Background
Significant racial disparities in cancer mortality have been well documented in the United States (US), with at least a 13% difference in 5-year survival observed between Blacks and Whites diagnosed with breast and colorectal cancer [
1]. Black adults are more likely to develop metabolic dysregulation and meet the formal criteria for metabolic syndrome (MetS). MetS is a cluster of interrelated biochemical abnormalities that include central obesity, insulin resistance, dyslipidemia and hypertension, and has been shown to significantly increase the risk of coronary heart disease, stroke and type-2 diabetes [
2‐
5]. Individual components of MetS, specifically obesity [
6,
7], diabetes [
8], and hypertension [
9,
10], have been associated with increased risk for cancer; however, the entire cluster of MetS components has only more recently been shown to be associated with cancer risk and outcome.
Several studies show a significant and independent positive association between MetS and cancer incidence [
11‐
13], distant metastasis [
14] and aggressive cancer phenotypes [
15‐
17]. A recent systematic review and meta-analysis of 43 studies found significant associations between MetS and risk of liver cancer, colorectal cancer, and bladder cancer [
18]. Importantly, the association between MetS and cancer risk has consistently been found to be larger than the corresponding associations for individual MetS components, suggesting that there may be independent, complex biological processes underlying this association. However, despite renewed interest in the role of MetS in cancer etiology and outcomes, many of the prior studies have suffered from significant lead-time, length and selection biases, with only a few studies examining racial differences in this association. Few prospective studies have examined objective measures of MetS at baseline in relation to cancer mortality, with racially diverse study populations to assess race-specific risk. Furthermore, although there is growing consensus regarding the biological importance of MetS, questions remain about the clinical definition, categorical cut-points and included components, especially in relation to cancer risk. In this study, we assess MetS and variables associated with metabolic dysregulation at baseline among Blacks and Whites, and examine the association with cancer mortality during follow-up.
This study takes advantage of the large REasons for Geographic and Racial Differences in Stroke (REGARDS) cohort, with individual assessment of MetS components at baseline, and cancer mortality identified prospectively during follow-up. The objective of the current study was to determine the association between MetS and metabolic dysregulation with cancer mortality among Black and White adults in the US.
Methods
Data source
REGARDS is one of the largest ongoing national longitudinal cohorts of community-dwelling adults in the US [
19]. Designed to examine factors contributing to racial and geographic differences in stroke mortality, the REGARDS study includes 30,239 participants ages 45 years and older at baseline; 45% male, 41% Black, and 69% >60 years old [
19]. REGARDS participants were randomly sampled from all US states and recruited via mail and telephone. REGARDS was originally designed to evaluate risk factors and racial disparities in stroke, therefore 30% of participants were recruited from the ‘Stroke Belt’ (Alabama, Arkansas, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, and Tennessee
), 20% from the ‘Stroke Buckle’ (Georgia, North Carolina, and South Carolina
) and approximately 50% from other US states. Participants were recruited between January 2003 and October 2007, and detailed information about demographics, health behaviors, chronic medical conditions, physical status, diet, and medications were collected. In addition, baseline in-home visits were scheduled for all enrolled patients, during which anthropometric data including body weight, waist circumference, height, and blood pressure were obtained. Participants were asked to fast overnight for 10–12 h before visitation and to have medications available during the time of visitation. Trained technicians collected blood and urine samples and assessed current medication use. Participants were subsequently contacted by telephone every 6-months to identify medical events or hospitalizations experienced since the prior contact. Medical events or deaths were ascertained using death certificates, medical records, and/or interviewed proxies to determine causes of the death.
Main exposure variables
We defined MetS based on the recently published consensus statement developed by multiple health and professional organizations [
20]. The joint harmonized criteria by Alberti et al. defined MetS as the presence of at least three of: 1) Diabetes: fasting glucose ≥126 mg per liter (mg/L) (or a glucose ≥200 mg/L for those not fasting) or the use of insulin or oral hypoglycemic agents; 2) High triglycerides: triglycerides ≥150 mg per deciliter (mg/dL) or reported use of medication for elevated triglycerides; 3) Dyslipidemia: low high-density lipoprotein (HDL) cholesterol <40 mg/dL for males and <50 mg/dL for females, or use of lipid lowering medications; 4) Hypertension: systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or the reported use of antihypertensive agents; and 5) Obesity: increased waist circumference (WC) >102 cm for males or >88 cm for females.
Secondary exposure variable
We additionally defined metabolic dysregulation based on factor analysis of 15 metabolic related variables (height, weight, WC, body mass index (BMI), log of triglycerides, cholesterol, HDL cholesterol, log of low-density lipoprotein (LDL) cholesterol, dyslipidemia, systolic blood pressure, diastolic blood pressure, hypertension, log of insulin, glucose, and diabetes). We used orthogonal rotation of each component in the factor analysis, examined the scree plots and employed an eigenvalue cut-point of approximately 1.0 for inclusion of final derived factors. We calculated final factor loadings of 6 factors based on the full sample, and presented results only for factors with absolute values for loadings >0.4. We named patterns based on the factor loadings that contributed most highly to each pattern (Additional file
1). For instance, Factor 1 was termed “Obesity” since it was loaded heavily by variables associated with body weight such as weight (0.91), waist circumference (0.86), BMI (0.94), and log insulin (0.61). Similarly, Factor 4 was termed “Lipids” since it was loaded heavily by log triglycerides (0.76), HDL cholesterol (−0.69), and dyslipidemia (0.70).
