Abstract
Background:
The prevalence of type 2 diabetes (T2D) is higher in Black than white Americans, and individuals with T2D have an increased cancer risk. We investigated the association of T2D with the risk of all cancer combined and 21 site-specific cancers, among predominantly low-income participants who experienced a disproportionately high risk of both T2D and cancer.
Materials and Methods:
The study included 76,121 participants (mean age: 52.0 years; 67.2% Black) from the Southern Community Cohort Study (SCCS). T2D was ascertained at the baseline survey. Incident cancer was ascertained via linkage to state cancer registries. Cox proportional hazard models were used to estimate the associations between T2D and cancer after adjusting for confounders.
Results:
Among participants, 21.2% (N=16,137) had baseline T2D, and 9.7% (N=7,376) were diagnosed with incident cancer. Compared to individuals without T2D, individuals with T2D had a significantly elevated risk of all cancer combined (HR: 1.07; 95% CI: 1.01–1.13) and several site-specific cancers, including cancers of the stomach, colorectum, pancreas, liver/intrahepatic bile duct, kidney, and renal pelvis as well as leukemia. The significant association for most cancers was largely observed within 15 years after T2D diagnosis, except for cancer of the pancreas and liver/intrahepatic bile duct, for which elevated risks remain statistically significant 15 to 30 years after T2D diagnosis
Conclusion:
T2D was associated with the risk of overall and certain site-specific cancers in this predominant low-income population.
Impact:
Preventive measures to reduce the burden from T2D could help reduce the risk of overall cancer and several site-specific cancers.
Keywords: diabetes, cancer, incidence, racial disparities, epidemiology
INTRODUCTION
The prevalence of type 2 diabetes (T2D) in the United States is 11.3%, with 37.3 million living with diagnosed T2D and an additional 8.5 million who remain undiagnosed(1). Among individuals who are ≥ 65 years, the prevalence of cancer is 17.3% for those with T2D compared to 16.4% for those without T2D; while among individuals aged 30–64 years, the prevalence of cancer is 7.6% in those with T2D compared to 5.3% for those without(2). Both T2D and cancer share certain modifiable risk factors, such as obesity, lack of physical activity, unhealthy diet, excessive alcohol consumption, and tobacco smoking(3). Previous research conducted among European and Asian populations suggests that individuals with T2D are at an increased risk of developing certain types of cancer(4,5). This could be due to the shared lifestyle risk factors and underlying biological mechanisms(6–10).
Additionally, race is also a factor that could disproportionately affect the T2D-cancer association(3,11). The prevalence of T2D is much higher in Black individuals with cancer compared to White individuals with cancer(12). Since the risk of having T2D is higher for Black individuals than for White individuals, and previous research has suggested that individual poverty could increase the risk of having T2D, there may be an increased risk among Black individuals who have T2D and also experience poverty(6,13–16).
Currently, limited prospective studies have been conducted to evaluate the associations between T2D and cancer incidence in underserved populations who bear a high prevalence of T2D. To bridge the gap in epidemiologic evidence, we conducted a prospective cohort study to investigate the association of T2D with overall cancer as well as 21 site-specific cancers among predominantly low-income Black and White participants in the Southern Community Cohort Study (SCCS).
MATERIALS AND METHODS
Study Population
The details of the SCCS are described elsewhere(17,18). In brief, it is a prospective cohort study of nearly 86,000 adults recruited across 12 states in the southeastern United States (Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia)(19); approximately 86% of them were recruited from community health centers that provide basic health care to low-income populations, and the remaining 14% were recruited from the general population via a mailed survey or a telephone interview. The cohort has a high representation of Black individuals (~65%) and women (~65%), and more than half of the participants have a household income of less than $15,000/year(20). The baseline prevalence of T2D in this cohort is approximately 20%. Demographic, socioeconomic, lifestyle, and anthropometric data, as well as personal medical history, were ascertained at cohort enrollment via standardized computer-assisted in-person interviews for community health center participants and self-administered mailed questionnaires for persons recruited from the general population. Enrollment in the cohort began in March 2002 and concluded in September 2009.
Assessment of T2D and Cancer
T2D was ascertained at the baseline survey by asking study participants whether they had been previously diagnosed by a physician with the condition. Incident cancer was ascertained via linkage to state cancer registries and was classified according to the primary site and histology categories defined by the Surveillance, Epidemiology, and End Results (SEER) Program (SEER Data Reporting Tools [cited 2023 January 23] Available from: https://seer.cancer.gov/siterecode/icdo3_d01272003/index.html). Cancers were classified using the International Classification of Diseases (ICD)-O version 3 (Southern Community Cohort Study ICD-O Codes [cited 2023 January 23] Available from: https://www.southerncommunitystudy.org/uploads/5/2/7/5/52750661/sccs_incident_cancer_11_5_2021.pdf). Only cancer sites with 50 incident cases or more were included in the analysis. The vital status of study participants was determined via linkage with the National Death Index.
Statistical Analysis
We excluded those participants with missing information at the baseline survey for a prior diagnosis of T2D (N=1,513; 1.8%) and cancer (N=2,684; 3.2%) or race (N=800; 0.9%), not mutually exclusive. Then we further excluded those participants who reported at baseline with cancer except for non-melanoma skin cancer (N=6,684; 7.9%). After exclusions, 76,121 participants remained for the current analysis. Among those, there were 16,137 individuals with T2D at baseline.
