Changing Epidemiology of TB in Shandong, China Driven by Demographic Changes

Abstract

Tuberculosis (TB) incidence has been in steady decline in China over the last few decades. However, ongoing demographic transition, fueled by aging, and massive internal migration could have important implications for TB control in the future. We collated data on TB notification, demography, and drug resistance between 2004 and 2017 across seven cities in Shandong, the second most populous province in China. Using these data, and age-period-cohort models, we (i) quantified heterogeneities in TB incidence across cities, by age, sex, resident status, and occupation and (ii) projected future trends in TB incidence, including drug-resistant TB (DR-TB). Between 2006 and 2017, we observed (i) substantial variability in the rates of annual change in TB incidence across cities, from -4.84 to 1.52%; (ii) heterogeneities in the increments in the proportion of patients over 60 among reported TB cases differs from 2 to 13%, and from 0 to 17% for women; (iii) huge differences across cities in the annual growths in TB notification rates among migrant population between 2007 and 2017, from 2.81 cases per 100K migrants per year in Jinan to 22.11 cases per 100K migrants per year in Liaocheng, with drastically increasing burden of TB cases from farmers; and (iv) moderate and stable increase in the notification rates of DR-TB in the province. All of these trends were projected to continue over the next decade, increasing heterogeneities in TB incidence across cities and between populations. To sustain declines in TB incidence and to prevent an increase in Multiple DR-TB (MDR-TB) in the future in China, future TB control strategies may (i) need to be tailored to local demography, (ii) prioritize key populations, such as elderly and internal migrants, and (iii) enhance DR-TB surveillance.

Keywords: Tuberculosis; age-period-cohort model; aging; heterogeneity; urbanization.

Figure 1

Changes in the notification rates for males, females, and total population from 2006–2017 in (A) Jinan, (B) Yantai, (C) Linyi, and (D) Liaocheng. In each panel, the blue, red, and black lines represent notification rate trends; the P-Values of the Two-Proportion Z-Test for males, females, and total population in each cities are shown in the top-right corner.

Figure 2

Case density changes in 2006 and 2017 for males (upper subpanels) and females (lower subpanels) in (A) Jinan, (B) Yantai, (C) Linyi, and (D) Liaocheng. In each panel, red dashed-and-dotted and green solid lines represent densities in 2006 and 2017, respectively. We show the changes in case proportions (%) of young adults (18–35) and seniors (60+) in 2006 and 2017 in the top-right corner for males and in the bottom-right corner for females; the corresponding P-Values for the decrease in case proportion for young adults and the increase in that for seniors between 2006 and 2017 via One-sided Two-Proportion Z Tests are also demonstrated, respectively.

Figure 3

The proportion (%) of farmer cases among migrant TB cases (in solid brown line) and TB notification rates among migrants (in dashed green line) from 2007–2017 in (A) Jinan, (B) Linyi, and (C) Liaocheng, (D) Weifang, (E) Jining, and (F) Dezhou. P-Values of One-sided Two-Proportion Z-Tests for significant increase between 2007 and 2017 in proportion of farmer cases and in TB notification rate among migrants are listed in the bottom-right corner and the top-left corner, respectively.

Figure 4

Trends of TB notification rates from 2006-2017 and forecasts from 2018 to 2027 in (A) Jinan, (B) Yantai, (C) Linyi, and (D) Liaocheng. In each panel, (i) upper and low subpanels show trends for males and females, respectively; (ii) dots, lines, and shades colored in red, blue, and black indicate reported trends, forecast trends, and 95% confidential intervals for forecasts of annual TB notifications (in the 100 thousand population) for populations aged 18-35 and over 60 years and the total population, respectively; (iii) red, blue, and black dashed lines are smoothed splines that indicate the trends of annual reported TB notifications for the populations aged 18–35 and over 60 years and the total population, respectively.

Figure 5

Historical trends from 2006–2017 and case density changes in 2006 and 2017 over age of DS-TB (A,D), INH/RFP-resistant TB (B,E), and MDR-TB (C,F). In (D,E), red dashed-and-dotted and green solid lines represent densities in 2006 and 2017, respectively. We show the changes in case proportions of young adults (18–35) and seniors (60+) in 2006 and 2017 in the top-right corner for males and in the bottom-right corner for females; the corresponding P-Values for the decrease in case proportion for young adults and the increase in that for seniors between 2006 and 2017 via One-sided Two-Proportion Z Tests are also demonstrated, respectively.

Figure 6

Reported trends from 2006-2017 and forecasts from 2018-2023 of DR-TB infections for (A) INH-resistant TB, (B) RFP-resistant TB, and (C) MDR-TB, among the total population in Shandong, China. Within each panel, dotted-and-dashed lines and solid lines colored in green/blue/purple indicate reported DR-TB infections and forecast trends, under assumed test ratios of 50/75/100%, respectively. We constructed APC models to predict future total DR-TB notifications and smoothed splines to predict the future population in Shandong.