Trends in the burden of most common obesity‐related cancers in 16 Southern Africa development community countries, 1990–2019. Findings from the global burden of disease study

Abstract Background Obesity‐related cancers in the 16 Southern African Development Community (SADC) countries is quite prominent. The changes and time trends of the burden of obesity‐related cancers in developing countries like SADC remain largely unknown. A descriptive epidemiological analysis was conducted to assess the burden of obesity‐related cancers, (liver, esophageal, breast, prostate, colon/rectal, leukemia, ovarian, uterine, pancreatic, kidney, gallbladder/biliary tract, and thyroid cancers) in SADC countries. Methods Data from the 2019 Global Burden of Diseases Study was used. Deaths extracted from vital registration, verbal autopsies and ICD codes. Cancer‐type, mortality and prevalence per 100,000 population and 95% uncertainty intervals (UIs) were calculated using the Cause of Death Ensemble model and Spatio‐Temporal Gaussian process with mixed effects regression models. Annual rates of change (AROCs) between 1990 and 2019 and the corresponding UIs were calculated. Results The top age‐standardized mortality rates per 100,000 in 2019 for males were leukemia, 20.1(14.4–26.4), esophageal cancer, 15.1 (11.2–19.1), and colon and rectal cancer, 10.3 (8.6–12.6). For females, breast cancer, 20.6 (16.6–25.0), leukemia, 17.1 (11.4–23.7), and esophageal cancer, 8.3 (5.5–10.7), had the leading mortality rates. For males, AROC substantial (p < 0.05) increase for kidney cancer for 11 of the countries (AROC from 0.41% to 1.24%), colon cancer for eight of the countries (from 0.39% to 0.92%), and pancreatic cancer for seven countries (from 0.26% to 1.01%). In females, AROC showed substantial increase for pancreatic cancer for 13 of the countries from (0.34%–1.67%), nine countries for kidney cancer (from 0.27% to 1.02%), seven countries each for breast cancer (0.35%–1.13%), and ovarian cancer (from 0.33% to 1.21%). Conclusions There is need for location‐specific and culturally appropriate strategies for better nutrition and weight control, and improved screening for all cancers. Health promotion messaging should target kidney, colon, pancreatic, and breast cancers and encourage clinically tested methods of reducing BMI such as increasing personal physical activity and adoption of effective dietary regimes.


