Documented non-tobacco nicotine dependence coding and incident age-related macular degeneration in a large multicenter electronic health record cohort

Abstract Background Cigarette smoking is a well-established risk factor for age-related macular degeneration (AMD), but the relationship between nicotine-related exposures recorded in electronic health records and AMD remains unclear. We evaluated whether documented non-tobacco nicotine dependence (NTND) coding was associated with incident AMD. Methods In this retrospective cohort study using the TriNetX Global Collaborative Network, adults aged ≥ 50 years with documented NTND were compared with individuals without documented nicotine dependence after exclusion of patients with recorded tobacco use. Propensity score matching (1:1) was performed to balance demographic characteristics, systemic comorbidities, ophthalmic conditions, healthcare utilization, and medication exposures. Primary outcomes were incident non-neovascular AMD and neovascular AMD over up to 10 years of follow-up. Sensitivity analyses included restriction to patients who survived the study period, five- and seven-year follow-up, and a landmark analysis indexed at the second qualifying event. Results After matching, 37,411 patients were included in each cohort. Documented NTND coding was associated with higher incidence of non-neovascular AMD (non-nvAMD; 1.74% vs. 0.92%; hazard ratio [HR] 1.50, 95% confidence interval [CI] 1.31–1.71; E-value 2.37) and neovascular AMD (nvAMD; 0.78% vs. 0.36%; HR 1.69, 95% CI 1.37–2.07; E-value 2.77). NvAMD remained significantly associated in the survivor-restricted, five-year, seven-year, and landmark analyses, whereas non-nvAMD remained significant only in the survivor-restricted analysis. The NTND cohort also demonstrated substantially higher risks of chronic obstructive pulmonary disease (HR 5.65, 95% CI 5.14–6.20; E-value 10.78) and lung cancer (HR 5.74, 95% CI 4.77–6.91; E-value 10.96). Conclusions Documented NTND coding was associated with increased AMD incidence in the primary analysis, with a more consistent association for nvAMD across sensitivity analyses. However, the exposure definition likely includes a heterogeneous population and may reflect a combination of nicotine-related exposures and incompletely captured prior cigarette smoking. These findings should therefore be considered hypothesis-generating and underscore the challenges of studying nicotine-related exposures using standardized clinical data.