Attribute risk burden and prediction analysis of cervical cancer under different sociodemographic transitions worldwide from 1990 to 2034

doi:10.21203/rs.3.rs-5012705/v1

Background

Cervical cancer is one of the four most prevalent cancers in women, posing a serious threat to women's health worldwide. In view of this, we aimed to investigate epidemiologic changes in the global burden of cervical cancer attributed to unsafe sex and smoking, respectively, and the relationship with social demographic index (SDI).

Methods

The data were obtained from the Global Burden of Disease 2019. Quantile regression and restricted cubic spline were employed to explore the relationship between age-standardized mortality rate (ASMR) or age-standardized Disability-Adjusted Life Years rate (ASDR) of cervical cancer and SDI under different risk factors. Additionally, we used the Nordpred model to predict the ASMR trends in five countries with different SDIs from 2020 to 2034.

Results

The global disease burden of cervical cancer was gradually decreasing. People aged 55–59 years and 95 + years had the highest age- specific incidence and mortality, respectively. The burden attributed to smoking showed an upward and then downward trend with increasing SDI, peaking at the SDI of 0.6, whereas the burden attributable to unsafe sex decreased as the SDI increased. The ASMR of cervical cancer around the world attributed to smoking and unsafe sex is projected to decrease to 0.58 and 6.39 per 100,000 by 2034, respectively. It is also predicted that the ASMR of cervical cancer attributed to smoking in China, and that attributed to unsafe sex in India and Argentina will increase in the future.

Conclusion

Unsafe sex is the main risk factor for cervical cancer, followed by smoking. People aged 55–59 years and over 95 years are the key protection groups for cervical cancer. Low SDI and middle-low SDI areas are the key prevention and control areas for cervical cancer. The projection suggests that China needs to strengthen rational control of smoking, and India and Argentina need to improve scientific prevention of unsafe sex.

cervical cancer

disease burden

smoking

unsafe sex

prediction

Cervical cancer is the fourth most prevalent cancer among women after breast, colorectal and lung cancers, and it is also the only gynecological malignant tumor with a clear cause that can be prevented and cured, and it has already been equipped with perfect three-level preventive and control measures, such as vaccines, screening, early diagnosis and early treatment^[1]. However, despite global, regional and national efforts to eradicate cervical cancer, it remains a major health problem in the world, and poses a great threat to women's health.

The results of the GLOBOCAN2020 database show that there will be 604,000 new cases of cervical cancer and 342,000 deaths around the world in 2020, accounting for 6.5% of new cases and 7.7% of deaths among females, with a crude incidence rate of cervical cancer of 15.6 per 100,000 and a crude mortality rate of 8.8 per 100,000 globally^{[2, 3]}. In many countries, this tumor leads the list of female tumors in terms of morbidity and mortality, but the mortality rate varies widely and the pattern of incidence is highly heterogeneous, with significant differences between countries and regions^[4]. Eastern Africa, Southern Africa, Central Africa, Melanesia and Western Africa have the highest incidence rates, While Western Europe, Northern Africa, North America, Australia, and Western Asia have the lowest age-standardized incidence rates.. Studies have indicated that there is a clear socio-economic gradient in cervical cancer incidence and mortality, with about 90% of cervical cancer cases occurring in low- and middle-income countries, where the mortality rate is 18 times higher than in rich countries^[5]. Similarly, the incidence of cervical cancer shows a significant increase in countries with a low Human Development Index (HDI). Morbidity and mortality in low HDI countries is three and six times higher than that in high HDI countries, respectively^[6]. The peak incidence of cervical cancer globally is 50–54 years of age, around 40 years in developed countries and 55–69 years in developing countries.^{[4, 7]}.

The main causative factor of cervical cancer is human papillomavirus (HPV), which is sexually transmitted. Persistent infection with high-risk HPV is essential for cervical cancer and its precancerous lesions^{[8, 9]}. There are 100 known subtypes of HPV, 16 of which are high-risk, and the main subtypes that cause 70% of cervical cancers are HPV 16 and 18. Factors such as smoking, early age of sexual debut, oral hormonal contraceptives, multiple sexual partners and unprotected sex increase the risk of HPV infection, which contributes to the disease burden of cervical cancer^[10]. At present, the prevention and control of cervical cancer are mainly primary and secondary prevention. Primary prevention is HPV vaccination and health education, and secondary prevention is cervical cancer screening, early diagnosis and early treatment^{[11, 12]}. The HPV vaccine is an effective preventive measure against cervical cancer. HPV vaccination can lead to a lower rate of HPV infection and effectively prevent most cervical cancers and genital warts^[13].

Although previous studies have explored the impact of SDI on the burden of disease in cervical cancer, they have not explicitly clarified the association under different risk factors. Therefore, this paper incorporates worldwide data on cervical cancer and utilizes various statistical analysis methods, focusing on studying the burden of attributable risk of cervical cancer at different levels of SDI globally and reasonably predicting the trend of attributable mortality of cervical cancer globally in the next 15 years.

2.1 Data sources

Data was obtained from the official website of the Global Burden of Disease Study 2019 (GBD2019) (http://ghdx.healthdata.org/gbd-results-tool), the study was led by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington. GBD2019 utilizes baseline datasets to quantify health levels and trends, and provides comprehensive estimates of the burden for 369 injuries, diseases and 87 risk factors in 204 countries and territories around the world^[14–16], making it the largest and most detailed global burden of disease study to date.

