resulting from LFHC versus HFLC diet on metabolic profiles and data on publication bias and their likely effect on the estimates
| A1C | FPG | 2-h glucose | Fasting insulin | 2-h fasting insulin | Total cholesterol | Triglycerides | HDL cholesterol | LDL cholesterol | |
|---|---|---|---|---|---|---|---|---|---|
| Trials (n) | 10 | 22 | 10 | 22 | 9 | 20 | 22 | 20 | 16 |
| Overall percent change | −1.5 | 0.3 | 10.3 | 8.4 | 12.8 | 1.6 | 13.4 | −5.6 | 0.1 |
| 95% CI | −5.3 to 2.3 | −2.8 to 3.4 | 6.7–13.9 | 1.3–15.6 | 5.2–20.4 | −1.3 to 4.5 | 7.1–19.8 | −8.4 to −2.9 | −3.8 to 4.1 |
| P | 0.70 | 0.87 | <0.001 | 0.02 | <0.001 | 0.27 | <0.001 | <0.001 | 0.94 |
| Publication bias | |||||||||
| Begg's | 0.80 | 0.82 | 0.25 | 0.30 | 0.40 | 0.85 | 0.48 | 0.75 | 0.86 |
| Egger's | 0.47 | 0.30 | 0.12 | 0.13 | 0.16 | 0.26 | 0.75 | 0.08 | 0.92 |
| Trim and fill | |||||||||
| Fill* | 7 | ||||||||
| Adjusted† | −7.6 | ||||||||
| 95% CI | −10.2 to −5.0 |
*Studies (n) added by the trim-and-fill method.
†Percent change after adjustment for publication bias by the trim-and-fill method. Begg's, Begg's adjusted rank correlation test; Egger's, Egger's regression asymmetry test.