Nutrition/Obesity
Causal effect of ultra-processed food intake on incident obstructive sleep apnoea: TMLE and AIPW analyses in UK Biobank Yohannes Adama Melaku* Yohannes Adama Melaku Melaku Melaku Melaku Melaku Flinders University
Background: Ultra-processed foods (UPF) are increasingly consumed and have been linked to obesity and cardiometabolic disease, but evidence on their causal role in obstructive sleep apnoea (OSA) is limited. We estimated the causal effect of UPF intake on incident OSA in UK Biobank.
Methods: We analysed 114,990 UK Biobank participants with dietary intake derived from the average of four 24-hour dietary recalls. UPF exposure was defined using the NOVA classification as the percentage contribution of UPF to total food intake (grams), evaluated at thresholds of ≥25%, ≥50%, and ≥75%. Incident OSA was ascertained via linkage to hospital admissions and primary care records. Causal risk ratios (RRs) were estimated using targeted maximum likelihood estimation (TMLE) and augmented inverse probability weighting (AIPW), adjusting for baseline sociodemographic, behavioural, and clinical confounders.
Results: During follow-up, 1,438 incident OSA cases were identified. Higher UPF consumption was associated with increased OSA risk in a graded manner. TMLE estimates were: UPF ≥25% RR 1.24 (95% CI 1.10–1.41), UPF ≥50% RR 1.54 (1.37–1.74), and UPF ≥75% RR 1.99 (1.57–2.53). AIPW results were highly consistent: UPF ≥25% RR 1.26 (1.12–1.40), UPF ≥50% RR 1.54 (1.38–1.73), and UPF ≥75% RR 1.96 (1.60–2.39).
Conclusions: Using two doubly robust causal estimators and repeated dietary assessment, higher UPF intake was associated with a dose–response increase in incident OSA risk. These findings support UPF reduction as a potential target for OSA prevention and motivate intervention and implementation research.