The ICD-10 codes for Cervical Cancer are C53, C53.0, C53.1, C53.3, C53.4, C53.8, C53.9 according to the Classification of Diseases codes. Risk factors associated with cervical cancer include smoking and unsafe sex^[17]. GBD smoking case definitions were current and former smoking of any tobacco products^{[16, 18]}. Current smokers were defined as individuals who currently use any smoked tobacco product on a daily or occasional basis, and exposure among current smokers was estimated for two continuous indicators: cigarettes per smoker per day and pack-years. Former smokers were defined as individuals who quit using all smoked tobacco products for at least six months, and exposure among former smokers was estimated for using years since cessation. Unsafe sex was defined as the risk of disease due to sexual transmission, including sexual intercourse without the use of contraception or without avoiding the contraction and transmission of sexually transmitted diseases or in an unhealthy state of mind^{[16, 19, 20]}. Age-standardized mortality rate (ASMR) and age-standardized Disability-Adjusted Life Years rate (ASDR) are standardized according to the GBD2019 World Standard Population Composition.

Disability-adjusted life years (DALYs), which represent the time lost due to disability or death caused by a specific disease, are considered an appropriate indicator to quantify the burden of disease and injury, cost-effectiveness and resource allocation^[21]. It is a combination of Years of Life Lost (YLL) and Years of Life with disability (YLD). YLL is the number of years of life lost due to premature death below life expectancy caused by a disease^[22]. YLD is the number of years of healthy life lost as a result of disability caused by a disease, it were calculated by a standardized method using prevalence of each disease sequelae and assigned disability weights (DW)^[23]. GBD2019 used comparative risk assessment (CRA) to estimate the proportion of deaths and DALYs attributable to risk factors by comparing the actual exposure of the target population with the Theoretical Minimum Risk Exposure Level (TMREL), holding other risk factors constant. In addition, population attributable fractions (PAFs) corresponding to age-sex-region-year were estimated for each risk factor to calculate the percentage of morbidity or mortality from a disease attributable to exposure to a particular risk factor out of the total morbidity or mortality from that disease in the total population^[16].

The Sociodemographic Index (SDI), which consists of three components: average education in individuals aged $\:\ge\:$15 years, per capita income and the total fertility rate of the population aged $\:<$25 years, is a composite indicator of development status correlated with health outcome. The value of SDI ranges from 0 to 1. 204 countries and territories were divided into 5 categories based on SDI, including low (0-0.454743), low-middle (0.454743–0.607679), middle (0.607679–0.689504), high-middle (0.689504–0.805129), and high (0.805129-1)^[24].

2.2 Statistical analysis

ASR and EAPC

We used standardized rates (ASR) to analyze trends in the burden of cervical cancer over the past 30 years, where standardized mortality rates (ASMR) and standardized Disability-Adjusted Life Years rates (ASMR) were calculated as follows^[25]:

In the ith age subgroup, a_i denotes the age-specific mortality or DALY rate, w is represented as the number (or weight) of people in the selected standard population, and A denotes the total number of age groups.

Annual percentage change (EAPCs) was adopted to accurately assess the epidemiological trend of cervical cancer mortality rate and disability adjusted life age.

EAPCs and its 95% uncertainty interval (UI) were calculated by a linear regression model as follows, where x refers to the calendar year:

$$\:\text{ln}\left(ASR\right)=\alpha\:+\beta\:x+\epsilon\:$$

$\:\text{E}\text{A}\text{P}\text{C}=100\times\:({e}^{{\beta\:}}-1$ )

If the upper limit and confidence interval of the EAPC are less than zero, the ASR represents the mortality or the disability adjusted life year of cervical cancer on the decline, and vice versa^[26].

Quantile regression

Quantile regression is a linear regression under different quantile conditions, which is based on the principle of weighted least squares method. It is a class of regression models rather than a single model^[27]. Quantile regression can fit multiple regression lines based on quantile points, providing more comprehensive information to help understand the full picture of the conditional distribution of the dependent variable, and has better robustness to outliers^[28].

Restricted cubic spline (RCS)

RCS is one of the most common methods for analyzing nonlinear relationships. It is a powerful and flexible nonparametric method that does not require prior specification of the functional form, and avoids overfitting while maintaining curve smoothness. RCS is essentially a regression spline with additional requirements. A regression spline is a segmented polynomial, which is generally required to be continuous and second-order derivable at each segmentation point in order to ensure the smoothness of the curve^[29]. These segmentation points are called "nodes" and the number of nodes affects the fit of the model. The recommended number of nodes is 3 to 5, and we used 3 nodes in this article.

Nordpred age-period-cohort

The prediction of chronic disease incidence and death is complicated by the influence of factors such as age, period, and birth cohort. The Nordpred age-period cohort model predicts based on the independent effects of age, period, and cohort, using different trend parameters for prediction and emphasizing the weighting of recent trends. It overcomes the problem of exponential growth of rates over time in the classical APC model and makes the prediction results more accurate^[30]. At the same time, this model has been shown to be a better predictor of estimated tumor incidence and mortality trends, and has been widely used in studies in different countries^{[31, 32]}. The Nordpred model makes predictions based on either a power function or a Poisson distribution function, where power function is more suitable for cancer incidence and mortality projection. We have divided 1990–2019 into six time periods, with five-year intervals, to predict the trends of cervical cancer mortality in countries with different SDI classes from 2020 to 2034. The R software 4.1.3 was applied for statistical analyses, and P $\:<$0.05 considered statistically significant.

