Quantifying Cancer Burden Attributable to Obesity: Highlighting the Disparities by Sex, Race, and Ethnicity in a Rural State with High Obesity and Cancer Burden

Abstract

Overweight and obesity is a complex, multifactorial disease that increases the risk of several cancers. The purpose of this study is to calculate the population-attributable fraction (PAF%) of obesity-associated cancer rates among adults from 2010–2019 in Arkansas by sex, race, and ethnicity. Obesity-associated cancer data for this period were obtained from the Arkansas Central Cancer Registry. The PAF% was calculated using obesity prevalence and global relative risks. Obesity prevalence data were gathered from the Arkansas Behavioral Risk Factor Surveillance System. Global relative risks for each obesity-associated cancer were gathered from large-scale epidemiological studies, meta-analyses, and systematic reviews in which body mass index (BMI) was ≥ 30 kg/m². Obesity-attributable cancer age-adjusted incidence rates (AAIRs) were calculated by multiplying the obesity-associated cancer's AAIR by the estimated PAF%. Breast, esophageal adenocarcinoma, gallbladder, kidney, and liver obesity-associated cancers each had a PAF% greater than 25% by sex, race, and ethnicity. Arkansas non-Hispanic (NH) Black women were disproportionately impacted by obesity-attributable cancers, with a disparity driven primarily by the higher incidence of breast and other female-specific cancers. Findings suggest that targeted screening among those with BMI ≥ 30 kg/m² could decrease the burden of obesity-associated and attributable cancers in Arkansas, particularly for breast and colorectal cancers.

Objective

Obesity is a modifiable risk factor for various chronic diseases.¹ It is a complex, multifactorial disease that increases the risk of several cancers.² The incidence of many of these cancers has been increasing, possibly due to the increasing prevalence of obesity. According to a Centers for Disease Control and Prevention (CDC) study, about 40% of all new cancer diagnoses in the United States are associated with overweight and obesity.³ Of the 13 cancer types associated with obesity, at least 5 (breast, colorectal, corpus uteri, kidney, liver, pancreatic, and thyroid) are among the 10 most commonly diagnosed cancers.⁴ While these observed rising cancer rates and obesity prevalence can only imply an association, continuing increases in obesity prevalence over time also suggest that the burden will likely increase in the decades to come.⁵ Evidence has also shown an increase in the risk of cancer recurrence and mortality among cancer survivors with obesity, underscoring the need for early interventions for patients before, during, and after a cancer diagnosis.^6,7 Data are needed to identify populations for whom the burden is highest in a rural, racially diverse, and high-poverty southern state (Arkansas) with a high obesity and cancer burden.⁸ Even though cancers associated with obesity quantify the relationship between exposure and disease (obesity-associated cancer), this association does not provide an estimate for cancer cases that are likely due to obesity (obesity-attributable cancers). Knowing the burden of cancers attributable to obesity can help clinicians assess the impact of obesity on cancer.⁹ These results may help community outreach programs to stress the importance of obesity and cancer prevention efforts.

For this study, data were used from the Arkansas Central Cancer Registry, a population-based registry that identifies cancer trends and rates at a state level. By using central cancer registry data, this study describes the burden of obesity-associated cancers and estimates attributable rates utilizing clinically reported cancer cases. Understanding the burden of cancer attributable to obesity is crucial, especially for Arkansas, due to its racial and ethnic diversity and its complex socioeconomic elements, which include poverty, rurality, inadequate nutrition, the limited physical activity of its population, and their limited access to quality healthcare.^10,11 The purpose of this study is to calculate obesity-attributable cancer rates derived from the population attributable fraction of adults from 2010–2019 in Arkansas by sex, race, and ethnicity.

Methods

Obesity-Associated Cancer Rates

This cross-sectional study used data obtained from the Arkansas Central Cancer Registry for adult patients (≥ 18 years of age). Data were evaluated by sex and 2 major racial groups (non-Hispanic [NH] Black and White). Age-adjusted incidence rates (AAIR) for obesity-associated cancers were computed using SEER*Stat software version 8.4.3 with corresponding 95% CIs calculated as modified gamma intervals on a central cancer registry imported dataset.¹² Cancers were identified according to the CDC's definitions of obesity-associated cancers, which is a predefined selection available in SEER*Stat.¹³ Cancers associated with obesity are defined by cancer type using an ICD-O-3 primary site and histology codes. Additional restrictions are applied to certain cancers: Esophageal adenocarcinoma (EAC) is restricted to cases microscopically confirmed, while corpus uteri (not otherwise specified) and ovarian cancer cases are restricted to women. Breast cancer is restricted to women who are over the age of 50 years and postmenopausal. For this analysis, meningioma (for all groups) and esophageal adenocarcinoma for (NH Black women) were excluded from analysis due to low case counts in Arkansas. Because the COVID-19 pandemic impacted cancer reporting for the diagnosis years 2020 and 2021, this study used prepandemic cancer diagnosis years 2010–2019. This study was determined to be exempt from Institutional Review Board review as it posed minimal risk to human subjects.