| BACKGROUND
It has been projected that by 2030, 70% of cancer deaths may occur in developing countries, and that these countries are expected to bear the bulk of the projected 24.1 million new cases annually. 1,24][5][6][7] A meta-analysis of 57 prospective studies with 900,000 adults with over 6.5 million person-years found that mortality was lowest for individuals with BMI within the normal range (22.5-25.0kg/m 2 ), and particularly for women. 8Each increase in BMI of 5 kg/m 2 conferred a 29% increase in the hazard for allcause mortality, hazard ratio [HR] 1.29 (95% confidence interval [CI]: 1.27-1.32),and a 10% increase in the hazard for cancer mortality, 1.10, (1.06-1.15). 8causal relationship has been established between overweight (BMI ≥30 kgs/m 2 ) and obesity (BMI: 25-<30 kg/m 2 ) and cancer, 9 but the role of increasing overweight and/or obesity on the trends of high-BMI-related cancers, particularly in developing countries, remains uncertain.Adults with obesity were reported to have a higher risk of cancer than those with a healthy weight in developed countries like the US. 6,7Globally, an annual 0.6% increase was observed for the age-standardized mortality rate and the burden of cancer attributable to obesity was heavier in regions with higher Socio-Demographic Index (SDI) levels. 10The rate of incident cancers not associated with overweight and obesity decreased by 13% 6,7 while age-adjusted rates of incident cancers associated with obesity [oesophagus, breast (in postmenopausal women), colon/rectal cancer; uterus, gallbladder, kidney, liver, ovary, pancreas, and thyroid cancers] in the US increased by 7% for the period 2005 to 2014.
Currently, there is no such comparable data for many developing countries, especially those of the SADC.Goal 3.4 of the United Nations General Assembly 2030 agenda for Sustainable Development Goals (SDGs) aims to reduce by one third premature mortality from non-communicable diseases (NCDs) with indicator 3.4.1 specifically seeking to reduce by one third the mortality rate attributed to conditions that are known to be causally associated with high BMI.
[15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30] Meta-analyses relative risks (RR) of the association ranged from 1.2 to 1.5 for overweight, and from 1.5 to 1.8 for obesity for cancers of the colon/rectal, 13,31 gastric cardia, 32 liver, 24 gallbladder, 33 pancreas 34 and kidney cancers. 21The RR for oesophageal adenocarcinoma was as high as 4.8 for individuals with a BMI of 40 kg/m 2 or more. 35However, 18 studies of the Asia Cohort Consortium found that the hazard ratios for oesophageal cancer and BMI formed a wide J-shaped association indicating mortality risk increase for underweight (BMI < 18.5 kg/m 2 : HR = 2.20, 95% CI = 1.80-2.70)and extreme obesity (BMI ≥ 35 kg/m 2 : 4.38, 2.25-8.52)relative to the referent BMI category of (23-25 kg/m 2 ). 36Other studies show substantial evidence of the positive association between increased BMI near the time of cancer diagnosis and reduced survival in patients with breast cancer [37][38][39] ; however, recipients of bariatric surgery for intentional weight loss reported reductions in cancer incidence and mortality indicating a negative association between cancers and low BMI. 402][43] The 16 countries that form the Southern Africa Development Community (SADC) shown in Figure 1  which breast (1.67 million), and colorectal (1.36 million) were among the top three most commonly diagnosed, while liver (745,000 deaths), and stomach cancers (723,000 deaths) were among the top three most common causes of cancer death. 44Projection estimates from GLOBOCAN 2020 Study indicated that of the 34 cancer types studied, 1.1 million new cases (95% UIs: 1.0 -1.3 million) and 711,429 (611,604 -827,547) deaths due to neoplasms occurred in Africa in 2020, and that by 2040, the burden of all neoplasms combined is expected to increase to 2.1 million new cases and 1.4 million deaths in Africa alone. 45In 2019, for all countries of Africa, the Institute for Health Metrics and Evaluation (IHME) estimated a total of 71,708 breast cancer deaths, 48,639.97deaths for colon/rectal cancer, 37,208 for liver cancer, 35,584.86 for oesophageal cancer 24,507 for pancreatic cancer, 17,624 for bladder cancer, 13,041 for ovarian cancer, 7366 for gallbladder and biliary tract cancer, 7201 for kidney cancer, 5901 for uterine cancer corresponding to agestandardized mortality rates (95% uncertainty interval (UI)) ranging from 10.89 (9.53-12.34)per 100,000 population for breast cancer to 1.02 (0.80-1.22) for uterine cancer. 46r previous report revealed that 47 age-standardized prevalence of overweight in adult females in SADC countries increased by 8.3% over a 30-year period from 31.4% (30.5-32.3) in 1990 to 39.7% (38.7-40.7) in 2019; and increased in males by 8.5% from 20.2% (19.5-20.8) to 28.7% (27.9-29.5). 48Obesity in adult females increased by >1.5-fold and nearly doubled in adult males.Despite these increasing trends in obesity and overweight, compounded by the well-established causal relationship between obesity and several cancers, there is no known in-depth analysis of trends in morbidity and mortality associated with high-BMI-related cancers for SADC countries.As a result, a descriptive epidemiological analysis of mortality and prevalence associated with high-BMI-related cancers for SADC countries was conducted using data from the Global Burden of Diseases (GBD), Injuries and Risk Factor Study. 47This information will aid key stakeholders to track progress and identify priorities for resource investment and/or develop corrective interventions to mitigate the interplay between obesity and related cancers.