Follow-up time started at the age of the baseline survey and ended at the time of cancer diagnosis, death, or the last date of linkage to identify incident cancer cases, whichever came first. The date of the last cancer registry linkage differed by states, ranging from 2016 to 2020, and mostly in 2018–2019. For this analysis, we used Cox proportional hazards models to evaluate the association of T2D with cancer risk. We used age as the time scale in the Cox regression model. All models were adjusted for sex (male or female), race (Black, White, or other), education (less than high school, completed high school or general education development or other training, some college or junior college, graduated from college or further education), annual household income (< $15,000, $15,000-$24,999, $25,000-$49,000, $50,000-$99,999, $100,000+), enrollment source (community health center in-person interview, general population mailed survey, or general population telephone interview), smoking status (current, former, or never), pack-years of smoking, and body-mass index (<18.5, 18.5–24.99, 25–29.99, 30–34.99, ≥35.0 kg/m2). Alcohol intake was not adjusted in the analysis as it was not a confounder as evaluated using the directed acyclic graph (DAG) analysis. Furthermore, there was no statistical difference among individuals who had diabetes or did not have diabetes regarding alcohol intake (P-value = 0.2679). Additionally, we conducted stratified analyses by race (white and black), years between the diagnosis of type 2 diabetes and the baseline survey (<10 years, 10–19.99 years, and >=10 years), and years since diabetes diagnosis (<15 years, 15–29.99 years, and >=30 years). The years since the diabetes diagnosis were treated as a time-varying variable in the Cox regression models. We only analyzed cancers with more than 50 cases to have stable estimates. The Cox proportional hazards assumption was formally assessed using graphical visuals and statistical tests based on the scaled Schoenfeld residuals, and this assumption generally held for the associations evaluated in this study.
Analyses were further stratified by gender, race, smoking status, BMI, and follow-up time (< 5 years and ≥ 5 years). For the analysis stratified by follow-up time, person-years for <5 years of follow-up were calculated from all eligible participants for the first five years of follow-up, while person-years for ≥ 5 years of follow-up were calculated excluding the first 5-year follow-up for those with≥ 5 years of follow-up. Interactions were assessed with likelihood ratio tests to compare the models with and without the addition of cross-product terms.
The missing values of these covariates (less than < 5% missing for all covariates in general) were first imputed using the multivariate imputation by chained equations to preserve the data(21). Of these, the missing categorical covariates were household income (N=2,618; 3.1%); education (N=1,848; 2.2%); and smoking status (N=2,095; 2.5%). The missing continuous covariate was BMI (N=2,393; 2.8%). All P values presented were two-sided, with statistical significance set at 0.05.
Data Availability Statement
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.
RESULTS
The descriptive and demographic characteristics of the study population for the overall study population and by T2D status are presented in Table 1. Among those with T2D, the mean baseline age was 52.0 years, 41.4% were males, 67.2% were Black adults, and 64.0% were former or current smokers. Among those without T2D, the mean baseline age was 54.8 years, 35.6% were males, 69.5% were Black adults, and 58.6% were either former or current smokers.
Table 1.
Selected baseline characteristics of study participants by comorbidity, Southern Community Cohort Study
| Baseline Characteristics | Total Cohort | Type 2 Diabetes |
|
|---|---|---|---|
| No | Yes | ||
| N = 76,121 | N = 59,984 | N = 16,137 | |
|
| |||
| Age (Mean (SD), yr) | 52.0 (8.67) | 51.2 (8.46) | 54.8 (8.87) |
| Race (N (%)) | |||
| White | 22,044 (29.0%) | 17,772 (29.6%) | 4,272 (26.5%) |
| Black | 51,121 (67.2%) | 39,913 (66.5%) | 11,208 (69.4%) |
| Other | 2,956 (3.9%) | 2,299 (3.8%) | 657 (4.1%) |
| Sex (Male (N), %) | 31,542 (41.4%) | 25,792 (43.0%) | 5,750 (35.6%) |
| Baseline body mass index (Mean (SD)) | 30.3 (7.56) | 29.3 (7.12) | 34.0 (8.04) |
| Education | |||
| Less than high school | 47,327 (62.2%) | 36,582 (61.0%) | 7,395 (45.8%) |
| Completed high school, GED, or other training | 4,239 (5.6%) | 3,358 (5.6%) | 881 (5.5%) |
| Some college or junior college | 14,904 (19.6%) | 12,019 (20.0%) | 2,885 (17.9%) |
| Graduated from college or further education | 9,651 (12.7%) | 8,025 (12.7%) | 1,626 (10.1%) |
| Income (N (%)) | |||
| < $15,000 | 42,362 (55.7%) | 32,503 (54.2%) | 9,859 (61.1%) |
| $15,000 - $24,999 | 16,217 (21.3%) | 12,861 (21.4%) | 3,356 (20.8%) |
| $25,000 - $49,999 | 10,573 (13.9%) | 8,611 (14.4%) | 1,962 (12.2%) |
| $50,000 - $99,999 | 5,229 (6.9%) | 4,444 (7.4%) | 785 (4.9%) |
| ≥ $100,000 | 1,740 (2.3%) | 1,565 (2.6%) | 175 (1.1%) |
| Smoking Status | |||
| Current | 31,584 (41.5%) | 26,886 (44.8%) | 4,698 (29.1%) |
| Former | 17,104 (22.5%) | 12,336 (20.6%) | 4,768 (29.5%) |
| Never | 27,433 (36.0%) | 20,762 (34.6%) | 6,671 (41.3%) |
| Packyears smoking (Mean (SD)) | 14.7 (21.2) | 14.6 (20.5) | 15.0 (23.6) |
Figure 1 portrays the associations between T2D and the risk of overall cancer or site-specific cancers. Compared to individuals without T2D, individuals with T2D had an elevated risk of all cancers combined (HR: 1.07; 95% CI: 1.01–1.13). Individuals with T2D also had an elevated risk of certain site-specific cancers, such as stomach, colorectal, pancreatic, liver, and intrahepatic bile duct, and kidney and renal pelvis cancers as well as leukemia, and a reduced risk of prostate cancers. Analyses were also performed after excluding those individuals with < 5 years of follow-up time (Supplemental Table 1).
