The impact of social and commercial determinants on the unequal increase of oral disorder disease burdens across global, regional, and national contexts | BMC Oral Health

The burden of oral disorders at global, regional, and National levels from 1990 to 2021

In 2021, oral disorders caused 23.24 million global DALYs (ASDR: 275.91) and affected nearly 3.7 billion people (prevalence: 3,692.29 million, ASPR: 45,908.54), with 3,739.81 million new cases (ASIR: 48,932.66). Males had slightly higher incidence, while females had higher DALYs and prevalence. From 1990 to 2021, global ASDR and ASPR slightly decreased (EAPCs: −0.08, −0.05), while ASIR increased (EAPC: 0.03). Regionally, South Asia had the highest cases (950.79 million) and Central Europe/Eastern Europe/Central Asia the highest ASIR (54,928.63) and ASPR (51,497.63). Southeast Asia/East Asia/Oceania had the highest prevalence (953.61 million) and DALYs (6.35 million). Latin America/Caribbean had the highest ASDR (373.93). Only South Asia and Latin America/Caribbean saw increases in ASDR and ASPR over time. Nationally, India recorded the most incident (724.27 million) and prevalent cases (713.05 million), China had the highest DALYs (4.4 million), Bolivia the highest ASPR (59,735.29) and ASDR (459.45), and Indonesia the highest ASIR (59,291.00). Significant regional heterogeneity existed in the burden (Fig. 1). (More detailed information on the oral disorders burden at global, regional, and national levels, as well as by sex, age, and specific disease subtypes, can be found in Appendices 1–6.)

Spatial distribution of DALYs, incidence and prevalence from oral disorders in 204 countries or territories in 2021. Panel (A) shows the ASDR of 204 countries or regions in 2021; Panel (B) shows the number of DALYs in 204 countries or regions in 2021; Panel (C) shows the ASIR of 204 countries or regions in 2021; Panel (D) shows the number of incidences in 204 countries or regions in 2021. Panel (E) shows the ASPR of 204 countries or regions in 2021; Panel (F) shows the amount of prevalence in 204 countries or regions in 2021

Trends in the disease burden of oral disorders across SDI quintiles

In the five SDI groups (Fig. 2), the ASIR in Low-middle and Low SDI regions declined rapidly between 2000 and 2010 but rose again from 2015 to 2018. In contrast, Middle and High-middle SDI regions experienced a sharp increase in ASIR from 2000 to 2005. The High-middle SDI region then stabilized from 2005 to 2015, followed by a decrease from 2015 to 2020, whereas the Middle SDI region continued its upward trend through 2020. For High SDI regions, 2010 marked a turning point, with ASIR decreasing before this year and increasing continuously afterward. The temporal changes in ASIR for oral disorders show significant heterogeneity across SDI groups, with the greatest increase in Middle SDI and the greatest decrease in Low SDI. Regarding ASPR, all SDI regions showed a steady decline from 1990 to 2000, followed by fluctuations from 2000 to 2015. From 2015 to 2021, only High SDI regions exhibited an increase, while others continued to decline. ASPR across SDI groups showed a stepwise distribution, with higher SDI levels associated with lower ASPR values. In Low-middle and Low SDI regions, ASDR followed an inverted V-shaped trend from 1990 to 2021, peaking in 2010, with Low-middle SDI having a significantly higher ASDR than other regions. Conversely, High-middle and High SDI regions exhibited a V-shaped trend with the lowest ASDR in 2010. Middle SDI regions saw a slow ASDR decline from 1990 to 2000, a rapid increase between 2000 and 2005, and subsequent stability, showing the smallest change among all SDI groups. The ASDR across SDI groups demonstrates a polarization trend, with 2010 as a critical point for change in low and high SDI regions.