3.1 Temporal trends in different SDI levels

In Table 1, we showed the number of cervical cancer deaths, ASMR, and EAPC globally, in the five SDI classes, and in 21 regions from 1990 to 2019. The number of deaths is increasing at a global level, with 184,527 (95% UI: 164,836 − 218,942) deaths in 1990 and 280,479 (95% UI: 238,864 − 313,930) deaths in 2019, representing a 52% growth of the total deaths rate of this disease. The ASMR showed a downward trend, with an average annual decrease of 0.93% (EAPC = -0.93; 95% UI: -0.98 - -0.88), from 8.48/100000 people (95% UI, 7.59–10.07) in 1990 to 6.51/100000 people (95% UI, 5.55–7.29) in 2019. For specific DALYs and ASDR in 1990 and 2019, please refer to Table 2. The overall DALY count is on an upward trend, growing from 6,176,248 DALY (95% UI: 5437672–7316926) in 1990 to 8,955,013 DALY(95% UI: 7547733–9978462) in 2019, with total DALY rate growth of 44.99%. The ASDR decrease obviously, with an average annual decrease of 0.95% (EAPC = -0.95, 95% UI: -1.00 - -0.90).

Table 1

Number of deaths and age-standardized death rates for cervical cancer globally, in SDI regions and geographic areas, 1990 and 2019, and percentage change in EAPC from 1990 to 2019
Characteristics	1990		2019		1990–2019
Characteristics	Deaths No.×10³(95%UI)	ASMR per 100,000 No.(95%UI)	Deaths No.×10³(95%UI)	ASMR per 100,000 No.(95%UI)	EAPC in ASMR No.(95%UI)
Global	184.53 (164.84−218.94)	8.48 (7.59–10.07)	280.48 (238.86−313.93)	6.51 (5.55–7.29)	−0.93 (−0.98–0.88)
Sex
Female	184.53 (164.84−218.94)	8.48 (7.59–10.07)	280.48 (238.86−313.93)	6.51 (5.55–7.29)	−0.93 (−0.98–0.88)
SDI region
High SDI	25.22 (23.28–26.19)	4.56 (4.22–4.71)	26.17 (22.82–28.15)	2.90 (2.60–3.10)	−1.57 (−1.68–1.46)
High-middle SDI	41.35 (38.69–48.40)	6.95 (6.50–8.13)	51.77 (41.66–57.87)	4.89 (3.92–5.47)	−1.25 (−1.31–1.19)
Middle SDI	52.53 (46.63–65.12)	9.32 (8.31–11.54)	90.10 (71.33–103.20)	6.78 (5.40–7.76)	−1.03 (−1.09–0.97)
Low-middle SDI	39.21 (32.46–50.05)	11.71 (9.73–15.05)	66.68 (57.27–81.24)	8.85 (7.62–10.83)	−1.04 (−1.12–0.96)
Low SDI	26.08 (20.23–32.11)	19.18 (15.00−23.66)	45.54 (35.80−56.26)	15.05 (11.92–18.46)	−0.90 (−0.94–0.87)
GBD region
Andean Latin America	2.33 (1.96–2.76)	20.39 (17.22–24.05)	4.28 (3.32–5.38)	14.37 (11.18–18.04)	−1.33 (−1.44–1.22)
Australasia	0.45 (0.39–0.48)	3.73 (3.16–3.92)	0.52 (0.45–0.58)	2.17 (1.88–2.40)	−1.57 (−1.95–1.19)
Caribbean	2.23 (1.75–2.55)	15.83 (12.55–18.04)	3.47 (2.72–4.26)	12.95 (10.11–15.96)	−0.65 (−0.72–0.58)
Central Asia	2.72 (2.48–2.89)	9.81 (8.95–10.43)	3.42 (3.00−3.93)	7.58 (6.68–8.70)	−0.75 (−0.88–0.62)
Central Europe	8.00 (7.60–8.53)	10.14 (9.63–10.79)	6.88 (5.82–7.99)	6.65 (5.59–7.75)	−1.57 (−1.68–1.47)
Central Latin America	9.59 (8.70−10.01)	20.35 (18.19–21.28)	13.83 (11.53–16.80)	10.65 (8.91–12.92)	−2.61 (−2.76–2.46)
Central Sub-Saharan Africa	3.72 (2.61–4.84)	26.27 (18.78–34.25)	7.30 (4.91–10.06)	21.67 (14.49–30.24)	−0.67 (−0.78–0.55)
East Asia	28.40 (22.32–46.14)	6.05 (4.77–9.76)	55.96 (33.19–71.36)	5.18 (3.09–6.59)	−0.05 (−0.28−0.18)
Eastern Europe	12.94 (10.95–13.86)	7.62 (6.50–8.19)	10.04 (8.47–11.91)	5.54 (4.62–6.61)	−1.38 (−1.54–1.21)
Eastern Sub-Saharan Africa	11.94 (9.08–14.92)	26.51 (20.07–33.45)	21.11 (15.48–27.86)	21.13 (15.15–27.62)	−0.90 (−0.96–0.84)
High-income Asia Pacific	4.64 (4.35–5.32)	4.20 (3.93–4.81)	5.60 (4.58–6.22)	2.70 (2.22–2.96)	−1.52 (−1.59–1.44)
High-income North America	6.74 (5.97–7.04)	3.71 (3.25–3.86)	8.80 (7.47–9.34)	2.99 (2.55–3.15)	−0.69 (−0.83–0.55)
North Africa and Middle East	3.97 (2.82–4.53)	4.37 (3.10–4.99)	7.00 (5.44–8.31)	3.15 (2.47–3.69)	−1.11 (−1.21–1.01)
Oceania	0.30 (0.22–0.42)	18.16 (13.38–25.22)	0.67 (0.45–0.91)	16.41 (11.50−22.19)	−0.18 (−0.27–0.08)
South Asia	33.33 (26.22–39.95)	10.49 (8.29–12.62)	53.30 (42.87–69.95)	7.01 (5.66–9.21)	−1.60 (−1.77–1.42)
Southeast Asia	16.71 (12.57–21.47)	11.05 (8.36–14.54)	25.13 (20.52–34.98)	7.36 (6.03–10.33)	−1.52 (−1.61–1.44)
Southern Latin America	3.07 (2.89–3.31)	12.34 (11.62–13.30)	4.18 (3.55–4.60)	9.64 (8.17–10.56)	−1.01 (−1.12–0.89)
Southern Sub-Saharan Africa	3.24 (2.48–4.07)	19.17 (14.63–24.27)	6.56 (5.39–7.75)	19.34 (15.82–22.77)	0.46 (0.19–0.72)
Tropical Latin America	7.69 (7.21–8.88)	14.72 (13.70−17.12)	11.58 (10.72–13.66)	8.69 (8.04–10.23)	−2.01 (−2.11–1.92)
Western Europe	13.13 (12.04–13.61)	4.36 (3.99–4.51)	11.75 (10.27–12.69)	2.65 (2.38–2.85)	−1.65 (−1.75–1.54)
Western Sub-Saharan Africa	9.38 (7.56–12.31)	19.74 (15.94–25.76)	19.09 (15.04–24.01)	16.83 (13.38−21.00)	−0.48 (−0.53–0.43)

