A large prospective cohort study investigates the association between high coffee consumption (≥5 cups per day) and risk of acute myocardial infarction (MI) over 10 years in adults aged 40–60. The crude relative risk (RR) for high versus low coffee consumption (<1 cup per day) is 2.0 (95% CI 1.6–2.5). Investigators suspect smoking may confound the relationship and stratify participants by smoking status. Among current smokers, RR = 1.05 (95% CI 0.85–1.30); among non-smokers, RR = 1.10 (95% CI 0.90–1.35). The Mantel–Haenszel adjusted RR is 1.07 (95% CI 0.92–1.25). Which of the following best explains these findings?