Figure 1: HRs and 95% CI of cancer associated with type 2 diabetes, Southern Community Cohort Study.
Model adjusted for sex, race, education, income, enrollment source, smoking status, pack-years, and body-mass index.
Table 2 shows the association between T2D and cancer risk stratified by race. White individuals with T2D had a significantly elevated risk of all cancers combined and several site-specific cancers, including colorectal and liver and intrahepatic bile duct cancers. Similar to White individuals, Black individuals with T2D also had a significantly elevated risk of any cancer, as well as liver and intrahepatic bile duct cancers. In addition, Black individuals with T2D also had an elevated risk for stomach, pancreatic, leukemia, and kidney and renal pelvis cancers. However, the racial difference in the strength of the T2D-cancer association was only statistically significant for all cancers combined and uterine body and liver and intrahepatic bile duct cancers (P for heterogeneity < 0.05).
Table 2.
HRs and 95% CI of cancer associated with type 2 diabetes by race, Southern Community Cohort Study
| Ever Diagnosed with Diabetes | Among White |
Among Black |
P for heterogeneity 3 | |||||
|---|---|---|---|---|---|---|---|---|
| Cases (n) | HR (95% CI)1 | HR (95% CI)2 | Cases (n) | HR (95% CI)1 | HR (95% CI)2 | |||
|
| ||||||||
| All Cancer | ||||||||
| No | 1,670 | 1.00 | 1.00 | 3,857 | 1.00 | 1.00 | ||
| Yes | 481 | 1.12 (1.01, 1.24) | 1.18 (1.06, 1.31) | 1,152 | 0.97 (0.90, 1.03) | 1.02 (0.95, 1.09) | 0.011 | |
| Oral and Pharynx | ||||||||
| No | 63 | 1.00 | 1.00 | 169 | 1.00 | 1.00 | ||
| Yes | 11 | 0.72 (0.38, 1.38) | 0.86 (0.44, 1.71) | 19 | 0.42 (0.26, 0.68) | 0.60 (0.37, 0.98) | 0.245 | |
| Esophagus | ||||||||
| No | 19 | 1.00 | 1.00 | 73 | 1.00 | 1.00 | ||
| Yes | 3 | 0.63 (0.18, 2.16) | 0.65 (0.18, 2.41) | 13 | 0.63 (0.35, 1.15) | 0.84 (0.45, 1.57) | 0.245 | |
| Stomach | ||||||||
| No | 21 | 1.00 | 1.00 | 64 | 1.00 | 1.00 | ||
| Yes | 8 | 1.39 (0.61, 3.17) | 1.22 (0.51, 2.95) | 30 | 1.44 (0.92, 2.24) | 1.71 (1.07, 2.72) | 0.919 | |
| Small Intestine | ||||||||
| No | 10 | 1.00 | 1.00 | 33 | 1.00 | 1.00 | ||
| Yes | 3 | 1.47 (0.39, 5.43) | 1.34 (0.33, 5.41) | 12 | 1.11 (0.56, 2.17) | 1.07 (0.54, 2.13) | 0.927 | |
| Colorectal | ||||||||
| No | 134 | 1.00 | 1.00 | 411 | 1.00 | 1.00 | ||
| Yes | 53 | 1.52 (1.10, 2.11) | 1.62 (1.15, 2.29) | 145 | 1.13 (0.93, 1.37) | 1.13 (0.92, 1.37) | 0.087 | |
| Pancreas | ||||||||
| No | 45 | 1.00 | 1.00 | 109 | 1.00 | 1.00 | ||
| Yes | 19 | 1.63 (0.94, 2.81) | 1.58 (0.89, 2.82) | 62 | 1.75 (1.27, 2.40) | 2.05 (1.47, 2.86) | 0.756 | |
| Larynx | ||||||||
| No | 32 | 1.00 | 1.00 | 80 | 1.00 | 1.00 | ||
| Yes | 4 | 0.45 (0.16, 1.29) | 0.61 (0.20,1.82) | 18 | 0.88 (0.52, 1.49) | 1.24 (0.72, 2.14) | 0.276 | |