Table 2

Number of DALY and age-standardized DALY rates for cervical cancer globally, in SDI regions and geographic areas, 1990 and 2019, and percentage change in EAPC from 1990 to 2019
Characteristics	1990		2019		1990–2019
Characteristics	DALY No.×10³(95%UI)	ASDR per 100,000 No.(95%UI)	DALY No.×10³(95%UI)	ASDR per 100,000 No.(95%UI)	EAPC in ASDR No.(95%UI)
Global	6176.25 (5437.67−7316.93)	275.05 (242.75−326.15)	8955.01 (7547.73−9978.46)	210.64 (177.67−234.85)	−0.95 (−1.00–0.90)
Sex
Female	6176.25 (5437.67−7316.93)	275.05 (242.75−326.15)	8955.01 (7547.73−9978.46)	210.64 (177.67−234.85)	−0.95 (−1.00–0.90)
SDI region
High SDI	725.85 (665.20−752.59)	143.23 (130.30−148.45)	672.11 (608.75–722.00)	89.72 (81.88–95.85)	−1.62 (−1.74–1.49)
High-middle SDI	1274.57 (1193.74−1497.01)	215.21 (201.46−252.85)	1543.70 (1236.00−1729.87)	154.69 (124.02−173.51)	−1.16 (−1.22–1.10)
Middle SDI	1790.63 (1588.48−2223.17)	287.82 (255.02−356.33)	2817.25 (2223.19−3217.72)	204.60 (161.92−233.49)	−1.11 (−1.17–1.05)
Low-middle SDI	1419.29 (1160.75−1789.13)	381.90 (315.26−485.47)	2282.24 (1948.33−2722.93)	285.64 (244.64−342.16)	−1.08 (−1.17–0.99)
Low SDI	961.20 (732.59−1179.64)	630.59 (487.61−777.41)	1632.49 (1271.61−2044.29)	477.53 (374.33−591.38)	−1.05 (−1.09–1.00)
GBD region
Andean Latin America	79.25 (66.33–92.90)	633.25 (529.46−743.28)	129.59 (99.42–165.40)	422.28 (323.97–538.40)	−1.54 (−1.66–1.42)
Australasia	13.93 (11.33–14.68)	120.58 (96.79−127.24)	13.58 (11.84–15.06)	65.47 (57.37–72.70)	−1.83 (−2.26–1.41)
Caribbean	77.36 (58.62–90.62)	526.18 (401.54−612.81)	114.71 (86.80−145.02)	438.19 (328.34−557.85)	−0.63 (−0.69–0.56)
Central Asia	89.50 (83.58–95.46)	317.62 (299.06−339.68)	119.72 (103.95−138.54)	249.41 (217.42−288.15)	−0.72 (−0.87–0.58)
Central Europe	251.14 (236.32−264.12)	333.96 (312.61−350.79)	190.26 (159.63−221.33)	212.08 (177.26−247.37)	−1.73 (−1.86–1.60)
Central Latin America	323.66 (302.50−337.95)	610.29 (562.25−636.93)	436.92 (361.75−538.52)	328.59 (272.54−404.33)	−2.48 (−2.63–2.32)
Central Sub-Saharan Africa	135.73 (92.58−179.24)	848.19 (592.20−1109.46)	261.63 (176.04−360.43)	678.72 (454.78−932.08)	−0.77 (−0.90–0.65)
East Asia	921.11 (714.66−1517.19)	182.74 (142.14−297.89)	1696.32 (972.55−2166.63)	159.12 (91.62−202.95)	0.02 (−0.20−0.24)
Eastern Europe	368.58 (319.59−396.93)	233.88 (203.92–253.70)	308.61 (255.79−369.57)	192.88 (156.89−231.51)	−0.89 (−1.09–0.69)
Eastern Sub-Saharan Africa	444.59 (335.91−562.71)	876.65 (666.35−1103.42)	758.61 (557.09−1022.32)	660.28 (484.65−874.17)	−1.13 (−1.21–1.06)
High-income Asia Pacific	136.98 (127.56−156.31)	126.77 (118.36−144.52)	133.64 (109.97−146.61)	85.67 (67.59–93.64)	−1.29 (−1.35–1.23)
High-income North America	206.88 (177.72.1−216.04)	123.61 (105.30−129.07)	245.96 (211.94−259.26)	96.55 (83.94−101.88)	−0.84 (−0.99–0.69)
North Africa and Middle East	133.00 (94.24–152.10)	130.16 (92.26−148.26)	221.93 (169.20−268.19)	88.28 (68.42−105.74)	−1.37 (−1.44–1.29)
Oceania	11.34 (7.93–15.07)	579.78 (417.98−793.65)	24.91 (16.09–34.06)	521.37 (347.49−709.69)	−0.17 (−0.27–0.08)
South Asia	1230.95 (969.02−1468.76)	345.39 (271.36–413.30)	1833.69 (1466.66−2370.91)	226.59 (181.92−292.64)	−1.64 (−1.83–1.45)
Southeast Asia	592.31 (440.72–733.90)	354.96 (266.86−447.24)	808.25 (653.21−1088.29)	223.36 (181.44−302.65)	−1.73 (−1.83–1.64)
Southern Latin America	101.76 (96.27−107.78)	413.34 (390.66−437.19)	127.49 (105.40−140.05)	317.23 (260.30−348.08)	−1.07 (−1.18–0.96)
Southern Sub-Saharan Africa	116.35 (89.85−142.79)	633.63 (488.91−782.66)	213.94 (173.97−254.95)	586.79 (476.20−698.37)	0.23 (−0.04−0.51)
Tropical Latin America	262.80 (247.91−303.91)	455.15 (428.99–524.50)	365.28 (340.28−419.75)	274.27 (255.50−314.34)	−1.95 (−2.05–1.85)
Western Europe	351.50 (321.56−363.85)	134.35 (119.97−138.99)	277.36 (248.48−299.74)	79.19 (71.88–85.34)	−1.77 (−1.90–1.64)
Western Sub-Saharan Africa	327.53 (261.59−424.73)	626.47 (502.32–815.10)	672.60 (524.74−854.86)	507.97 (398.99−640.76)	−0.69 (−0.74–0.64)