Trends in the disease burden of oral disorders across SDI quintiles. Panels A, B, C show the trends in ASDR, ASIR and ASPR for each SDI quintile from 1990 to 2021. Panel D illustrates the EAPC value in ASDR, ASIR and ASPR for each SDI quintile from 1990 to 2021

In 2021, a nonlinear correlation was identified between the SDI and the ASIR across 204 countries or territories (Fig. 3). As the SDI increased from 0.5 to 0.75, moving from Middle to High-middle SDI, there was a substantial rise in the ASDR, ASIR, and ASPR of oral disorders. This trend is particularly pronounced in countries such as Brazil, Peru, and Bolivia within the Latin America and Caribbean region, where ASDR, ASIR, and ASPR are significantly elevated compared to other regions. These findings are consistent with the descriptions in the Lancet series regarding the relationship between SDI and disease burden. (Appendix Part 6 provides further descriptions)

The relationship between the SDI and the burden of oral disorders in 204 countries or territories in 2021. In bubble plots A, B, and C, the relationships between the SDI and the ASDR, ASIR, and ASPR for each country or territory in 2021 are shown, respectively. The color of the bubbles represents their corresponding super-regions

Temporal trends in the burden of oral disorders based on the Age-Period-Cohort model

The net drift measures the annual percentage change in the incidence and DALYs rates of oral disorders throughout the study period, while local drift represents the average annual percentage change for specific age groups (Fig. 4). Globally, and in most regions, there was a decline in the net drift of oral disorders. However, in regions characterized by middle and middle-high SDI levels, the net drift increased, with the most significant rise observed in the middle SDI region at 0.3274. The incidence rate for the 30–40 years age group showed an increasing trend across all regions except in low SDI regions, with all age groups in the middle SDI region experiencing a rise in incidence. Regarding DALYs, only the net drift in the middle and middle-low SDI regions increased, with the middle SDI region showing the largest increment at 0.1401. In the 0–5 years age group, DALYs decreased only in high and middle-high SDI regions, while the middle-low SDI region exhibited the highest growth rate. In the 15–20 years age group, the middle SDI region was the first to enter a growth phase, maintaining this trend until the 60–65 years age group, peaking at the 35–40 years age group. Globally and in other regions, excluding low SDI regions, the peak of DALYs growth was observed in the 30–40 years age group.

Net and local drift of oral disorders across different SDI groups from 1992 to 2021. The left panel displays the net and local drift of oral disorders incidence, while the right panel shows the net and local drift of DALYs

From 1992 to 2021, the effects of age, period, and cohort on the incidence and DALYs rates of oral disorders were evaluated (Fig. 5). Concerning age effects, the trends for incidence and DALYs were completely opposite. From 0 to 10 years, the incidence rate rose rapidly with age and then gradually decreased, whereas the DALYs rate consistently increased with age. Notably, the incidence rate was highest in high SDI regions at ages 20–25, while in other age groups, the incidence was highest in low and middle-low SDI regions. Before age 50, the DALYs rates in middle-low and low SDI regions were consistently higher than those in other regions, but after age 50, the DALYs rates in middle and middle-high SDI regions rose rapidly, with the middle SDI region maintaining the highest DALYs rates in subsequent age groups.

Concerning the period effect, high-middle and middle SDI regions experienced a continuous decline in incidence risk from 1992 to 1996. However, from 1997 to 2011, the risk of incidence rose rapidly, and from 2012 to 2021, the risk continued to rise in middle SDI regions. Conversely, the other three SDI regions saw a continuous decline in incidence risk from 1996 to 2011.Additionally, the DALYs rate showed the highest risk in low-middle SDI regions during 2007–2011, while middle SDI regions maintained a relatively high risk of DALYs rate from 2007 to 2021. Regarding cohort effects, before 1952, middle-high SDI regions had the highest risk of incidence, which gradually decreased over time. In contrast, middle SDI regions experienced a rise in incidence risk before 1952, which continued to increase post-1952, eventually becoming the region with the highest risk. Meanwhile, incidence risk in low-middle and low SDI regions gradually decreased after 1952.

Age-period-cohort effects of oral disorders across different SDI groups from 1992 to 2021. Panel A represents the age-period-cohort effects on the incidence of oral disorders, while Panel B illustrates the age-period-cohort effects on DALYs

Furthermore, we conducted an APC analysis on edentulism, which bears the heaviest burden of disease. The variations observed across different SDI differ from those associated with oral disorders, as detailed in Appendix Section seven.

