Table 3.
Logistic Regression Models of the Effects of mtDNA Background Haplotype and the DXS8090(166)-DXS1068(258) Nuclear Haplotype in Males Homoplasmic for 11778A→G, 14484T→C, and 14495A→G[Note]
| A. Independent Effect of the Nuclear Haplotype DXS8090(166)-DXS1068(258), with χ2 = 31.42, P < .001 | ||
| Variables in the Equation | P | OR (95% CI) |
| DXS8090(166)-DXS1068(258) | .001 | 35.25 (4.62–268.73) |
| Constant | .001 | .03 |
| B. Independent Effect of the mtDNA Haplotype, with χ2 = 7.53, P = .006 | ||
| Variables in the Equation | P | OR (95% CI) |
| Non mtDNA haplotype J | .007 | 2.89 (1.34–6.19) |
| Constant | .001 | .43 |
| C. Synergistic Interaction between DXS8090(166)-DXS1068(258) and Non-J mtDNA Background, with χ2 = 27.38, P < .001 | ||
| Variables in the Equation | P | OR (95% CI) |
| (Non mtDNA haplogroup J) × DXS8090(166)-DXS1068(258) | .65 | 9,340.46 (0–1.91 × 1021) |
| Constant | .65 | 0 |
| D. Antagonistic Interaction between DXS8090(166)-DXS1068(258) and Non-J mtDNA Background, with χ2 = 2.86, P = .09 | ||
| Variables in the Equation | P | OR (95% CI) |
| (mtDNA haplogroup J) × DXS8090(166)-DXS1068(258) | .152 | 4.73 (.56–39.61) |
| Constant | .031 | .467 |
Note.— The additive model is shown in the text and has the greatest χ2 of 35.68 (P < .0001).