To interpret divergent findings from these surveys, it is essential to recognize their distinct trajectories and design evolutions. KNHANES, initiated in 1998 as a periodic national survey, underwent a major redesign in 2007, becoming a continuous annual survey with standardized field operations and examination protocols. This transition strengthened its role as the Republic of Korea’s primary instrument for tracking population means over time.

The CHS was launched in 2007 as an annual survey to support local health governance. With a large, stable sample in every municipality, it provides reliable estimates at the si/gun/gu (city/county/district) level, enabling monitoring of regional variation and community-based public health interventions. KYRBS, initiated in 2005 as an annual school-based survey, has served as a core surveillance tool for adolescent health behaviors. It introduced a panel component in 2019, strengthening its capacity to examine longitudinal behavioral patterns and to inform early preventive strategies.

These surveys constitute a maturing national surveillance architecture. Apparent inconsistencies in their findings should be understood not as methodological flaws, but as the result of purpose-built systems operating at different levels of the health system.

KNHANES, in particular, is fundamentally a survey of national means. Examining approximately 10,000 individuals annually and reporting pooled 3-year statistics, it is optimized to track long-term trends in population averages. Because extremes can disproportionately influence the mean, worsening or stagnating national averages are not “noise”; they signal that structural risk factors—linked to labor conditions, demographic aging, food environments, and social norms—remain inadequately addressed. KNHANES data have revealed a troubling paradox: average caloric intake has remained largely unchanged, yet the prevalence of obesity and hypercholesterolemia has risen steadily. High-risk alcohol consumption remains widespread, while vigorous physical activity has declined rather than increased [1].

However, averages are blunt instruments. They obscure heterogeneity, dampen local variation, and may remain stable even when interventions produce meaningful changes for many people in specific regions or contexts.

The CHS, by contrast, is designed to detect such changes. With a sample of approximately 230,000 adults across all municipalities, it emphasizes medians and interregional comparisons, making it more sensitive to changes affecting the “typical” community resident. Community-level interventions—such as recreational infrastructure, more walkable environments, and health center–led programs targeting alcohol consumption and physical activity—may therefore be visible in CHS before they influence national means. The 2024 CHS findings, showing lower median levels of high-risk drinking, higher median physical activity, and reduced regional gaps, suggest that such interventions may be beginning to take effect. From a policy perspective, this is encouraging, implying that sustained and adequately resourced local action can influence population health behaviors [2].

Yet this apparent progress must be interpreted with caution. A reduction in disparities does not necessarily indicate improved population health. Regional gaps can narrow because disadvantaged areas improve, but also because previously advantaged areas deteriorate. Mean and median based indicators provide little information about changes in the distribution’s spread or about outcomes among the most vulnerable groups. Without examining variance, interquartile ranges, and distributional tails, it is impossible to distinguish genuine equity gains from downward convergence [3].

This tension between national averages and regional medians reflects a well-recognized phenomenon in population health. Geoffrey Rose famously argued that most cases of disease arise not from a small group at extreme risk, but from the large number of people exposed to modest risk [4]. Sustainable improvements therefore require strategies that shift the entire distribution of risk. Such shifts occur gradually and are often first detectable as changes in medians and dispersion, rather than as immediate improvements in the mean [5].

Viewed through this lens, the divergent signals from KNHANES and CHS are not contradictory. They suggest a transitional phase: early community-level distributional changes that are not yet sufficient to alter national averages. CHS may be capturing the initial stages of a population-wide shift driven by local interventions, while KNHANES indicates that these shifts have not reached the scale or intensity required to overcome entrenched structural determinants.

The implications for policy and research are substantial. The central question is no longer which survey is correct, but how insights from different surveillance systems can be integrated. Reliance on means, medians, or simple maximum–minimum gaps alone is inadequate. Instead, analyses must incorporate measures of variance, interquartile ranges, and gradients across social determinants of health, including income, education, occupation, and housing conditions [6].

Moreover, translating community-level improvements into national gains requires action beyond the health sector. Improvements in physical activity and alcohol consumption cannot be sustained through health programs alone; they depend on labor policies shaping working hours, urban planning that enables active transport, housing policies influencing neighborhood environments, and regulatory approaches addressing alcohol availability and marketing. Medical cost subsidies and intensified treatment of high-risk individuals, while important, are insufficient to shift population averages. Equally important, success cannot be declared solely on the basis of improved medians or reduced regional gaps. If disease prevalence continues to rise, convergence of indicators may mask a worsening overall burden. Conversely, improvements in national averages that leave inequalities unchanged are unlikely to be durable. Population health improvement and health equity are inseparable objectives.

Therefore, policymakers, researchers, and practitioners face a challenge: are we content to monitor averages, or are we prepared to confront the full distribution of health and its social determinants?

Citations to this article as recorded by