Spatial heterogeneity in the burden of oral disorders based on the geographically weighted regression model

This study analyzed the ASIR of Edentulism, the most burdensome oral disease, across 204 countries or territories, utilizing the social and commercial determinants framework of oral health. The analysis employed traditional OLS as well as TWR, GWR, and GTWR models. Comparative results from these models are presented in Table 2, which includes the regression outcomes of the basic OLS model, collinearity tests, non-stationarity tests, and descriptive statistics of regression coefficients (see Appendix Table S4-S6). Typically, the corrected Akaike Information Criterion (AICc), R², and residual sum of squares (RSS) are used for model comparison, with smaller AICc and RSS values indicating better models, while a larger R² signifies better fit. As shown in Table 2, the GWR model outperformed the others, with the smallest AICc and RSS and the highest R² (0.8873). Consequently, this study selected the GWR model results for further analysis, with spatial visualizations of regression coefficients displayed in Fig. 6.

Regression coefficients of factors influencing the ASIR of edentulism in 204 countries or territories based on the GWR model. Panels A–I show the spatial distribution of regression coefficients for LDI_pc, Education_yrs_pc, Prop_urban, Dentists_pc, Pollution_pm25, Alc_gday_cont, Cigarettes_pc, Neg_exp_index, and Sugar_g, respectively

For traditional social determinants such as LDI_pc, Education_yrs_pc, and Prop_urban, significant differences in coefficient values were observed among LMICs. In many Asian and African countries, the coefficients for LDI_pc and Prop_urban were negative, while those for Education_yrs_pc were positive. Notably, the absolute values of these coefficients were higher in East Asia than in other regions. In Africa, the coefficients for these three factors displayed an opposite direction in West Africa compared to other regions, with significantly larger absolute values. In Southern Latin America (Chile, Argentina, and Uruguay), the coefficients for LDI_pc, Education_yrs_pc, and Prop_urban were opposite in sign to those in Asian and African countries, with the largest absolute values. In most HICs, the coefficients for these factors were negative, though their absolute values were closer to zero than those in LMICs. While income and urbanization effectively reduce the ASIR of Edentulism in most LMICs, education appears to have a negative impact; the influence of traditional social determinants is markedly lower in HICs.

Regarding the external social environment’s influence on the ASIR of Edentulism, factors such as Dentists_pc, Pollution_pm25, and Neg_exp_index exhibited significant spatial differences across various countries or regions. In HICs, Dentists_pc generally had a negative coefficient, whereas Pollution_ pm25 was positively correlated in North American and European nations. The coefficient for Neg_exp_index was negative in Western Europe but positive in Eastern Europe and other high-income regions. In LMICs, these three factors showed the greatest heterogeneity in Africa, where both positive and negative coefficients were present, particularly in West Africa, where they coexisted with larger absolute values. For example, within West African countries, the coefficient for Dentists_pc was 38.12 in Sierra Leone but − 16.04 in Senegal. In Asia, coefficients for Pollution_ pm25 and Neg_exp_index were completely opposed, with the former being negative and the latter positive. Moreover, Dentists_pc showed marked differences across Asia, with negative coefficients in Central and South Asia, positive in East Asia, and varying results in mainland Southeast Asia compared to island nations. Overall, Dentists_pc effectively reduces the ASIR in most countries, particularly in HICs; Pollution_ pm25 tends to increase ASIR in HICs but decreases it in many LMICs; Neg_exp_index generally increases ASIR across most countries.

Concerning commercial determinants of health, the coefficients for Alc_gday_cont and Cigarettes_pc displayed polarized trends across North America, East Asia, and Southeast Asia. Alc_gday_cont was significantly positive in North America but negative in East Asia and Southeast Asia. Conversely, Cigarettes_pc exhibited a negative coefficient in North America and a positive coefficient in East Asia and Southeast Asia. For Sugar_g, coefficients were positive in both North America and East Asia/Southeast Asia. Additionally, Alc_gday_cont demonstrated positive coefficients in Oceania, countries along the Red Sea, and some Western and Northern European nations, while remaining negative in most other regions. Cigarettes_pc showed positive coefficients in most African, Caribbean, and Central American countries, but negative values in most high-income nations, with the largest negative coefficients found in Western Europe. Notably, Sugar_g had negative coefficients in Caribbean and South American nations, as well as in Central Africa, Central Asia, and South Asia, while exhibiting variability among HICs, being negative in Japan, Korea, and Central Europe, contrary to trends in other HICs. The impact of health-related commercial determinants on the ASIR in LMICs is greater than in HICs. Compared to Alc_gday_cont, Cigarettes_pc and Sugar_g significantly increase the ASIR in LMICs, particularly within Africa, East Asia, and Southeast Asia.