| Lung | ||||||||
| No | 377 | 1.00 | 1.00 | 701 | 1.00 | 1.00 | ||
| Yes | 89 | 0.79 (0.62, 0.99) | 1.15 (0.90, 1.46) | 144 | 0.65 (0.54, 0.78) | 0.92 (0.76, 1.11) | 0.044 | |
| Melanoma | ||||||||
| No | 70 | 1.00 | 1.00 | 6 | 1.00 | 1.00 | ||
| Yes | 15 | 0.92 (0.52, 1.62) | 0.92 (0.51, 1.65) | 1 | (0.49 (0.06, 4.14) | 0.41 (0.05, 3.60) | 0.574 | |
| Breast | ||||||||
| No | 266 | 1.00 | 1.00 | 579 | 1.00 | 1.00 | ||
| Yes | 76 | 1.14 (0.88, 1.48) | 1.04 (0.79, 1.36) | 227 | 1.17 (1.00, 1.37) | 1.12 (0.96, 1.32) | 0.967 | |
| Uterine Cervix | ||||||||
| No | 14 | 1.00 | 1.00 | 57 | 1.00 | 1.00 | ||
| Yes | 6 | 1.80 (0.68, 4.76) | 1.69 (0.60, 4.69) | 9 | 0.48 (0.23, 0.97) | 0.49 (0.24, 1.01) | 0.060 | |
| Uterine Body | ||||||||
| No | 47 | 1.00 | 1.00 | 120 | 1.00 | 1.00 | ||
| Yes | 30 | 2.52 (1.57, 4.02) | 1.55 (0.96, 2.51) | 48 | 0.96 (0.68, 1.34) | 0.74 (0.52, 1.04) | 0.033 | |
| Bladder | ||||||||
| No | 61 | 1.00 | 1.00 | 70 | 1.00 | 1.00 | ||
| Yes | 16 | 0.98 (0.56, 1.70) | 1.24 (0.69, 2.23) | 25 | 1.11 (0.70, 1.77) | 1.16 (0.72, 1.88) | 0.632 | |
| Brain | ||||||||
| No | 15 | 1.00 | 1.00 | 21 | 1.00 | 1.00 | ||
| Yes | 7 | 1.75 (0.71, 4.36) | 1.85 (0.70, 4.86) | 6 | 0.90 (0.36, 2.28) | 0.82 (0.32, 2.11) | 0.360 | |
| Non-Hodgkin Lymphoma | ||||||||
| No | 59 | 1.00 | 1.00 | 92 | 1.00 | 1.00 | ||
| Yes | 20 | 1.33 (0.79, 2.23) | 1.13 (0.66, 1.94) | 34 | 1.30 (0.87, 1.95) | 1.25 (0.82, 1.90) | 0.848 | |
| Multiple Myeloma | ||||||||
| No | 23 | 1.00 | 1.00 | 81 | 1.00 | 1.00 | ||
| Yes | 5 | 0.84 (0.31 2.23) | 0.60 (0.22, 1.66) | 31 | 1.05 (0.69, 1.60) | 0.90 (0.59, 1.39) | 0.626 | |
| Liver and Intrahepatic Bile Duct | ||||||||
| No | 44 | 1.00 | 1.00 | 203 | 1.00 | 1.00 | ||
| Yes | 22 | 2.00 (0.19, 3.38) | 2.94 (1.66, 5.20) | 58 | 1.10 (0.82, 1.49) | 1.43 (1.05, 1.95) | 0.049 | |
| Leukemia | ||||||||
| No | 45 | 1.00 | 1.00 | 47 | 1.00 | 1.00 | ||
| Yes | 15 | 1.38 (0.76, 2.51) | 1.44 (0.77, 2.67) | 29 | 1.81 (1.12, 2.90) | 1.65 (1.01, 2.69) | 0.329 | |
| Ovary | ||||||||
| No | 20 | 1.00 | 1.00 | 35 | 1.00 | 1.00 | ||
| Yes | 2 | 0.36 (0.08, 1.54) | 0.40 (0.09, 1.78) | 9 | 0.64 (0.30, 1.34) | 0.62 (0.29, 1.32) | 0.435 | |
| Prostate | ||||||||
| No | 206 | 1.00 | 1.00 | 725 | 1.00 | 1.00 | ||
| Yes | 44 | 0.98 (0.70, 1.36) | 1.04 (0.74, 1.46) | 163 | 0.83 (0.70, 0.98) | 0.77 (0.64, 0.92) | 0.420 | |
| Kidney and Renal Pelvis | ||||||||
| No | 61 | 1.00 | 1.00 | 123 | 1.00 | 1.00 | ||
| Yes | 24 | 1.61 (0.99, 2.59) | 1.36 (0.82, 2.24) | 53 | 1.46 (1.05, 2.03) | 1.43 (1.02, 2.01) | 0.930 | |
| Thyroid | ||||||||
| No | 38 | 1.00 | 1.00 | 58 | 1.00 | 1.00 | ||
| Yes | 9 | 1.03 (0.50, 2.16) | 0.94 (0.43, 2.01) | 16 | 0.98 (0.56, 1.72) | 0.86 (0.48, 1.52) | 0.936 | |
Model adjusted for sex, education, income, and enrollment source.
