1. You mention inverse probability weighting to adjust for selection bias due to missing tissue, but this assumes missingness is at random conditional on observed covariates. Given that tissue collection depends on participants undergoing biopsy/resection at various hospitals (often with unknown retrieval success rates), how do you rule out non‑ignorable missingness? Without validation, the generally lower degree of selection bias claim remains unsubstantiated. 2. All three flagship cohorts (NHS, NHS II, HPFS) consist predominantly of non‑Hispanic White health professionals. The PCIBM’s unique etiological insights may not apply to other racial/ethnic groups or socioeconomic strata. Table 1 lists no diverse population biobanks. How can the field justify decades of investment in PCIBM before establishing incident‑tumor biobanks in underrepresented populations, given the known disparities in cancer incidence and outcomes?