Population Attributable Fraction

The population attributable fraction (PAF) was calculated using published relative risks for each obesity-associated cancer and obesity prevalence by sex, race, and ethnicity group. Obesity prevalence for each group was obtained from the 2011 Arkansas Behavioral Risk Factor Surveillance System (BRFSS), as 2011 is the first year that the current BRFSS weighting methodology was implemented. Global relative risks for each obesity-associated cancer were gathered from large-scale epidemiological studies, metaanalyses, and systematic reviews where body mass index (BMI) was ≥ 30 kg/m² (Table 1). To calculate the obesity-attributable cancer rate, the AAIR of obesity-associated cancers was multiplied by the PAF (%)⁹:

Table 1.

Relative Risk for Obesity-Associated Cancers Gathered from Meta-analysis, Systematic Review, and Large-Scale Epidemiological Studies Where Obesity Has ≥ 30 Body Mass Index Classification

Obesity-Associated Cancer Type	Relative Risk (95% CI)
Breast, postmenopausal*	1.13 (1.05–1.22)1
Colorectal	1.33 (1.25–1.42)2
Corpus uteri*	3.22 (2.91–3.56)1
Esophageal adenocarcinoma	3.29 (1.82–5.95)3
Gallbladder	1.58 (1.43–1.75)4
Gastric cardia	1.13 (1.03–1.24)5
Kidney	1.76 (1.61–1.91)6
Liver	1.77 (1.56–2.01)7
Multiple myeloma	1.23 (0.99–1.52)8
Ovary*	1.28 (1.20–1.36)1
Pancreas	1.47 (1.23–1.75)9
Thyroid	1.09 (0.98–1.22)10

^*Female-specific.

Results

Obesity-Associated Cancer Rates

In Arkansas, 58,698 obesity-associated cancers were diagnosed from 2010–2019. Overall, NH Black women had the highest rate of obesity-associated cancers (292.5 cases per 100,000 population). Among female-specific cancers, postmenopausal breast cancer had the highest incidence rate (NH Black women, 317.3 cases per 100,000 population; NH White women, 319.3 cases per 100,000 population). Among cancers that affect all groups, colorectal cancer had the highest frequency of cancer, with NH Black adults having the highest incidence rate (NH Black men = 73.9 cases per 100,000 population; NH Black women = 60.7 cases per 100,000 population) (Table 2).

Table 2.

Number and Age-Adjusted Incidence Rates (AAIRs) of Obesity-Associated Cancers by Race/Ethnicity and Sex, Arkansas, 2010–2019

	Women				Men
	Non-Hispanic Black		Non-Hispanic White		Non-Hispanic Black		Non-Hispanic White
	Count	AAIR (95% CI)	Count	AAIR (95% CI)	Count	AAIR (95% CI)	Count	AAIR (95% CI)
Overall	5,624	292.5 (284.7–300.5)	34,544	266.0 (263.1–268.9)	2,415	166.0 (158.9–173.2)	16,115	143.7 (141.4–146.0)
Cancer Type
Breast, postmenopausal*	2,233	317.3 (303.9–331.2)	15,490	319.3 (314.3–324.5)	-	-	-	-
Colorectal	1,118	60.7 (57.1–64.5)	5,927	45.7 (44.5–46.9)	1,070	73.9 (69.2–78.8)	6,556	59.0 (57.5–60.5)
Corpus uteri*	616	31.1 (28.6–33.7)	3,753	29.9 (29.0–31.0)	-	-	-	-
Esophageal adenocarcinoma	†	†	111	0.9 (0.7–1.0)	25	1.6 (1.0–2.4)	789	6.8 (6.3–7.3)
Gallbladder	35	1.9 (1.3–2.6)	185	1.4 (1.2–1.7)	16	1.2 (0.6–2.0)	85	0.7 (0.6–0.9)
Gastric cardia	19	1.0 (0.6–1.7)	132	1.0 (0.8–1.2)	40	2.7 (1.9–3.8)	519	4.5 (4.2–5.0)
Kidney	370	19.9 (17.9–22.1)	2,056	16.8 (16.0–17.5)	444	29.9 (27.0–33.0)	3,251	29.2 (28.2–30.3)
Liver	79	4.1 (3.3–5.2)	499	3.7 (3.4–4.1)	210	13.0 (11.2–15.1)	1,298	10.9 (10.3–11.6)
Multiple myeloma	289	15.2 (13.5–17.1)	760	5.7 (5.3–6.1)	253	18.4 (16.0–21.0)	945	8.4 (7.8–8.9)
Ovary*	206	11.1 (9.6–12.7)	1,681	13.8 (13.1–14.5)	-	-	-	-
Pancreas	399	22.3 (20.1–24.6)	1,662	12.2 (11.6–12.8)	295	21.3 (18.7–24.0)	1,887	16.6 (15.9–17.4)
Thyroid	256	13.4 (11.8–15.2)	2,273	22.4 (21.5–23.4)	62	4.1 (3.1–5.3)	767	7.4 (6.8–7.9)

^*Female only.

