Fifty Years of Data Show Nasopharyngeal Cancer Falling in a Chinese Hotspot

In one of the world’s most stubborn cancer hotspots, the tide is finally turning. A new analysis of fifty years of registry data from Sihui City in southern China’s Guangdong Province shows that nasopharyngeal carcinoma, a cancer of the throat long endemic to the region, has entered a sustained decline. The study, published in BMC Cancer, tracked incidence from 1971 through 2020 and projects that rates will keep falling through 2040, offering one of the clearest long-term pictures yet of how a traditionally high-burden cancer responds to decades of economic and environmental change.

Nasopharyngeal carcinoma arises in the nasopharynx, the passage connecting the back of the nose to the throat. It is unusual among cancers in that its distribution is intensely geographic: rates in parts of southern China, Hong Kong, and Southeast Asia have historically been tens of times higher than in Europe or North America. The disease is closely associated with Epstein-Barr virus infection, along with dietary and environmental risk factors such as salted fish consumption and occupational exposures. Sihui, a city in the heart of the endemic zone, has maintained one of the longest-running population-based cancer registries in China, making it an ideal laboratory for studying the disease’s trajectory.

The research team, led by Yong-Rui Ling of Guangdong Medical University and Su-Mei Cao of Sun Yat-sen University Cancer Center, drew on registry records covering every diagnosed case in Sihui over five decades. They calculated age-standardized incidence rates using the world standard population, which adjusts for changes in the age structure of the population so that trends reflect genuine shifts in cancer risk rather than demographic drift. Across the full 1971 to 2020 period, the overall age-standardized incidence rate was 31.19 cases per 100,000 person-years in males and 13.26 per 100,000 in females, a male-to-female ratio of roughly 2.4 to 1 that is characteristic of the disease.

To dissect the temporal patterns, the researchers applied joinpoint regression, a statistical technique that fits a series of straight lines to incidence data on a log scale and identifies the years at which the trend changes direction. Each segment yields an estimated annual percent change, and the average of these segments across the whole period gives the average annual percent change. This approach allowed the team to pinpoint exactly when the decline began rather than relying on visual inspection of curves. The male incidence rate, they found, declined continuously and gradually throughout the entire fifty-year span, while female rates held steady in the early decades before dropping markedly between 2009 and 2020.

The heart of the study is an age-period-cohort analysis, a modeling framework designed to separate three intertwined influences on cancer risk that are otherwise impossible to disentangle from raw incidence data. Age effects capture the rise in risk as people grow older. Period effects reflect influences operating at a particular calendar time, such as a screening campaign or an environmental change affecting everyone alive at once. Cohort effects capture the lifetime risk carried by people born in the same era, shaped by exposures early in life such as childhood diet or infection patterns. Because age, period, and cohort are mathematically linked, with period equal to cohort plus age, the model relies on constraints and Poisson regression to estimate each component’s independent contribution.

The results of this decomposition are striking. The relative risk of nasopharyngeal carcinoma rose with advancing age, peaking in the 50 to 59-year age group and declining thereafter, a pattern suggesting that the oldest individuals may either be less susceptible or represent survivors of a generation with different exposure histories. The period effect, by contrast, was relatively flat for both sexes, indicating that no single calendar moment produced a sharp change in risk across the whole population. Instead, the cohort effect told the decisive story: each successive birth cohort, for both males and females, carried a progressively lower risk of the disease than the one before it.

This cohort pattern is epidemiologically meaningful because it implies that the drivers of the cancer operate early in life and have been diminishing generation by generation. The authors presume that reduced exposure to environmental risk factors resulting from economic development is the most plausible reason. Over the decades covered by the study, Sihui and the wider Guangdong region underwent profound transformations: diets diversified away from traditional preserved foods such as salted fish, housing and occupational conditions improved, and living standards rose steadily. If childhood exposures to these factors seeded the cancer risk carried through life, then each cohort born into a cleaner environment should indeed show lower incidence, exactly what the model reveals.

Looking forward, the team used age-period-cohort models implemented in the NORDPRED package, a projection tool developed for cancer registry forecasting, to estimate incidence through 2040. NORDPRED applies drift terms and cohort-based projections to extend observed trends while accounting for the aging of the population. The projections indicate that by the period 2036 to 2040, the incidence rate will fall to 14.52 per 100,000 in males and 4.65 per 100,000 in females, less than half the fifty-year average for males and a dramatic reduction for females. The authors conclude that the decline, which remained relatively stable before 2010 and decreased significantly thereafter, is expected to continue over the next twenty years.

The findings carry implications well beyond Sihui. For other endemic regions in southern China and Southeast Asia, the study suggests that the cancer burden will ease as economic development reduces early-life exposures, without requiring a single dramatic intervention. At the same time, the persistence of elevated rates for decades to come, even in decline, underscores the continued importance of early detection programs in the region. Sihui has long served as a site for screening studies of Epstein-Barr virus markers, and the new projections provide health planners with quantitative targets for how screening and prevention resources should be allocated as the demographic profile of risk shifts.

For cancer epidemiology more broadly, the study demonstrates the power of half a century of continuous registry data combined with rigorous statistical decomposition. Fifty years ago, few registries in low- and middle-income settings could sustain the decades of dedicated data collection, registration, and quality control that the Sihui Cancer Registry staff provided, an effort the authors explicitly credit for making the analysis possible. As endemic regions worldwide undergo similar transitions, the Sihui record stands as a template for measuring how a virally linked, environmentally shaped cancer yields, generation by generation, to a changing world.

Subject of Research: Fifty-year age-period-cohort analysis of nasopharyngeal carcinoma incidence trends and projections in Sihui City, southern China

Article Title: Fifty-year trends and future projections of nasopharyngeal carcinoma incidence in Sihui City, southern China: an age-period-cohort analysis (1971–2020)

Article References: Ling, Y.-R., Li, T., Ling, W., Wu, Z.-C., Wen, X.-J., Xie, S. H., Chen, H., Ye, J.-P., Cao, L., Cao, S.-M., & Du, J.-L. (2026). Fifty-year trends and future projections of nasopharyngeal carcinoma incidence in Sihui City, southern China: an age-period-cohort analysis (1971–2020). BMC Cancer. https://doi.org/10.1186/s12885-026-17195-5

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17195-5

Keywords: nasopharyngeal carcinoma, cancer epidemiology, age-period-cohort analysis, Sihui City, southern China, incidence trends, Epstein-Barr virus, cancer registry, joinpoint regression, cancer projection, BMC Cancer, Guangdong