Figure 1A and Fig. 1B presented the ASMRs trends of cervical cancer attributable to smoking and unsafe sex at five different SDI levels from 1990 to 2019, respectively. The ASMRs for cervical cancer attributable to both risk factors showed an overall decreasing trend around the world and in different SDI levels. The values of ASMRs attributed to unsafe sex were significantly higher than those of smoking. The ASMR are highest in low SDI areas, where the ASMR attributable to smoking decreased from 1.43/100,000 to 0.95/100,000 from 1990 to 2019, and the ASMR attributable to unsafe sex declined from 19.18/100,000 to 15.05/100,000. The greatest decreases in ASMR attributed to smoking were observed in the high SDI countries, which declined by 47.80% (from 1.43 to 0.75 per 100000 people ). While the ASMR values attributed to unsafe sex decreased the most in low SDI countries, from 19.18 to 15.05 per 100000 people. The ASDRs for cervical cancer attributed to these two risk factors followed similar trends to those of ASMRs (Supplementary Material, e-Figure1).

3.2 The age-specific mortality and incidence rates by age group

The age-specific mortality rates for cervical cancer attributable to unsafe sex and smoking from five different SDI levels globally are displayed in Fig. 2 and e-Figure 2. The age-specific mortality rates for cervical cancer attributable to both risk factors showed a downward trend across SDI levels, and the age-specific mortality rates attributable to unsafe sex were significantly higher than those of smoking. The age group with the highest age-specific mortality from cervical cancer associated with unsafe sex was predominantly over 95 years in five SDI levels. The mortality of cervical cancer attributable to unsafe sex in the people aged over 95 years is highest in low SDI areas, increasing from 74.04 per 100000 people in 1990 to 89.00 per 100000 people in 2019. Between 1990 and 2019, the global age-specific mortality rate attributable to unsafe sex decreased from 45.50 per 100,000 to 40.45 per 100,000 (Fig. 2), and the rate attributable to smoking showed a downtrend from 6.55 per 100,000 to 3.89 per 100,000 (e-Figure 2).

We also collected data from five SDI levels in the world to investigate the age-specific incidence of cervical cancer attributable to the two risk factors, and the rates are presented in Fig. 3. The incidence rates in 2019 were lower than those in 1990 in most age groups, and the age-specific incidence of 55-59-years was observed the highest. The age-specific incidence of 55-59-years for this cancer showed the lowest in high SDI quintile, decreasing from 27.00 per 100000 in 1990 to 19.47 per 100000 in 2019. While the highest age-specific incidence of this age group from cervical cancer was observed in the low SDI level. The incidence was 79.06 per 100000 and 66,32 per 100000 in 1990 and 2019, respectively.