Model adjusted for sex, education, income, enrollment source, smoking status, pack-years, and body mass index.
P for heterogeneity calculated for model 2 adjustments.
We also assessed cancer incidence stratified by years between the baseline survey and diagnosis of T2D (Table 3). In order to provide stable risk estimates, we included in this analysis only cancer types with >50 cases diagnosed with T2D. Results show that individuals with T2D have a slightly decreasing risk of cancer as the years increase, where those with T2D for ≥ 20 years have an HR of 0.96 (95% CI: 0.83–1.09). Among individuals with T2D, there is a significantly decreasing risk of lung cancer as the years increase.. We also assessed associations of T2D diagnosis with cancer risk according to years since T2D diagnosis using time-varying Cox regression models (Table 4). Similarly, we included in this analysis only cancer types with >50 cases diagnosed with T2D. Results show that individuals with T2D have an increasing risk of all cancers combined within 10 years after T2D diagnosis, with an HR of 1.10 (95% CI: 1.02–1.17). The risk begins to decrease with an increasing duration since T2D diagnosis. A similar pattern of the associations was found for virtually all site-specific cancers evaluated in Table 4, except for cancers of the pancreas and liver/intrahepatic bile duct, for which elevated risks remain statistically significant even 15 to 30 years after T2D diagnosis.
Table 3.
HRs and 95% CI of cancer associated with diabetes by years between diagnosis of type 2 diabetes and baseline survey, Southern Community Cohort Study
| Cancer Type and Diabetes Status | Total Cases (n)2 | <10 years |
10–19.99 years |
≥ 20 years |
|---|---|---|---|---|
| HR (95% CI)1 | HR (95% CI)1 | HR (95% CI)1 | ||
|
| ||||
| All Cancer | ||||
| No | 5,682 | 1.00 | 1.00 | |
| Yes | 1,694 | 1.10 (1.03, 1.18) | 1.02 (0.92, 1.13) | 0.96 (0.83, 1.09) |
| Colorectal | ||||
| No | 562 | 1.00 | 1.00 | |
| Yes | 202 | 1.27 (1.04, 1.55) | 0.96 (0.69, 1.34) | 1.31 (0.91, 1.87) |
| Pancreas | ||||
| No | 159 | 1.00 | 1.00 | |
| Yes | 81 | 1.99 (1.43, 2.75) | 1.30 (0.74, 2.27) | 2.01 (1.14, 3.52) |
| Lung | ||||
| No | 1,114 | 1.00 | 1.00 | |
| Yes | 247 | 1.04 (0.87, 1.24) | 0.96 (0.73, 1.26) | 0.96 (0.68, 1.34) |
| Breast | ||||
| No | 868 | 1.00 | 1.00 | |
| Yes | 307 | 1.06 (0.90, 1.25) | 1.15 (0.90, 1.45) | 1.02 (0.75, 1.39) |
| Uterine Body | ||||
| No | 170 | 1.00 | 1.00 | |
| Yes | 82 | 1.14 (0.84, 1.56) | 0.54 (0.29, 0.99) | 1.17 (0.68, 2.01) |
| Non-Hodgkin Lymphoma | ||||
| No | 157 | 1.00 | 1.00 | |
| Yes | 55 | 1.28 (0.88, 1.65) | 1.12 (0.63, 2.00) | 0.59 (0.22, 1.62) |
| Liver and Intrahepatic Bile Duct | ||||
| No | 253 | 1.00 | 1.00 | |
| Yes | 86 | 1.92 (1.42, 2.58) | 1.83 (1.14, 2.92) | 1.07 (0.50, 2.29) |
| Prostate | ||||
| No | 954 | 1.00 | 1.00 | |
| Yes | 216 | 0.83 (0.69, 1.00) | 0.75 (0.56, 0.99) | 0.66 (0.44, 1.00) |
| Kidney and Renal Pelvis | ||||
| No | 196 | 1.00 | 1.00 | |
| Yes | 80 | 1.49 (1.08, 2.04) | 1.52 (0.95, 2.44) | 0.96 (0.45, 2.06) |
Model adjusted for sex, race, education, income, enrollment source, smoking status, pack-years, and body mass index.
Reference group is individuals without diabetes.
Table 4.
