Table 4. Normal versus abnormal laboratory findings.
| Laboratory test, with cutoff for normal range | US Cases % abnormal | US Controls % abnormal | p-value (cases vs controls) | Chilean Cases % abnormal | Chilean controls % abnormal | p-value (cases vs controls) |
|---|---|---|---|---|---|---|
| TBA (Clinical) (< 10 umol/L) | 81% elevated | — | — | — | ||
| TBIL (Clinical) (≤ 1.0 mg/dL) | 13% elevated | — | 14% | — | ||
| TBIL (Research) (≤ 1.0 mg/dL) | 3% elevated | 0% | 0.4967 | 3% | 0% | 0.2464 |
| DBIL (Clinical) (≤ 0.3 mg/dL) | 18% elevated | — | (89%) | — | ||
| DBIL (Research) (≤ 0.3 mg/dL) | 22% elevated | 1% | <0.0001 | 18% | 6% | 0.0095 |
| ALT (Clinical) (≤ 35 U/L) | 67% elevated | — | 34% | — | ||
| ALT (Research) (≤ 35 U/L) | 60% elevated | 3% | <0.0001 | 21% | 3% | <0.0001 |
| AST (Clinical) (≤ 41 U/L) | 53% elevated | — | 29% | — | ||
| AST (Research) (≤ 41 U/L) | 53% elevated | 7% | <0.0001 | 22% | 4% | <0.0001 |
| gGT (Clinical) (5–36 U/L) | (0%) | — | 16% | — |
Table 4 notes. For some biochemical parameters, results from medical records were available from a subset of controls; those data are not presented, as there is likely bias in selection of which control women would undergo testing for clinical purposes. Therefore, comparisons between cases and controls are performed only for data from the study samples. Normal range definitions: Since study samples were assayed for a number of parameters at the SFGH clinical laboratory, we used normal ranges in use at SFGH for the analyses presented in Table 4, except for γGT, for which the normal range at the Chilean site is indicated, as that is the only site where appreciable numbers of samples were assayed for γGT in a clinical laboratory. The normal ranges used are the standards for nonpregnant women; as designation and application of pregnancy-specific ranges for these tests is inconsistent between laboratories, we did not use them. To note, for results of biochemical tests from medical records, normal ranges sometimes differ modestly between study sites; while using such site-specific normal ranges for reporting of the ‘clinical’ data in Table 4 could change percentages somewhat, the general trends remain, so the same ‘consensus’ ranges were used for data from all sites, to avoid complexities in presentation. For ‘research’ TBA and γGT, normal ranges are not indicated, and results are only presented in Table 3, as the tests performed were not clinical laboratory assays with well-defined normal ranges. Percentages in parentheses in Table 4 are of questionable meaning, due to small sample size. P-values are from Fisher’s exact test.