Figure 1.

Sample selection criteria in a study of childhood mortality in family members of men with poor semen quality. We identified 26 147 men with semen analysis (SA) performed during our study period. We excluded 1424 men who did not link to another record in the Utah Population Database (UPDB), 434 with inadequate follow-up, 449 with cancer prior to SA, and 41 with an analysis prior to age 20. There were 10 910 men without complete information on their parents, and therefore a total of 12 889 men were available with first-degree relative (FDR) information. This cohort of men was used to study the childhood mortality in FDR. Another 4857 men did not have full information on their grandparents, which left 8032 men with complete second-degree relative (SDR) information available for analysis.