The study drew on incidence data reported through the infectious disease reporting system of the China Disease Prevention and Control Information System, covering the years 2005 through 2024 in Shandong Province. Across the entire twenty-year window, the average annual reported incidence of brucellosis stood at 1.88 cases per 100,000 population. That figure, modest in absolute terms, conceals a dramatic story of rise and fall. Using joinpoint regression, a technique that identifies statistically meaningful inflection points in trend data, the researchers calculated the annual percentage change (APC) in incidence across distinct segments of the epidemic, and the average annual percentage change (AAPC) as a summary measure of the overall trajectory.

The joinpoint analysis revealed four distinct phases. From 2005 to 2009, reported incidence declined slightly, at an annual rate of 0.7 percent. Then came a startling reversal: between 2009 and 2015, incidence surged at an average annual rate of 75.3 percent, one of the steepest sustained climbs documented for this disease in a provincial Chinese population. The tide turned again from 2015 to 2019, when incidence fell at 12.4 percent per year, before resuming a slower upward march of 6.7 percent annually from 2019 to 2024. This alternating pattern of decline, explosive growth, retreat, and renewed increase suggests that the drivers of brucellosis transmission in Shandong have shifted repeatedly, responding to changes in animal husbandry practices, surveillance intensity, and control measures.

To move beyond simple trend description, the research team applied an age-period-cohort (APC) model, a statistical framework designed to separate three effects that are otherwise mathematically entangled. Age effects capture how susceptibility or exposure changes across the human lifespan. Period effects reflect influences that operate on the entire population at a given moment, such as outbreaks, policy interventions, or diagnostic changes. Cohort effects track the shared experiences of people born in the same era, whose lifelong exposure patterns differ from those of earlier or later generations. Because the three dimensions are linearly dependent, APC modeling requires careful statistical constraints, and the authors report that their model fit the data strongly, with the period effect alone yielding a chi-squared statistic of 795.63 and a p-value below 0.001.

The age effect emerged as the most clinically consequential finding. The risk of brucellosis increased rapidly with advancing age group, and the risk coefficient reached its highest point among people over 80 years old. This pattern is notable because brucellosis has traditionally been framed as an occupational disease of working-age farmers, herders, and slaughterhouse workers. An age gradient that climbs all the way into the oldest age brackets suggests that in Shandong, exposure pathways may extend well beyond occupational contact, potentially reflecting household food handling, consumption of unpasteurized dairy products, or the cumulative effect of repeated low-level exposures over a lifetime. It also flags elderly residents, whose immune function typically wanes with age, as a population deserving targeted protection.

The period effect told an equally vivid story. Incidence risk rose slowly at first, then climbed sharply before declining, and the model placed the peak in 2017, when the relative risk stood at 2.779 (95 percent confidence interval: 2.568 to 3.007) compared with the reference period. The lowest point fell in 2007, with a relative risk of just 0.149 (95 percent confidence interval: 0.123 to 0.181). In other words, the risk of contracting brucellosis in Shandong at the height of the epidemic was nearly nineteen times higher than at its low point a decade earlier. That the peak in 2017 came after the joinpoint analysis showed incidence beginning to fall in 2015 underscores a subtlety of APC modeling: period effects capture year-specific risk net of age and cohort composition, and they can peak at moments that raw trend lines only hint at.

The cohort effect completed the picture. The analysis found that incidence risk increased with successive birth cohorts, meaning that people born in more recent years carried a higher lifetime risk of brucellosis than those born earlier, a relationship that was itself statistically robust (chi-squared = 126.74, p < 0.001). This generational gradient is epidemiologically significant. It implies that the conditions giving rise to brucellosis exposure in Shandong have intensified over the decades in which younger cohorts grew up, rather than fading as the province modernized. Rapid expansion of sheep, goat, and cattle farming, growing market circulation of live animals and animal products, and changing patterns of rural livelihood could all contribute to such a cohort pattern, and the finding gives weight to the argument that structural changes in the livestock economy, not merely short-term outbreaks, have driven the province’s epidemic.

Brucellosis itself remains a formidable public health challenge worldwide. Caused by bacteria of the genus Brucella, most commonly Brucella melitensis in regions where sheep and goats are raised, the disease infects humans primarily through direct contact with infected animals or their secretions, or through consumption of contaminated dairy. Its clinical presentation, with undulating fevers, arthralgia, fatigue, and in chronic cases osteoarticular complications, makes it easy to miss and difficult to eradicate once established in an animal reservoir. Because the bacteria persist in livestock populations, human incidence is ultimately a downstream reflection of veterinary control. China has experienced a nationwide resurgence of human brucellosis since the 1990s, and Shandong, though historically a lower-incidence province compared with the pastoral regions of the north, has not been spared the broader epidemic wave that the 2009 to 2015 surge so clearly documents.

The authors conclude that from 2005 to 2024, reported brucellosis incidence in Shandong Province displayed a significant age-period-cohort effect, with an overall trajectory that first increased and then decreased. Risk rose rapidly with age, climbed slowly at first and then sharply before declining across calendar periods, and increased steadily across birth cohorts. On the strength of these patterns, they recommend that health education be targeted toward the groups the model identifies as most vulnerable, particularly the elderly, while surveillance and control efforts remain attentive to the renewed upward drift in incidence observed since 2019. The renewed increase, though modest at 6.7 percent per year, arrives at a moment when the population at risk is aging, which the age effect suggests could amplify its impact.

For epidemiologists, the study is also a demonstration of method. Joinpoint regression and APC modeling answer different questions, and used together they convert a raw surveillance time series into a layered account of an epidemic: when it turned, how sharply, and which dimensions of human experience carried the risk. As brucellosis continues to circulate in livestock systems across much of the world, the Shandong experience suggests that the tools to read its patterns already exist in routine surveillance data. What remains is the harder work of translating statistical risk coefficients into interventions that reach the aging farmers, rural households, and food consumers whose exposures, accumulated across decades and generations, the model has now made visible.

Subject of Research: Epidemiological trends and age-period-cohort analysis of human brucellosis incidence in Shandong Province, China, from 2005 to 2024

Article Title: Trend of brucellosis incidence and age-period-cohort model analysis in Shandong Province from 2005 to 2024

Article References: Xiaolin, Y., Maowen, L., Lu, L., Ming, F., Xiaolin, L., Yan, L., Shujun, D., & Zengqiang, K. (2026). Trend of brucellosis incidence and age-period-cohort model analysis in Shandong Province from 2005 to 2024. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14307-1

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14307-1

Keywords: brucellosis, epidemiology, age-period-cohort model, joinpoint regression, Shandong Province, China, zoonotic disease, public health surveillance, incidence trends, infectious disease, livestock, elderly health