^†Excluded due to low counts.

Population Attributable Fraction (PAF) Estimates

The obesity crude prevalence in Arkansas was 59.9% (95% CI: 49.0–70.8) for NH Black women, 48.7% (95% CI, 44.2–53.3) for NH White women, 40.10% (95% CI, 29.2–51.0) for NH Black men, and 51.3% (95% CI, 46.7–55.8) for NH White men. Breast, esophageal adenocarcinoma, gallbladder, kidney, and liver obesity-associated cancers each accounted for at least one group with a PAF greater than 25%. Among female-specific cancers, NH Black women had a higher PAF than NH White women for obesity-attributable breast, corpus uteri, and ovarian cancer. Excluding NH Black women, the PAF for esophageal adenocarcinoma attributable to obesity among all groups ranged from approximately 47.9–54.0% (Table 3).

Table 3.

Estimated Population Attributable Fraction (PAF) (%) for Sex, Race, and Ethnicity by Cancer Type, Arkansas

	Women		Men
	Non-Hispanic Black, PAF%	Non-Hispanic White, PAF%	Non-Hispanic Black, PAF%	Non-Hispanic White, PAF%
Obesity-associated cancer type
Breast, postmenopausal*	11.64%	9.68%	-	-
Colorectal	16.67%	13.99%	11.81%	14.63%
Corpus uteri*	60.53%	55.49%	-	-
Esophageal adenocarcinoma	†	52.72%	47.87%	54.02%
Gallbladder	25.78%	22.02%	18.87%	22.93%
Gastric cardia	7.22%	5.95%	4.95%	6.25%
Kidney	31.28%	27.01%	23.36%	28.05%
Liver	31.56%	27.27%	23.59%	28.32%
Multiple myeloma	12.11%	10.07%	8.44%	10.55%
Ovary*	17.74%	14.92%	-	-
Pancreas	21.97%	18.63%	15.86%	19.43%
Thyroid	5.12%	4.20%	3.48%	4.41%

^*Female only

^†Excluded due to low counts

Estimated Obesity-Attributable Cancer Rates

Overall, NH Black women had a higher obesity-attributable cancer rate for breast cancer (36.9 cases per 100,000 population), colorectal cancer (10.1 cases per 100,000 population), cancer of the corpus uteri (18.8 cases per 100,000 population), gallbladder cancer (0.5 cases per 100,000 population), multiple myeloma (1.8 cases per 100,000 population), and pancreatic cancer (4.9 cases per 100,000 population). Among cancers that affect both men and women, NH White men had a higher obesity-attributable cancer rate for esophageal adenocarcinoma (3.67 cases per 100,000 population), gastric cardia (0.28 cases per 100,000 population), and kidney cancer (8.19 cases per 100,000 population) (Figure 1).

Figure 1.

Estimated obesity-attributable cancers AAIR by sex, race and ethnicity, Arkansas, 2010–2019

Discussion

Overall, publicly available data from 2010–2019 show that Arkansas and the United States have a similar obesity-associated rate of approximately 172 cancer cases per 100,000 population.⁸ Moreover, cancer rates in Arkansas closely mirror national patterns by race, ethnicity, and sex. For example, NH Black women had the highest obesity-associated cancer compared to all other major groups, though one group, NH White women, whose rate is higher in the United States overall than in Arkansas, showed slightly lower rates. Utilizing population-based cancer registry data made it possible to quantify Arkansas-specific obesity prevalence and to describe race and ethnicity-specific PAF by sex and cancer type. NH Black women are disproportionately impacted by obesity-attributable cancers, a disparity driven primarily by their higher incidence of breast and other female-specific cancers. Nevertheless, disparities by sex, race, and ethnicity were not statistically different across all types of cancer. Moreover, NH White men had a significantly higher rate of esophageal cancer. Greater use of smokeless tobacco products, especially among rural NH White men, may be one factor contributing to this disparity.¹⁴

This study has several implications for Arkansas. First, obesity is a well-established risk factor for both breast and colorectal cancer. Research shows that obesity increases the risk of breast cancer recurrence among patients and can negatively impact a survivor's quality of life.^15–17 Patients with obesity were more likely to be diagnosed with late-stage colorectal cancer, with a potential poor prognosis, and obesity has been linked to a 14% increase in colorectal-cancer-specific and all-cause mortality rates.¹⁸ Timely screenings are recommended for breast and colorectal cancer in Arkansas. Although Arkansas ranks 31 in breast and 41 in colorectal cancer screening compared to other US states, evidence-based efforts are needed to increase screenings in the state, especially for at-risk populations with high rates of obesity.^19–21