| Data sources for cancer mortality
A description of GBD data sources and processing steps for the cause of death database, case definitions, input data for morbidity and mortality and modeling strategies for each cancer type are provided on pages 805 to 819 of Supplementary appendix 1 for GBD 2020 article. 47

| Input data
GBD obtained data from vital registration, verbal autopsies, and International Classification of Diseases (ICD) codes for the years 1990 and 2019. 49Cause of Death Ensemble model (CODEm) and Spatio-Temporal Gaussian regression was used to estimate mortality due to individual cancer types.All ICD-9 codes for cancer (140-209), and all ICD-10 codes (C00-C96, except for Kaposi's sarcoma (ICD-10: C46) were included in the estimates for "malignant neoplasms", specific details are described elsewhere. 50

| Adjustment for country-specific covariates
The CODEm modeling strategy for each cancer type was weighted for country-specific covariates such as healthcare access and quality index (HAQI), education years per capita, SDI, 10 age-and sex-specific summary exposure variable for alcohol use, etc.A full description of coding of covariates and their influences can be found on pages 195 to 213 of Supplementary appendix 1 for GBD 2020 article 47

| Change in mortality rates over time
For each cancer type the slopes of the mortality rate from 1990 to 2019 were assessed using an annualized age-standardized rate of change (AROC) as the percent difference in the natural logarithm of the rate in 1990 and 2019 divided by 30 (i.e., 100*[ln (2019 Rate/ 1990 Rate)/30]).AROC (%) represent a measure of a trend (increasing, decreasing or flat) over 30 years.A positive AROC indicates an increasing trend/slope or acceleration of mortality over the 30 years, and a negative AROC indicates decreasing mortality rate.To put negative AROC into perspective, two locations with different negative AROCs (e.g., −1.0% and −2.0%) indicates that the decline in mortality rate is less robust in the location with the lesser absolute AROC (−1.0%) than the decline in the location with the larger absolute AROC (−2.0%).

| Uncertainty analysis
Uncertainty for each outcome was quantified using 95% uncertainty intervals (UIs) based on 1000 bootstrap draws from the posterior distribution. 52,53UIs were determined by the 25th and 975th ordered values of the posterior distribution of the 1000 draws, and point estimates were computed from the mean.Changes over time were considered statistically significant when the 95% UI of the percentage change did not include zero.

| Reporting guidelines
The study did not require Institutional Review Board ethical review or informed consent as it used public GBD results and data.The GBD study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines (Elm E v 2007).
In addition, all GBD estimates adhere to the 14 Guidelines on Accurate and Transparent Health Estimate Reporting (GATHER).GATHER recommends making available statistical code, why some sources are used and others are not, and how primary data are adjusted. 54

| Role of the funder
The funders of this study had no role in the study design, data collection, data analysis, data interpretation, or the writing of the report.The corresponding author (PNG) had full access to the data in the study and final responsibility for the decision to submit for publication.

| RESULTS
Our recent report found that obesity in adult females increased >1.5fold from 12.0% (95% UI: 11.5-12.4) in 1990 to 18  (23.8-25.8) in adult females. 45Results presented in this paper should be considered in the above context.

| Mortality rates
Table 1 shows age-standardized mortality rates per 100,000 population due to the six leading cancers studied here (breast, oesophageal, colon/rectal, liver, prostate, and leukaemia).(18.1-41.8).Relative to the SADC, the 2019 female breast cancer mortality rate was 1.44-fold higher than 1990 for Namibia, 1.39-fold higher for Botswana, and 1.38-fold higher for Lesotho.

| Oesophageal cancer
The second most common type of cancer among males was oesophageal cancer with mortality rate (95% UI) at 17.5 (11.4-22.1)

| Age standardized annualized rate of change for mortality
Table 4 shows age-standardized prevalence per 100,000 population for five leading high-BMI-related cancers (breast, colon/rectal, oesophageal, prostate, and liver) for the period 1990 and 2019.