HRs and 95% CI of cancer associated with diabetes by years since diabetes diagnosis, Southern Community Cohort Study
| Cancer Type and Diabetes Status |
Total Cases (n)2 | <15 years |
15–29.99 years |
≥30 years |
|---|---|---|---|---|
| HR (95% CI)1 | HR (95% CI)1 | HR (95% CI)1 | ||
|
| ||||
| All Cancer | ||||
| No | 5,682 | 1.00 | 1.00 | |
| Yes | 1,694 | 1.10 (1.02, 1.17) | 1.04 (0.95, 1.14) | 0.88((0.73, 1.05) |
| Colorectal | ||||
| No | 562 | 1.00 | 1.00 | |
| Yes | 202 | 1.20 (0.98, 1.48) | 1.19 (0.91, 1.55) | 1.22 (0.77, 1.94) |
| Pancreas | ||||
| No | 159 | 1.00 | 1.00 | |
| Yes | 81 | 1.92 (1.37, 2.69) | 1.84 (1.21, 2.82) | 1.32 (0.58, 3.01) |
| Lung | ||||
| No | 1,114 | 1.00 | 1.00 | |
| Yes | 247 | 1.08 (0.91, 1.29) | 0.84 (0.65, 1.09) | 1.02 (0.68, 1.53) |
| Breast | ||||
| No | 868 | 1.00 | 1.00 | |
| Yes | 307 | 1.03 (0.87, 1.22) | 1.19 (0.97, 1.47) | 0.96 (0.65, 1.43) |
| Uterine Body | ||||
| No | 170 | 1.00 | 1.00 | |
| Yes | 82 | 1.22 (0.89, 1.66) | 0.56 (0.33, 0.95) | 1.16 (0.61, 2.21) |
| Non-Hodgkin Lymphoma | ||||
| No | 157 | 1.00 | 1.00 | |
| Yes | 55 | 1.20 (0.82, 1.75) | 1.28 (0.78, 2.10) | 0.24 (0.03, 1.74) |
| Liver and Intrahepatic Bile Duct | ||||
| No | 253 | 1.00 | 1.00 | |
| Yes | 86 | 1.95 (1.44, 2.66) | 1.66 (1.09, 2.51) | 1.18 (0.48, 2.87) |
| Prostate | ||||
| No | 954 | 1.00 | 1.00 | |
| Yes | 216 | 0.86 (0.71, 1.03) | 0.70 (0.54, 0.91) | 0.59 (0.33, 1.04) |
| Kidney and Renal Pelvis | ||||
| No | 196 | 1.00 | 1.00 | |
| Yes | 80 | 1.57 (1.14, 2.16) | 1.43 (0.93, 2.18) | 0.42 (0.10, 1.70) |
Model adjusted for sex, race, education, income, enrollment source, smoking status, pack-years, and body mass index.
Reference group is individuals without diabetes.
We also performed stratified analyses by sex (Supplemental Table 2), smoking status (Supplemental Table 3), BMI (Supplemental Table 4), and by both race and income (Supplemental Table 5 and Supplemental Table 6). Overall, no apparent modifying effect was observed, and the tests for interactions were generally not significant after adjusting for multiple comparisons.
DISCUSSION
In this large prospective cohort study, we found an increased risk of overall cancer in association with T2D. Those with T2D had an elevated risk of certain site-specific cancers, such as leukemia as well as stomach, colorectal, pancreatic, liver and intrahepatic bile duct, and kidney and renal pelvis cancers; and a reduced risk of prostate cancer. This study provides supporting evidence that some findings observed previously in other populations regarding the association between T2D and cancer risk can be extended to the low-income and Black population, while some other findings are unique.
Limited studies have assessed the associations of T2D with the overall risk of developing cancer in Black populations. However, there are studies analyzing this association in other cohorts. Results from a study by Bjornsdottir et al. using data from a Swedish cohort found that the risk of any cancer among T2D patients had an HR of 1.19 (95% CI: 1.09–1.12). Another study by Inoue et al. using data from a Japanese cohort found similar results with an HR of 1.27 95% CI: 1.14–1.42)(4,5). Data combined from previous observational studies and meta-analyses have shown that T2D is positively associated with increased risk of all cancers and several site-specific cancers, including pancreatic, liver, kidney, bladder, breast cancer, colorectal, and endometrial, while being negatively associated with risk of prostate cancer(2,22). We found a weak positive association of T2D with all cancers combined (HR: 1.07; 95% CI: 1.01–1.13). The reasons for this inconsistency are unclear. Our findings for site-specific cancers are similar to those reported previously in other populations.
With a few exceptions, the associations between T2D and cancer risk were found to be similar between Black and White individuals in this study. The association of T2D with the risk of overall cancer and cancer of the liver and intrahepatic bile duct appeared to be stronger in White than Black study participants (P for heterogeneity < 0.05). The reasons for this finding are unclear. It is possible that other cancer risk factors are more prevalent in Black than white participants in this study, which partially masked the positive association of T2D with overall cancer risk in Black participants. Surprisingly, a prior T2D diagnosis showed a positive association with uterine body cancer in White participants but an inverse association in Black participants (P for heterogeneity < 0.05). Again, the reasons for this difference are unclear. However, we could not rule out possible chance findings due to multiple comparisons.
Overall, we found a weaker association between T2D and cancer risk in this study than previously reported. It is possible that the risk of cancer in this low-income population, including those without prior diabetes, may have masked some of the association signals with T2D in our study, resulting in a weaker association observed in this study compared with some previous studies conducted in other populations. Nevertheless, our study did provide additional evidence for a positive association of T2D with the risk of certain cancers, which can help understand implications from T2D on cancer in low-income and minority populations whom previous analyses may not well represent.
