Age Patterns, Cohort-proxy Trends, and Scenario-based U.S. Coffee Participation through 2040

Abstract

Aims: To quantify age-related differences in U.S. coffee consumption and assess how demographic ageing and cohort-compatible persistence patterns may shape adult coffee participation and espresso-based beverage (EBB) incidence through 2040. Study design: Repeated cross-sectional secondary-data analysis with demographic scenario projection. Place and Duration of Study: United States; National Health and Nutrition Examination Survey cycles from 2003-2004 through August 2021-August 2023, supplemented by 2026 National Coffee Data Trends statistics and U.S. Census population projections through 2040. Methodology: Survey-weighted point estimates of coffee participation, intake volume, and caffeine intake were calculated for adults aged 18-24, 25-39, 40-59, and 60+ years. A population-weighted cohort proxy was evaluated using three historical anchors. Future participation was compared under a static age-specific-rate scenario and a 2023-anchored cohort-proxy sensitivity scenario. The dynamic scenario used the median annualised endpoint deviation across the three anchors, with zero modelled drift at the 2023 base year. Formal sampling uncertainty was not propagated into the projections. Results: In 2021-2023, coffee participation increased from 22.4% at ages 18-24 to 68.7% at ages 60+, whereas mean intake among consumers peaked at ages 40-59 (546 g/day; 202 mg caffeine/day) and declined at ages 60+ (521 g/day; 179 mg/day). From 2003-2004 to 2021-2023, participation rose 56.7% for ages 18-24 and 25.8% for ages 25-39 while volume and caffeine intake generally declined. By 2040, the static and 2023-anchored scenarios yielded approximately 167.6 and 167.8 million adult coffee participants, respectively; corresponding EBB estimates were 12.1 and 15.2 million, a difference of about 26%. Conclusion: Demographic ageing alone raises projected coffee participation, while cohort-compatible drift has little effect on the headline total under the anchored specification but a larger effect on modelled beverage-format composition. Because NHANES uses a repeated cross-sectional design, the 2021-2023 dietary cycle involved a change in interview mode, and formal sampling uncertainty was not propagated, all long-horizon estimates are descriptive scenario outputs rather than causal forecasts.