| DISCUSSION
This study was heavily guided and inspired by the serious lack of information on trends in morbidity and mortality associated with high-BMI-related cancers in SADC countries.It was also inspired by our report of 2021. 48To our knowledge, this study represents the most comprehensive study so far that has reported the estimates of the burden of cancer due to obesity in SADC countries.Escalation of such diseases and other NCDs in the SADC countries requires information to highlight the impact and changes of high-BMI-related cancer burden on the populations for each of the countries to inform strategies and guide interventions to reduce the burden.The generally negative AROCs in males and females for oesophageal and leukaemia cancers provides evidence that mortality for these cancers is trending in the right direction but at a pace not rapid enough to meet the WHO UN SDG Target 3.4.
6][57] The other reason, in part, is the aging populations.The demographic transition theory suggests that as populations live longer and age substantially, there seems to be a correlated increase to lifetime risks for cancer.While the demographic transition proposes that neoplasms occur mostly in middle and old age, it is now known that they may not primarily or necessarily be caused by age-related biological processes of developing countries. 58The lifetime probability of developing cancer and cancer-related mortality in developed countries has already reached 56.9% and 27.6% in men, and 51.9% and 21.7% in women respectively. 59African countries are at different stages of the demographic transitions, and even in-country transitions are at different levels that is, transitions are not monolithic, therefore interventions should be targeted to specific demographic patterns for each country and within country by region, such as rural/urban and other country-relevant subnational demarcations.
A major risk factor in developing cancer is obesity and/or high BMI.Generally, obesity is rising in developing countries; and the SADC is no exception.Among the many known and unknown reasons for this rise includes unhealthy diets and lack of preventive practices at both personal and policy levels 60 combined with rising disposable income. 61Observed obesity increase in Africa is largely caused by increasing urbanization, 62 high changes in diet composition such as readily available and relatively cheaper fast foods confounded by dialogs comparing healthier foods as being harder to access and costly 63 reduced physical exercise, and sedentary life styles. 49SADC countries should, as a matter of urgency start funding and developing public health and behavioral research interventions targeting these unhealthy behaviors with a focus on diet quality interrogations, caloric intake and physical activity, and the effect of rapid urbanization on childhood obesity and thereby suggest solutions as a way forward. 64art from putting functional policies in place to effectively halt and reverse the occurrence of high-BMI-related cancers and mortality, African governments, especially in the SADC require investments in prevention and early detection and cancer care services.Currently, although the demand for cancer care has risen sharply, cancer care services are still limited and therefore, access to such is extremely difficult for many low-income cancer patients.For example, overweight and obesity were strongly associated with the later stages of breast cancer at diagnosis (stage III, IV) in Egypt, 65 however and perhaps due to lack of screening awareness among women with obesity, breast cancers were discovered quite late and when irreversible.Breast cancer survival for most cases is driven by early diagnosis.However, cancer survival rates are worse in the African population than in developed countries, with the 5-year survival rate of women with breast cancer in Europe being 82%, while it ranges between 12% and 46% in parts of Africa. 66The call and need for SADC countries to invest in widespread cancer screening and early diagnosis and treatment together with preventive efforts at primary care facilities cannot be overstated.3) 2.3 (1.5-3.4)0.4 (0.3-0.5) 0.7 (0.5-1.0) 1.2 (0.7-1.7) 1.4 (0.9-2.0) 0.9 (0.6-1.2) 1.1 (0.7-1.6) Namibia 2.6 (1.9-3.6)3.7 (2.6-5.2) 2.2 (1.6-3.0)2.3 (1.6-3.2) 1.1 (0.9-1.4) 2.6 (1.9-3.6)0.9 (0.7-1.1) 1.3 (0.9-1.8) 0.8 (0.6-1.1) 0.8 (0.5-1.0) 0.5 (0.4-0.7) 0.5 (0.3-0.7) Seychelles 5.5 (4.5-7.9)8.9 (7.0-10.8)2.9 (2.4-3.7)3.0 (2.4-3.8)2.5 (2.2-2.9)3.8 (3.2-4.5)0.8 (0.7-0.9) 1.0 (0.8-1.2) 1.5 (1.1-1.8)1.3 (1.0-1.7)0.4 (0.3-0.5) 0.4 (0.3-0.5)South Africa 3.6 (2.9-4.2) 4.7 (3.6-5.7)1.8 (1.4-2.2) 2.3 (1.6-2.7)2.7 (2.2-3.4)3.6 (3.1-4.1)0.8 (0.7-0.9) 0.8 (0.7-0.9) 0.9 (0.7-1.1) 0.9 (0.6-1.0) 0.5 (0.4-0.6) 0.4 (0.4-0.5) Tanzania 3.7 (2.3-5.3)5.3 (4.1-6.6)2.7 (1.6-3.5)2.8 (1.8-3.6)1.5 (1.2-1.8)2.4 (1.9-2.9)0.6 (0.5-0.8) 1.1 (0.9-1.4) 1.3 (1.0-1.7)1.3 (1.0-1.8)0.9 (0.7-1.1) 1.1 (0.8-1.4)The consumption of processed foods in Africa has increased in the last decades, with the downside impact of increased obesity 73 and cancer.SADC regions may require stringent and enforceable regulations to reduce uncontrolled access to ultra-processed foods. 