Serum trace-element categories and reported first-degree family history of malignancy: an exploratory analysis of the SPES cohort

Background We examined whether contemporary serum trace-element categories were associated with reported first-degree family history of malignancy, rather than a relative’s historical exposure or the participant’s future cancer risk. Methods This secondary exploratory cross-sectional analysis used archived data from 4,188 adults in SPES (Studio di Esposizione nella Popolazione Suscettibile), a population-based human-biomonitoring programme conducted in Campania, Italy, in 2016–2017. Relative-level reports were classified using documented rules and oncologist adjudication; a second registered physician independently reviewed a random 760-record subsample in a post hoc assessment during revision. Principal mixed-effects logistic models used observed median categories for 18 channels and quantified-versus-below-limit-of-quantification status for Be-9 and Pb-208, adjusting for age modelled with a restricted cubic spline, sex, participant smoking and SPES cluster. The 20 channels represented 19 elements. The original analysis inventory contained 258 hypothesis slots across nine separately corrected families; reviewer-requested sensitivities were reported separately. Results Family history was positive in 1,040 of 4,185 classifiable participants (24.9%). Principal odds ratios ranged from 0.832 to 1.109 (minimum q = 0.730). Original upper-quartile and conditional malignancy-category findings did not survive their respective corrections. Category models compared a target malignancy with other malignancies among composite-positive histories, with incomplete site ascertainment. Additional stratum-adjusted, expanded-adjustment, quantified-range, broader-comparator and restrictive-definition principal/category families yielded no FDR-supported association. Cobalt met the within-set BH criterion only in a selected three-contrast expanded-Q4 sensitivity ( q = 0.019), not a panel-wide analysis. Independent binary agreement was 99.87% (757/758; κ = 0.9969); two indeterminate ratings were reported separately. Conclusion The tested serum categories did not identify a broad reported first-degree family-history signature. Nominal secondary patterns are hypotheses for independent replication, not validated exposure–cancer relationships. Classification reproducibility does not establish diagnostic accuracy or registry confirmation.