3.3 The proportion of YLL in DALY

As shown in Fig. 4, the component ratios of YLL for cervical cancer attributed to unsafe sex and smoking in DALYs were above 95% worldwide, and both were on a downward trend. The proportions of YLL attributed to unsafe sex was higher in high-middle SDI areas, while those of smoking was higher in other SDI quintiles. The percentage of YLL of cervical cancer attributed to both risk factors declined progressively with increasing SDI level.

3.4 The relationship between SDI and ASMR or ASDR of cervical cancer

We used RCS and quantile regression to analyse the relationship between ASMR, ASDR of cervical cancer and SDI values in 21 regions classified by GBD, which was presented in Fig. 5. When the SDI was close to 0.6, the ASMR of cervical cancer attributed to smoking was the highest. The results of quantile regression(P₂₅) was statistically significant, and the ASMR attributed to smoking showed a downward trend with the increase in SDI. In addition, five quintiles were statistically significant, and lower ASMRs for cervical cancer were attributed to unsafe sex as the SDI increase. A similar result was obtained for the relationship between ASDR and SDI (e-Figure 3).

3.5 Predictions of ASMR for cervical cancer in different SDI countries from 2020 to 2034

The projected ASMR trends for cervical cancer attributed to both two risk factors in five different SDI countries were exhibited in Fig. 6. In the next 15 years, the ASMR for cervical cancer attributed to unsafe sex was predicted to decrease in global, Canada, China and Afghanistan, while that in Argentina and India were projected to increase to 11.30 per 100000 and 7.79 per 100000 in 2034, respectively. The ASMR for cervical cancer attributed to smoking was projected to show an upward trend to 0.53 per 100,000 in China. In contrast, the values in other four countries and global were projected to decline.

We found the disease burden of cervical cancer shows a progressive reduction from 1990 to 2019, with a more significant burden attributable to unsafe sex than smoking. This result ties well with previous studies wherein epidemiological trends and attributable risk burden of cervical cancer. In addition, we also explored the age groups with the highest incidence rate and the highest attributed mortality rate, which were 55–59 years and over 95 years respectively..When comparing our results to those of older studies, we innovatively found that the burden of death attributed to both risk factors accounted for more than 95% of DALY, and the proportions of years lived with disability (YLD) showed an upward trend. This may be due to the fact that the disease burden of cervical cancer arises primarily from years of life lost (YLL) because of premature mortality.. Additionally early screening reduces the risk of premature death from cervical cancer, which in turn increases the burden of loss of healthy life expectancy because of disability.

This paper has presented the results that the burden of cervical cancer attributed to smoking peaks at an SDI close to 0.6 (Low-middle SDI). In 2005, the World Health Organization promulgated the Framework Convention on Tobacco Control^[33], which was formally adopted for implementation in many countries and regions. However, the situation with regard to smoking-attributable diseases among women remains critical due to wide disparities in the implementation of tobacco control in different countries and regions, especially in low and low-middle SDI areas, where the issue of women's tobacco use is not yet fully recognized^[34]. Global cancer statistics 2018 showed that 290,000 of the 570,000 new cases of cervical cancer worldwide occur in women in low- and middle-income countries^[35]. A study by Vaccarella et al^[36] also exhibited that 86% of the cervical cancer burden in the world occurs in Africa, Latin America and the Caribbean and Asia, with a higher concentration of low SDI countries in these regions. Thus, smoking may be one of the reasons for the more serious burden of cervical cancer among women in low and low-middle SDI countries.

In addition, our study found that the burden of cervical cancer attributed to unsafe sex was highest in low SDI countries and lowest in high SDI countries. The HPV virus is the most important risk factor for cervical cancer, and the main route of infection is through unsafe sex. The overall prevalence of HPV in Latin America and the Caribbean is two times higher than the global average^[37], and the disease burden of cervical cancer is among the highest in the world. However, in France (high SDI), the organization of national screening for cervical cancer^[38], and the implementation of a gender-neutral vaccination program^[39], have significantly reduced the burden of cervical cancer..An study in Indian^[40] (low-middle SDI) showed that the level of adequacy of knowledge, attitudes, and practice of cervical cancer screening among Indian women was only 20%, 40.8%, and 13.8%, respectively, which was much lower than a similar study in Argentina (high-middle SDI). It has also been demonstrated that high rates of HIV infection will greatly increase the burden of cervical cancer^[41]. In Uganda (low SDI), high HIV prevalence and incidence as well as low treatment level increase the burden of cervical cancer attributed to unsafe sex^[42]. In contrast, the United States (high SDI), where antiretroviral treatment and comprehensive care have been widely available for about half of all HIV-diagnosed patients for decades^[43], has a much lower burden of cervical cancer than low SDI countries. Overall, the large variation in the burden of cervical cancer attributable to unsafe sex across SDI quintiles may be related to numerous factors, such as the allocation of health care resources, the delivery of health education, the treatment and management of HIV, the prevalence of HPV vaccination and cervical cancer screening.

Projections suggest that cervical cancer mortality attributable to both risk factors will decrease over the next 15 years globally, in Canada, and in Afghanistan, and similar results were obtained in the global multi-country study by Shujuan Lin et al^[44]. However, in contrast to that study, our results indicated that the mortality of cervical cancer attributable to smoking in China and to unsafe sex in India and Argentina will increase. The probable reason for this is that Shujuan Lin's study was an overall mortality prediction for cervical cancer, whereas the present study was a further refinement, with separate mortality predictions for cervical cancer attributable to the two risk factors.