73The influence of ultra-processed foods (both imported and locally made) and poor food legislature, in addition to regulation enforcement may be notable factors driving the burden of obesityrelated cancers.In developed countries, an increase in colorectal cancers have been attributed, in part, to high consumption of ultraprocessed foods. 74The consumption of processed foods has increased in SADC countries.In addition to locally processed foods, substantial increases in importation of poorly regulated processed drinks and foods from developed countries into the SADC region have dramatically increased.This is likely driven by a combination of the public need for cheaper food alternatives, and the need for increased business opportunities for regional and local importers, who create much needed employment and income opportunities. 75cal food quality control has also been found lacking in SADC countries.Despite the well documented carcinogenic and health risk properties of mycotoxins, secondary fungal metabolites known to contaminate major staple foods in these regions, legislation controlling mycotoxin contamination of food is still limited in a number of SADC countries. 76In addition, SADC populations, young and old, will benefit from awareness campaigns on food quality and conscious consumption through food label evaluation, specifically for those foods that may be associated with obesity and cancer.
Specific HIV treatment regiments have been associated with Obesity. 77Considering that sub-Saharan Africa has the highest burden of people living with HIV and who are on anti-retroviral treatment, globally, 78 it is plausible that HIV treatment associated increases in BMI could also be driving increases in BMI-associated cancers in SADC.In addition, significantly improved survival of people living with HIV has resulted in notable increases in aging populations who may be at high risk of developing non-AIDS cancer, 79 and generally, increased risk of developing cancer due to advancing age. 80her (HR = 0.62; CI: 0.49-0.78). 81On account of such overwhelming evidence, it is incumbent upon SADC countries to invest in surgical cancer care services as well.However, location-specific and culturally appropriate non-surgical interventions and strategies for weight loss are the most important since most SADC countries are generally poor and may not afford quick investment in surgical or indeed complex medical procedures.
This study reinforces conclusions of the IHME's Total Cancers report 47 and the GLOBOCAN 2020 analysis of 34 cancer types showing that high cancer mortality rates in Africa demand a holistic approach toward control and management such as increasing cancer awareness through health education, adoption of primary and secondary prevention methods through health promotion, mitigating risk factors, improving cancer infrastructure and robust screening, diagnosis, and timely treatment. 45,82A large portion of the most rapidly rising cancers are avoidable by implementing public health.Such information is critical to help highlight the impact, changes and trends of high-BMI-related cancer burden on the populations for each of the countries to inform strategies and interventions to act to hopefully reduce the burden.
This study has some limitations.Firstly, period and cohort effects were not examined, which ideally would have allowed differentiation of the shift of cancer-related mortality risk by time periods and birth cohorts for each country so as to better appreciate effects of epidemiological and demographic transitions.To adequately capture time trends in mortality for each age-group adjusting for period effects, future analysis should use age-period-cohort models to analyze time trends in the burden of high-BMI-related cancers. 83Secondly, specific population and country characteristics were not available to adjust for in the analyses.Future studies should tease out the intersection of obesity, different types of cancers (especially for breast cancer in females, and leukaemia, oesophageal, colon/rectal cancer in both sexes), and risk factors in the environment such as industrial pollution, air and water pollution as well as climate change.
Other limitations of GBD estimates have been documented elsewhere. 84,85Finally, information on country-specific policies were not available to address cancers, cancer-related care and mitigation.