Second, current knowledge elucidates proposed biological mechanisms for an obesity-to-cancer progression.^1,22 For example, adipose tissue functions as an organ, releasing chemical mediators and enzymes, which can lead to excess production of estrogen. This excess production has been associated with a higher risk of developing female-specific cancers, including breast, endometrial, and ovarian cancer.^23–26 Another example is the increase of insulin and insulin-like growth factor 1 (IGF-1). High levels of insulin and IGF-1 are commonly seen among obese individuals and may contribute to the development of associated cancers, such as colon, renal, and endometrial cancer.^23,27,28 Finally, by weakening tumor immunity and altering the mechanical properties of the tissue surrounding growing tumors, obesity may increase cancer risk. Obesity has been linked with metastasis through various factors, such as adipokines, immune cell modulation, systemic inflammation, angiogenesis, metabolic changes, extracellular matrix alterations, and extracellular vesicles.^29,30 Considering Arkansas' statistically significant increase in adult obesity, there is a concerning association with obesity contributing to cancer-specific tumor development, growth, recurrence, and survival.^6,7,31–33 It is likely that the development of cancer in relation to obesity involves a variety of mechanisms; more research is needed to further understand them.

Third, weight management and interventions continue to be important not only in the prevention of cancer and reduction of overall mortality in cancer survivors, but also in comorbid illnesses and other chronic diseases.^6,34–36 Thus, it is important to address risk factors that lead to obesity.^37,38 The findings in this study suggest the continued need for obesity prevention initiatives at the state level that promote physical activity and access to preventive care. For example, it has been shown that, among type 2 diabetes patients, recently approved weight loss drugs (ie, glucagon-like peptide 1 receptor agonists [GLP-IRAs]) decrease the risk of certain obesity-associated cancers compared to insulin and metformin. However, lack of insurance coverage and high cost can hinder low-income, uninsured adults from utilizing GLP-IRAs.^39,40

Moreover, evidence shows that combining structured exercise with dietary support for weight loss leads to greater weight loss than either exercise or diet alone.^41–45 This approach also has the most significant effect on blood biomarkers associated with common cancers, including insulin resistance, circulating levels of sex hormones, leptin, and inflammatory markers.^41–45 Considering that approximately 44.9% of Arkansans do not meet the recommended physical activity guideline and that 18.9% are food insecure, lack of access to safe and convenient places for physical activity and high-quality groceries may be contributing to obesity rates and disparities.^46–49 As not all populations have the same availability to resources for sustainable weight loss or healthy weight maintenance, interventions may require a culturally tailored, multifaceted approach to improve access. Future work is needed to understand the variation in state- and demographic-specific cancers associated with obesity and to create focused approaches to decrease population obesity through regular physical activity and access to healthy foods in communities.

Limitations

Although the study has many strengths, it has at least three limitations. First, anthropometric data on weights and heights are self-reported in BRFSS, resulting in potential underreporting of obesity prevalence.⁵⁰ As a result, the burden of obesity-associated and attributable cancer in Arkansas is likely higher. Furthermore, BMI is commonly used as a surrogate measure for obesity but does not consider the varying distribution of excess adipose tissue in the body.⁵¹ While there are studies using waist circumference and waist-hip ratio measures as better predictors of cancer risk, most research has utilized BMI due to its low cost and ease of measurement.¹ Second, the data assembled were not adequate to identify disparate age-adjusted incidence rates among all minoritized populations within the state, specifically excluding the state's growing Hispanic population and Marshallese community. Nor was data adequate to distinguish disparities across the urban-rural divide. Third, compared to Islami, et al., this study's findings showed noticeable discrepancies in PAF, which may be due to differences in the methodological approach and data sources.⁵ Notably, colorectal, corpus uteri, esophageal adenocarcinoma, and ovarian cancer had a higher PAF, while the PAF was lower for gastric cardia and thyroid cancer. In general, states with lower cancer incidence rates that also have high rates of obesity have a higher PAF compared to other states, which may be the case for Arkansas.⁵

Conclusions

In conclusion, this study calculated a high obesity-attributable rate for cancers among women, especially among NH Black women for female-specific cancers. Men experienced a high obesity-attributable rate for kidney and liver cancer. Regular breast and colorectal screenings are suggested for individuals with obesity. Research also suggests that obesity is a contributing factor for cancer rates, which is a concern for Arkansas as a state that has consistently ranked among the highest in the nation in obesity prevalence. Addressing obesity in Arkansas through public health initiatives, such as access to healthy foods, access to physical activity opportunities, and education could play a crucial role in reducing the state's cancer burden.