Certain mechanisms may explain the increased risk of cancer among diabetic individuals. These plausible mechanisms affecting cancer etiology and accelerating this progression could be hyperglycemia, hyperinsulinemia, and inflammation(3,23). Hyperglycemia could promote the proliferation of cancer cells and cell growth(3). Similarly, previous studies have shown that hyperinsulinemia could increase the expression of insulin growth factor-1 and activate the receptor, leading to further cancer cell stimulation and growth(24). For example, individuals with diabetes could be at an increased risk of pancreatic cancer due to hyperinsulinemia, glucose toxicity, excess iron in the circulation, exogenous insulin therapy, and impaired B-cell function(23,25–27). For the SCCS population, the effects of these biological mechanisms, along with socioeconomic and sociodemographic factors among individuals with diabetes, could be accelerating the cancer risk.
The main noticeable strength of our study is that two-thirds of our study participants represent the underserved and understudied populations. Our study also has a higher baseline prevalence of T2D (20%) coupled with a comprehensive assessment of potential demographic, lifestyle, and behavioral major disease risk factors, which allowed us to enable control for a wide range of potential confounders and carefully evaluate the association. However, a limitation of our study is that T2D was self-reported, and the specific date of T2D diagnosis is unclear, which could introduce potential exposure misclassification. Previous SCCS validation studies have shown that self-reported T2D is highly specific(23,28). We did not specifically ask for T2D diagnosis in the study, thus, some diabetes diagnoses could be type 1 diabetes. Given that type 1 diabetes prevalence is substantially lower than T2D in middle-aged to older populations, we could assume that the vast majority of diabetes diagnoses should be T2D. Some T2D patients may be undiagnosed or unreported in the population(28), which could also attenuate the true association between the exposure and outcomes. Some T2D patients may change their lifestyle after diagnosis, which could confound the association of prior T2D with the risk of certain cancers. Future studies with repeat surveys of lifestyle factors could be helpful to address this concern.
In conclusion, in this large prospective cohort study, we found that T2D was associated with the risk of overall cancer and certain site-specific cancers. Our findings were generally consistent with previous studies and provide strong support that T2D is a significant risk factor for cancer in low-income and minority populations. Our study stresses the importance of T2D prevention in reducing the incidence of these cancers.
Supplementary Material
ACKNOWLEDGEMENTS
We thank all study participants who took part in the study. Also, we would like to thank Rachel Mullen in the Division of Epidemiology, Vanderbilt University Medical Center, for assistance with preparation.
Funding:
This study is supported, in part, by the National Institutes of Health U01 CA 202979 and Anne Potter Wilson Chair endowment funds to Vanderbilt University which were received by W. Zheng.
Role of the funder/sponsor:
The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Footnotes
Conflict of interest: The authors have no conflict of interest to declare.
Disclaimer: The research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number U01CA202979 which was received by W. Zheng. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Additional information: Data for cancer diagnosis used in this study were provided by the Alabama Statewide Cancer Registry, Kentucky Cancer Registry, Tennessee Department of Health, Office of Cancer Surveillance, Florida Cancer Data System, North Carolina Central Cancer Registry, North Carolina Division of Public Health, Georgia Comprehensive Cancer Registry, Louisiana Tumor Registry, Mississippi Cancer Registry, South Carolina Central Cancer Registry, Virginia Department of Health, Virginia Cancer Registry, and the Arkansas Department of Health Cancer Registry. The Arkansas Central Cancer Registry is fully funded by a grant from National Program of Cancer Registries of the CDC. Data on SCCS cancer cases from Mississippi were collected by the Mississippi Cancer Registry, which participates in the National Program of Cancer Registries of the CDC. Data on SCCS cancer cases from West Virginia were provided by the West Virginia Cancer Registry.
REFERENCES
- 1.National Diabetes Statistics Report 2020. Estimates of diabetes and its burden in the United States. [Internet]. Centers for Disease Control and Prevention; 2020. page 32. Available from: https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf [Google Scholar]
- 2.Yeh H-C, Golozar A, Brancati FL. Cancer and Diabetes. In: Cowie CC, Casagrande SS, Menke A, Cissell MA, Eberhardt MS, Meigs JB, et al. , editors. Diabetes Am [Internet]. 3rd ed. Bethesda (MD): National Institute of Diabetes and Digestive and Kidney Diseases (US); 2018. [cited 2023 Jan 5]. page 85–96. Available from: http://www.ncbi.nlm.nih.gov/books/NBK568016/ [Google Scholar]