of 18 -
(Angola, Botswana, Comoros, Democratic Republic of Congo [DRC], Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, Eswatini [formerly Swaziland], Tanzania, Zambia, and Zimbabwe) are no exception and they generally lack important information regarding the morbidity and mortality trends due to high-BMI-related cancers.As yet, there has not been any systematic understanding in SADC countries of the distribution of BMI related cancers and time trend analysis of the same.On the global scale, the International Agency for Research on Cancer's Global Cancer Observatory (GLOBOCAN) series estimated that there were 14.1 million new cancer cases and 8.2 million deaths worldwide in 2012 of 2 GONA ET AL.
All data from GBD are anonymized and can be accessed on the website of the Institute for Health Metrics and Evaluation 47 at the University of Washington Seattle.The GBD study uses F I G U R E 1 Map of Southern African development community countries.Source: http://ontheworldmap.com/africa/map-of-southern-africa.jpg.deidentified, aggregated data and a waiver of informed consent approved by the University of Washington Institutional Review Board.
Studies providing national or sub-national representative estimates for each type of cancer systematically searched on Medline are shown on Page 11 of the appendix, 2019.Search terms, selection criteria, and flow diagrams of screening and other details are provided elsewhere. 50,51However, for illustration purposes, here is an example of the search description: to find the proportion of liver cancer cases (example) the following search string was used: "("liver neoplasms" [All Fields] OR "HCC" [All Fields] OR "liver cancer" [All Fields] OR "Carcinoma, Hepatocellular" [Mesh]) AND (("hepatitis B" [All Fields] OR "Hepatitis B" [Mesh] OR "Hepatitis B virus" [Mesh] OR "Hepatitis B Antibodies" [Mesh] OR "Hepatitis B Antigens" [Mesh]) OR ("hepatitis C"[All Fields] OR "Hepatitis C"[Mesh] OR "hepatitis C antibodies"[MESH] OR "Hepatitis C Antigens"[Mesh] OR "Hepacivirus"[Mesh]) OR ("alcohol"[All Fields] OR "Alcohol Drinking"[Mesh] OR "Alcohol-Related Disorders" [Mesh] OR "Alcoholism"[Mesh] OR "Alcohol-Induced Disorders"[Mesh])) NOT (animals [MeSH] NOT humans[MeSH])".

and
Lesotho were 8.1 and 5.8 times, respectively relative to the combined SADC rate of 5.2 (4.6-5.8).In females, the 2019 liver cancer mortality rate for Zimbabwe was 4.5 times the combined SADC rate of 12.6 (8.7-18.0).
age-standardized prevalence per 100,000 population for ovarian, uterine, pancreatic, kidney, gallbladder/biliary T A B L E 1 (Continued) Abbreviations: BMI, body mass index (kg/m 2 ); DRC, Democratic Republic of the Congo.T A B L E 2 Age-standardized annual rates of change (AROC) % for obesity-related cancers in Southern Africa development community countries, 1990 and 2019.