There were some advantages in our study. Firstly, this study is a large and systematic global analysis describing the epidemiological characteristics and shifts in the attributable burden of two risk factors for cervical cancer across different SDI levels of the world. Secondly, existing studies have focused on the descriptive analysis of disease burden for cervical cancer, while this study further explored the relationship between attributable burden of cervical cancer and SDI by using more comprehensive statistical methods such as quantile regression and restricted cubic spline. Thirdly, we predicted the mortality trends of cervical cancer caused by smoking and unsafe sex respectively, providing a scientific basis for future disease prevention and control of cervical cancer. This study also had several limitations. Firstly, predicting the burden of cervical cancer attributable to two risk factors in 204 countries would require a significant amount of work, so only five countries representing different SDI quintiles were selected for this study. Secondly, this study is based on population- observations and analysis, lacking the individual information. Thirdly, this study has not reported the burden of disease for different case types of cervical cancer.

The global burden of cervical cancer is on the decline, but the situation remains serious. The primary burden was mainly from the loss of years of life due to premature death. Unsafe sex is the main risk factor for cervical cancer, followed by smoking. People aged 55–59 years and over 95 years are the key protection groups for cervical cancer. Low SDI and middle-low SDI areas are the key prevention and control areas for cervical cancer. It is necessary to take targeted measures on risk factors of cervical cancer in specific countries in the future.

Ethics approval and consent to participate

Not applicable

Competing interests

The authors have no conflict of interest to declare.

Funding

This research was funded by the 11th batch of innovative provincial construction projects in 2023 in Hunan Province (Grant 2023ZK4111) .

Author Contribution

JH acquired the data, performed the analysis of data, and wrote the manuscript. YW contributed to the analysis of data and the coding of the statistical analysis. ZC designed and evaluated the whole work. All authors read and approved the final manuscript.

Acknowledgement

Thanks to GBD for providing data resources.