- 3.Giovannucci E, Harlan DM, Archer MC, Bergenstal RM, Gapstur SM, Habel LA, et al. Diabetes and Cancer. Diabetes Care. 2010;33:1674–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bjornsdottir HH, Rawshani A, Rawshani A, Franzén S, Svensson A-M, Sattar N, et al. A national observation study of cancer incidence and mortality risks in type 2 diabetes compared to the background population over time. Sci Rep. 2020;10:17376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Inoue M, Iwasaki M, Otani T, Sasazuki S, Noda M, Tsugane S, et al. Diabetes Mellitus and the Risk of Cancer: Results From a Large-Scale Population-Based Cohort Study in Japan. Arch Intern Med. 2006;166:1871–7. [DOI] [PubMed] [Google Scholar]
- 6.Bancks MP, Kershaw K, Carson AP, Gordon-Larsen P, Schreiner PJ, Carnethon MR. Association of Modifiable Risk Factors in Young Adulthood With Racial Disparity in Incident Type 2 Diabetes During Middle Adulthood. JAMA. 2017;318:2457–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.LeRoith D, Novosyadlyy R, Gallagher EJ, Lann D, Vijayakumar A, Yakar S. Obesity and type 2 diabetes are associated with an increased risk of developing cancer and a worse prognosis; epidemiological and mechanistic evidence. Exp Clin Endocrinol Diabetes Off J Ger Soc Endocrinol Ger Diabetes Assoc. 2008;116 Suppl 1:S4–6. [DOI] [PubMed] [Google Scholar]
- 8.Ferguson RD, Gallagher EJ, Scheinman EJ, Damouni R, LeRoith D. The epidemiology and molecular mechanisms linking obesity, diabetes, and cancer. Vitam Horm. 2013;93:51–98. [DOI] [PubMed] [Google Scholar]
- 9.Kang C, LeRoith D, Gallagher EJ. Diabetes, Obesity, and Breast Cancer. Endocrinology. 2018;159:3801–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gallagher EJ, LeRoith D. Obesity and Diabetes: The Increased Risk of Cancer and Cancer-Related Mortality. Physiol Rev. 2015;95:727–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, 2009. CA Cancer J Clin. 2009;59:225–49. [DOI] [PubMed] [Google Scholar]
- 12.Lam C, Cronin K, Ballard R, Mariotto A. Differences in cancer survival among white and black cancer patients by presence of diabetes mellitus: Estimations based on SEER‐Medicare‐linked data resource. Cancer Med. 2018;7:3434–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gaskin DJ, Thorpe RJ, McGinty EE, Bower K, Rohde C, Young JH, et al. Disparities in diabetes: the nexus of race, poverty, and place. Am J Public Health. 2014;104:2147–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Brancati FL, Whelton PK, Kuller LH, Klag MJ. Diabetes mellitus, race, and socioeconomic status. A population-based study. Ann Epidemiol. 1996;6:67–73. [DOI] [PubMed] [Google Scholar]
- 15.Brancati FL, Kao WHL, Folsom AR, Watson RL, Szklo M. Incident Type 2 Diabetes Mellitus in African American and White AdultsThe Atherosclerosis Risk in Communities Study. JAMA. 2000;283:2253–9. [DOI] [PubMed] [Google Scholar]
- 16.Cheng YJ, Kanaya AM, Araneta MRG, Saydah SH, Kahn HS, Gregg EW, et al. Prevalence of Diabetes by Race and Ethnicity in the United States, 2011–2016. JAMA. 2019;322:2389–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Signorello LB, Hargreaves MK, Blot WJ. The Southern Community Cohort Study: Investigating Health Disparities. J Health Care Poor Underserved. 2010;21:26–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.About the SCCS [Internet]. South. Community Cohort Study. [cited 2025 May 20]. Available from: https://www.southerncommunitystudy.org/about-the-study.html [Google Scholar]
- 19.Signorello LB, Hargreaves MK, Steinwandel MD, Zheng W, Cai Q, Schlundt DG, et al. Southern community cohort study: establishing a cohort to investigate health disparities. J Natl Med Assoc. 2005;97:972–9. [PMC free article] [PubMed] [Google Scholar]
- 20.Signorello LB, Hargreaves MK, Blot WJ. The Southern Community Cohort Study: Investigating Health Disparities. J Health Care Poor Underserved. 2010;21:26–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res. 2011;20:40–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ballotari P, Vicentini M, Manicardi V, Gallo M, Chiatamone Ranieri S, Greci M, et al. Diabetes and risk of cancer incidence: results from a population-based cohort study in northern Italy. BMC Cancer. 2017;17:703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Conway RBN, Hudson AG, Munro H, Fu D, McClain DA, Blot WJ. Diabetes and pancreatic cancer risk in a multiracial cohort. Diabet Med J Br Diabet Assoc. 2023;e15234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Xu C-X, Zhu H-H, Zhu Y-M. Diabetes and cancer: Associations, mechanisms, and implications for medical practice. World J Diabetes. 2014;5:372–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wolpin BM, Bao Y, Qian ZR, Wu C, Kraft P, Ogino S, et al. Hyperglycemia, Insulin Resistance, Impaired Pancreatic β-Cell Function, and Risk of Pancreatic Cancer. JNCI J Natl Cancer Inst. 2013;105:1027–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Fisher WE, Boros LG, Schirmer WJ. Insulin Promotes Pancreatic Cancer: Evidence for Endocrine Influence on Exocrine Pancreatic Tumors. J Surg Res. Elsevier; 1996;63:310–3. [DOI] [PubMed] [Google Scholar]
- 27.Bao W, Rong Y, Rong S, Liu L. Dietary iron intake, body iron stores, and the risk of type 2 diabetes: a systematic review and meta-analysis. BMC Med. 2012;10:119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Conway BN, Han X, Munro HM, Gross AL, Shu X-O, Hargreaves MK, et al. The obesity epidemic and rising diabetes incidence in a low-income racially diverse southern US cohort. PloS One. 2018;13:e0190993. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.