Trends and Age-Period-Cohort Effects on the Prevalence, Incidence and Mortality of Hepatocellular Carcinoma from 2008 to 2017 in Tianjin, China

Chengyu Liu, Jing Wu, Zheng ChangInternational Journal of Environmental Research and Public Health 2021 · DOI

TL;DR

This study analyzes trends in hepatocellular carcinoma (HCC) prevalence, incidence, and mortality in Tianjin, China, from 2008-2017. It reveals a rising prevalence, stable incidence, and decreasing mortality, driven by significant age and cohort effects, with implications for targeted screening and treatment strategies.

What Problem Does This Paper Address?

Hepatocellular carcinoma (HCC) is a major global health concern, with China bearing a disproportionately high burden of cases and deaths. Despite this, detailed epidemiological data for HCC in mainland China, particularly concerning temporal trends and the influence of age, period, and birth cohort, has been scarce. Understanding these dynamics is crucial for effective public health interventions, resource allocation, and predicting future disease patterns. This paper aims to fill this knowledge gap by examining HCC epidemiology in Tianjin, a major northern Chinese city, over a decade.

Methods

The researchers employed two primary statistical methods:

  • Joinpoint Regression Analysis: This technique was used to identify statistically significant changes in the trends of age-standardized prevalence, incidence, and mortality rates over the study period. It allows for the detection of "joinpoints" where the rate of change shifts, providing a more nuanced understanding of temporal trends than a simple linear model.
  • Age-Period-Cohort (APC) Analysis: This sophisticated modeling approach was used to disentangle the independent effects of age (the individual's age), period (the calendar year of diagnosis), and birth cohort (the year of birth) on HCC rates. By using an intrinsic estimator (IE) within a Poisson log-linear model, the study aimed to overcome the inherent linear dependency between these three factors and provide rate ratios for each effect.

Datasets

The study utilized data from the Urban Employee Basic Medical Insurance (UEBMI) claims database in Tianjin, China, covering the period from 2003 to 2017. This database includes information on inpatient, outpatient, and pharmacy claims for approximately 5.95 million individuals in Tianjin. A 30% random sample of these enrollees was used for the analytical sample. The data was de-identified and did not involve direct patient interaction, thus waiving the need for explicit ethical approval.

Key Insights and Ideas

  • Prevalence is Climbing, Incidence is Stable, Mortality is Falling: While the rate of new HCC cases (incidence) remained relatively constant, the overall number of people living with HCC (prevalence) significantly increased. Encouragingly, the death rate from HCC (mortality) showed a notable decrease.
  • Age is a Dominant Factor: The study strongly confirms that age is a primary driver of HCC risk. Prevalence and incidence rates rise sharply with age, peaking in older age groups, while mortality also increases significantly with age.
  • Birth Cohort Matters Significantly: Distinct birth cohorts exhibit different risk profiles. Earlier birth cohorts (e.g., those born in the 1920s) had substantially higher prevalence and incidence rates compared to more recent cohorts (e.g., born in the 1990s). This suggests that factors experienced during specific historical periods or early life exposures have a lasting impact.
  • Period Effects are Less Pronounced: While there was a general upward trend in prevalence over the study period, the influence of the "period" (calendar year) on HCC rates was less pronounced compared to age and cohort effects. This suggests that broader societal or environmental changes over the decade had a more subtle impact than individual aging or generational exposures.
  • Gender Disparities Persist: Men consistently showed approximately twice the prevalence, incidence, and mortality rates of women, likely attributable to factors like higher rates of alcohol consumption in men.

Key Results

  • Prevalence: Increased annually by 5.13% (AAPC) from 20.12/100,000 in 2008 to 30.49/100,000 in 2017.
  • Incidence: Remained largely unchanged, with an AAPC of -0.82% from 13.91/100,000 in 2008 to 14.09/100,000 in 2017.
  • Mortality: Decreased annually by 1.80% (AAPC) from 8.18/100,000 in 2008 to 7.34/100,000 in 2017.
  • Age Effects: Prevalence and incidence peaked around age 60 and decreased after age 70. Mortality increased from age 40 to 85.
  • Cohort Effects: Prevalence and incidence decreased linearly from the 1925 cohort to the 1990 cohort. Mortality also decreased across cohorts, with some fluctuations.

Limitations and Future Work

The authors acknowledge several limitations:

  • "Prevalent-treatment cases": The data relies on claims, meaning only individuals who sought treatment for HCC were captured. This might underestimate the true prevalence and incidence, though it likely doesn't significantly affect trend analyses.
  • UEBMI population: The study population consists of employees and retirees, excluding individuals under 18 and those with other insurance types. This could lead to an underestimation of overall prevalence and incidence, and potentially an overestimation of mortality if younger individuals have lower mortality rates.
  • Geographic scope: The study is limited to Tianjin, and while it's a major northern city, its findings may not be universally generalizable to all of China.
  • Limited timeframe: The ten-year study period is relatively short. Longer-term data would be beneficial for observing more pronounced period effects.

Future work could involve extending the analysis to longer timeframes to capture the impact of new interventions (like Sorafenib and Lenvatinib) and expanding the study to include broader insurance populations to provide a more comprehensive picture of HCC epidemiology in China.

Summary updated

Abstract

<i>Objectives:</i> China is the country most afflicted by hepatocellular carcinoma in the world. However, little is known about the epidemiology of hepatocellular carcinoma in China. This study aimed to examine the trends of the prevalence, incidence, and mortality of hepatocellular carcinoma in China, and to investigate the effects of age, period, and birth cohort on the epidemiological trend. <i>Methods:</i> The data were obtained from the Urban Employee Basic Medical Insurance claims database (2003-2017) in Tianjin, China, which covers 5.95 million individuals. The average annual percentage change of the prevalence, incidence, and mortality were accessed using joinpoint regression. Age-period-cohort models were produced to quantify the effects of age, period, and cohort. <i>Results:</i> The hepatocellular carcinoma prevalence rate increased by 5.13% annually from 20.12/100,000 in 2008 to 30.49/100,000 in 2017, and the incidence rate was almost unchanged, from 13.91/100,000 in 2008 to 14.09/100,000 in 2017, but mortality decreased by 1.80% annually from 8.18/100,000 in 2008 to 7.34/100,000 in 2017. The age-period-cohort analysis revealed that the prevalence rate was remarkably increased from age 25, peaked in age 60, and decreased at age 70 and over. In the period index, the prevalence rate increased gradually from 2008 to 2016, and decreased a little in 2017. In the cohort index, the prevalence rate decreased approximately linearly from the 1925 cohort to the 1990 cohort. The result for the incidence was similar to the prevalence. The mortality rate increased approximately linearly from age 45 to 85, decreased from the 1925 cohort to the 1990 cohort, but it changed a little with the change of period. <i>Conclusions:</i> The findings of this study could inform the necessity of conducting earlier screening for high-risk individuals and improving the treatment of hepatocellular carcinoma, which may also help to predict future changes in hepatocellular carcinoma epidemiology.
BibTeX
@article{liu2021trends,
  author = {Chengyu Liu and Jing Wu and Zheng Chang},
  title = {{Trends and Age-Period-Cohort Effects on the Prevalence, Incidence and Mortality of Hepatocellular Carcinoma from 2008 to 2017 in Tianjin, China}},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2021},
  doi = {10.3390/ijerph18116034},
}
Figure