Yang M, Du J, Lu H, et al. Global trends and age-specific incidence and mortality of cervical cancer from 1990 to 2019: an international comparative study based on the Global Burden of Disease[J]. BMJ Open. 2022;12(7):e055470.
Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries[J]. CA Cancer J Clin. 2021;71(3):209–49.
Singh D, Vignat J, Lorenzoni V, et al. Global estimates of incidence and mortality of cervical cancer in 2020: a baseline analysis of the WHO Global Cervical Cancer Elimination Initiative[J]. Lancet Glob Health. 2023;11(2):e197–206.
Arbyn M, Weiderpass E, Bruni L, et al. Estimates of incidence and mortality of cervical cancer in 2018: a worldwide analysis[J]. Lancet Glob Health. 2020;8(2):e191–203.
Shen TT, Long CY, Wu MP. Favorable cervical cancer mortality-to-incidence ratios of countries with good human development index rankings and high health expenditures[J]. BMC Womens Health. 2023;23(1):284.
Singh D, Vignat J, Lorenzoni V, et al. Global estimates of incidence and mortality of cervical cancer in 2020: a baseline analysis of the WHO Global Cervical Cancer Elimination Initiative[J]. Lancet Glob Health. 2023;11(2):197–206.
Saito E, Hori M. Age-specific cervical cancer incidence rate in the world[J]. Jpn J Clin Oncol. 2020;50(10):1229–30.
Burd EM. Human papillomavirus and cervical cancer[J]. Clin Microbiol Rev. 2003;16(1):1–17.
Yuan Y, Cai X, Shen F, et al. HPV post-infection microenvironment and cervical cancer[J]. Cancer Lett. 2021;497:243–54.
Momenimovahed Z, Mazidimoradi A, Maroofi P, et al. Global, regional and national burden, incidence, and mortality of cervical cancer[J]. Cancer Rep (Hoboken). 2023;6(3):e1756.
Pimple SA, Mishra GA. Global strategies for cervical cancer prevention and screening[J]. Minerva Ginecol. 2019;71(4):313–20.
Lim TSC, Ismail-Pratt I, Goh LH. Cervical cancer screening and vaccination: understanding the latest guidelines[J]. Singap Med J. 2022;63(3):125–9.
Ahmad M, Asrar R, Ahmed I, et al. HPV vaccination: A key strategy for preventing cervical cancer[J]. J Infect Public Health. 2024;17(3):474–5.
Collaborators GD. Global age-sex-specific fertility, mortality, healthy life expectancy (HALE), and population estimates in 204 countries and territories, 1950–2019: a comprehensive demographic analysis for the Global Burden of Disease Study 2019[J]. Lancet. 2020;396(10258):1160–203.
Collaborators GDaI. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019[J]. Lancet. 2020;396(10258):1204–22.
Collaborators GRF. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019[J]. Lancet. 2020;396(10258):1223–49.
Yao H, Yan C, Qiumin H et al. Epidemiological Trends and Attributable Risk Burden of Cervical Cancer: An Observational Study from 1990 to 2019[J]. Int J Clin Pract, 2022, 2022: 3356431.
Yuan R, Ren F, Xie Y, et al. The Global, Regional, and National Burdens of Cervical Cancer Attributable to Smoking From 1990 to 2019: Population-Based Study[J]. JMIR Public Health Surveill. 2022;8(12):e40657.
Qiu P, He H, Zhao Y, et al. Global disease burden attributed to unsafe sex in 204 countries and territories from 1990 to 2019: results from the Global Burden of Disease Study 2019[J]. Sci Rep. 2023;13(1):12900.
Marcus JL, Snowden JM. Words Matter: Putting an End to Unsafe and Risky. Sex[J] Sex Transm Dis. 2020;47(1):1–3.
Saxena N, Sethia D. Decomposition of years of life lost due to premature death (YLL): a method for spatial and temporal comparative assessment[J]. Arch Public Health. 2020;78:91.
Prevalence and attributable health burden of chronic respiratory diseases. 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017[J]. Lancet Respir Med. 2020;8(6):585–96.
Salomon JA, Vos T, Hogan DR, et al. Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010[J]. Lancet. 2012;380(9859):2129–43.
Ferrari G, Farías-Valenzuela C, Guzmán-Habinger J, et al. Relationship between socio-demographic correlates and human development index with physical activity and sedentary time in a cross-sectional multicenter study[J]. BMC Public Health. 2022;22(1):669.
Cao G, Liu J, Liu M, Global. Regional, and National Incidence and Mortality of Neonatal Preterm Birth, 1990–2019[J]. JAMA Pediatr. 2022;176(8):787–96.
Pu L, Wang L, Zhang R, et al. Projected Global Trends in Ischemic Stroke Incidence, Deaths and Disability-Adjusted Life Years From 2020 to 2030[J]. Stroke. 2023;54(5):1330–9.
Liu Y, Wen H, Bai J, et al. Disease Burden and Prediction Analysis of Tracheal, Bronchus, and Lung Cancer Attributable to Residential Radon, Solid Fuels, and Particulate Matter Pollution Under Different Sociodemographic Transitions From 1990 to 2030[J]. Chest. 2024;165(2):446–60.
Staffa SJ, Kohane DS, Zurakowski D. Quantile Regression and Its Applications: A Primer for Anesthesiologists[J]. Anesth Analg. 2019;128(4):820–30.
Wu M, Si J, Liu Y, et al. Association between composite dietary antioxidant index and hypertension: insights from NHANES[J]. Clin Exp Hypertens. 2023;45(1):2233712.
He Q, Mok TN, Sin TH, et al. Global, Regional, and National Prevalence of Gout From 1990 to 2019: Age-Period-Cohort Analysis With Future Burden Prediction[J]. JMIR Public Health Surveill. 2023;9:e45943.
Son M, Yun JW. Cancer Mortality Projections in Korea up to 2032[J]. J Korean Med Sci. 2016;31(6):892–901.
Fukui K, Ito Y, Nakayama T. Trends and projections of cancer mortality in Osaka, Japan from 1977 to 2032[J]. Jpn J Clin Oncol. 2019;49(4):383–8.
Chung-Hall J, Craig L, Gravely S, et al. Impact of the WHO FCTC over the first decade: a global evidence review prepared for the Impact Assessment Expert Group[J]. Tob Control. 2019;28(Suppl 2):s119–28.
Amos A, Greaves L, Nichter M, et al. Women and tobacco: a call for including gender in tobacco control research, policy and practice[J]. Tob Control. 2012;21(2):236–43.
Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin. 2018;68(6):394–424.
Vaccarella S, Laversanne M, Ferlay J, et al. Cervical cancer in Africa, Latin America and the Caribbean and Asia: Regional inequalities and changing trends[J]. Int J Cancer. 2017;141(10):1997–2001.
Almonte M, Albero G, Molano M, et al. Risk factors for human papillomavirus exposure and co-factors for cervical cancer in Latin America and the Caribbean[J]. Vaccine. 2008;26(Suppl 11):L16–36.
Hamers FF, Duport N, Beltzer N. Population-based organized cervical cancer screening pilot program in France[J]. Eur J Cancer Prev. 2018;27(5):486–92.
Majed L, Bresse X, El Mouaddin N, et al. Public health impact and cost-effectiveness of a nine-valent gender-neutral HPV vaccination program in France[J]. Vaccine. 2021;39(2):438–46.
Taneja N, Chawla B, Awasthi AA, et al. Knowledge, Attitude, and Practice on Cervical Cancer and Screening Among Women in India: A Review[J]. Cancer Control. 2021;28:10732748211010799.
Stelzle D, Tanaka LF, Lee KK, et al. Estimates of the global burden of cervical cancer associated with HIV[J]. Lancet Glob Health. 2021;9(2):e161–9.
Choudhry V, Ambresin AE, Nyakato VN, et al. Transactional sex and HIV risks - evidence from a cross-sectional national survey among young people in Uganda[J]. Glob Health Action. 2015;8:27249.
Hall HI, Brooks JT, Mermin J. Can the United States achieve 90-90-90?[J]. Curr Opin HIV AIDS. 2019;14(6):464–70.
Lin S, Gao K, Gu S, et al. Worldwide trends in cervical cancer incidence and mortality, with predictions for the next 15 years[J]. Cancer. 2021;127(21):4030–9.

No competing interests reported.

supplementmaterials.docx

Attribute risk burden and prediction analysis of cervical cancer under different sociodemographic transitions worldwide from 1990 to 2034

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

1. Introduction

2. Method

2.1 Data sources

2.2 Statistical analysis

3. Results

3.1 Temporal trends in different SDI levels

3.2 The age-specific mortality and incidence rates by age group

3.3 The proportion of YLL in DALY

3.4 The relationship between SDI and ASMR or ASDR of cervical cancer

3.5 Predictions of ASMR for cervical cancer in different SDI countries from 2020 to 2034

4. Discussion

5. Conclusion

Declarations