Application of the age-period-cohort model in tuberculosis

The age-period-cohort (APC) model, despite facing challenges like unidentifiability due to collinearity among its variables, offers a valuable lens for examining long-term trends in diseases such as tuberculosis (TB). While it has been widely used in analyzing diseases like influenza and HIV/AIDS, its full potential in TB research has not been extensively explored. This study aims to fill that gap by reviewing literature from multiple databases, summarizing how the APC model has been applied in TB research, and highlighting its capabilities and limitations in tracking disease patterns across different time dimensions.

The APC model works by dissecting the effects of age, period, and cohort—three intertwined but distinct time-related factors. “Age” captures individual physiological and social changes over life stages, “period” reflects events and trends at a specific point in time (like pandemics or policy changes), and “cohort” encapsulates the shared experiences of people born around the same time. This framework allows researchers to interpret fluctuations in disease incidence or mortality by isolating and analyzing these temporal components independently, giving more nuanced insight into epidemic trends.

Mathematically, the APC model is built on a generalized linear model framework where incidence or mortality is predicted using age, period, and cohort as independent variables. The model typically assumes a Poisson distribution for observed events and calculates relative risk (RR), net drift (average annual change), and local drift (age-specific trends). However, a major issue is the linear dependency among the three factors (period = age + cohort), which creates a collinearity problem and makes it difficult to uniquely estimate their effects. Over time, various methodological advances—like the intrinsic estimator (IE) and estimable function approaches—have been developed to overcome this limitation.

Compared to traditional trend analysis methods, the APC model excels at addressing confounding among variables and offers a more robust framework for analyzing disease dynamics across age, period, and cohort. This makes it particularly valuable in public health and social science for studying phenomena such as aging, economic shifts, or life-threatening diseases. However, practical issues like overlapping cohorts and interpretability of parameters persist. Despite these, improvements in modeling techniques and computational tools have made APC analyses more accessible and accurate, with notable applications in TB research emerging in recent years.

In conclusion, the APC model stands out as a powerful statistical tool to monitor, predict, and understand TB epidemics by parsing out the temporal components that influence disease patterns. As data collection and modeling strategies continue to improve, the model’s use in identifying high-risk groups and informing targeted interventions will become even more critical. Nonetheless, its results should be interpreted with caution and combined with contextual and qualitative research to guide policy decisions and public health responses effectively. 

Source: Luo, D., Wang, F., Chen, S., Zhang, Y., Wang, W., Wu, Q., Ling, Y., Zhou, Y., Li, Y., Liu, K. and Chen, B., 2025. Application of the age-period-cohort model in tuberculosis. Frontiers in Public Health, 13, p.1486946.