Cancer mortality
Cancer mortality was identified through semi-annual telephone follow-up, death information from participant proxies, linkages with the Social Security Death Index (SSDI) as well as the National Death Index (NDI). Date of death was confirmed using death certificates, SSDI and/or NDI, and cause of death was adjudicated by a committee of experts using all available information as recommended by national guidelines [
21]. Follow-up data for this analysis was available through December 31, 2012.
Participant characteristics
Demographic information used for analysis included age, race, gender, income, education, and geographic location. Health behaviors included tobacco and alcohol use. Chronic medical conditions assessed included: atrial fibrillation, chronic lung disease, chronic kidney disease, coronary artery disease, deep vein thrombosis, diabetes, dyslipidemia, hypertension, myocardial infarction, obesity, peripheral artery disease, and stroke. In addition, we included serum high-sensitivity C-reactive protein (hsCRP) as a covariate, since multiple studies have demonstrated racial differences in CRP [
22‐
24].
Statistical analysis
We compared baseline characteristics by race using chi-square tests for categorical characteristics, analysis of variance (ANOVA) for continuous variables, and Kruskal-Wallis test for non-normal continuous variables. We categorized each metabolic factor into quartiles based on the distribution among study participants, with the highest quartile corresponding to participants at the highest distribution of the factor. Thus, participants in the fourth quartile of the cholesterol factor had the highest mean cholesterol levels in the study. To estimate the hazards of cancer mortality, we fit Cox proportional hazard models examining each MetS component independently and jointly in relation to cancer mortality. We a priori specified examination of race-stratified models, and stratified statistical models by race groups after adjusting for socio-demographics, health behaviors, baseline medical conditions, and hs-CRP. In sensitivity analysis we adjusted models for socio-demographics and health behaviors. The results of all models were expressed as adjusted hazard ratios (AHR) and the corresponding 95% confidence intervals (CI). Individuals were censored at the time of death, loss to follow-up, or the end of cancer mortality ascertainment (December 31, 2012). SAS version 9.4 and STATA version 13 were used for all statistical analysis. We considered two-sided p values <0.05 as statistically significant.
Discussion
In a large prospective cohort of community-dwelling adults, we observed significant differences in the prevalence of MetS and metabolic dysregulation by race. This study fills a significant gap in the literature regarding the role of metabolic dysregulation, defined here using the harmonized MetS criteria as well as factor analysis, in relation to cancer mortality. Black participants in this study had a higher prevalence of MetS and dysregulated metabolic components when compared with White participants. Overall, participants who met the criteria for MetS were at significantly higher risk for cancer mortality, and the risk increased with increasing number of components. Of the factor analysis-derived metabolic components, elevated glucose levels was strongly associated with increased mortality, while obesity and high cholesterol were associated with reduced cancer mortality. In race-stratified analysis, Blacks with MetS had about a 2-fold increased risk of cancer mortality compared with those without MetS, however there were no significant associations among Whites.
The joint harmonized criteria for MetS was developed to generate consensus regarding the clinical importance and standardized definition of this condition[
20], further stimulating renewed attention into its impact on health outcomes. The link between individual metabolic components, i.e. central obesity, insulin resistance, dyslipidemia and chronic diseases have been known for many decades, although only recently has compelling research studies begun to emerge on the biological consequence the MetS cluster in chronic diseases. Metabolic syndrome has been associated with significantly increased risk of coronary heart disease and stroke in multiple research studies [
2‐
5]. Furthermore, individual components of MetS – specifically obesity [
6,
7], diabetes [
8], and hypertension [
9,
10] – have also been associated with increased risk of multiple cancers, and MetS has been shown to be as a strong risk factor for breast cancer, with odds ratios ranging from 2.50 in Brazil, (95% CI: 1.17–5.30) [
11], to 6.28 in Italy (95% CI 2.79–14.11) [
13]. Strikingly, women with MetS who develop breast cancer were more than twice as likely to develop distant metastasis [
14], and aggressive tumors [
16,
17]. What has remained less controversial is the inextricable link between poor diet and obesity with MetS, and an understanding that these predisposing factors are more prevalent among Blacks compared with Whites- a trend that was observed in this study and has remained consistent for several decades [
25,
26].
The consistent associations observed in this study between having MetS and increased risk of cancer mortality, as well as higher risk with increasing number of MetS components suggests that prevention strategies focused on these modifiable factors, i.e. obesity, cholesterol, blood pressure and glucose, are warranted and should be considered as part of comprehensive cancer control and prevention plans. Unfortunately, past strategies to reduce the prevalence of MetS risk factors may have had limited success due to individual and community level factors such as poverty, lack of availability of fresh food, safe walking environments and routine access to preventive care in many communities [
27,
28]. More research will be needed to better identify race-specific public health strategies that have the best chance of eliminating individual MetS components as well as clinical strategies to control the entire cluster of MetS, since those strategies may have significant potential to reduce racial disparities in cancer mortality.
Certain limitations are relevant to the interpretation of this study. First, although metabolic factors were measured at baseline prior to cancer diagnosis or mortality, observed values may have been subject to information biases. Second, although we adjusted for confounding due to several baseline covariates, physical activity variables that may affect lipid and blood pressure levels were not included in the analysis. In addition, the total number of cancer deaths may be underestimated in this population as the REGARDS study was primarily intended to identify incident stroke events was not specifically focused on cancer outcomes. We did not have information regarding cancer stage or treatment and therefore were unable to make statistical adjustments for these differences in our models. We were likely underpowered due to small sample sizes in the race-stratified analysis, and were unable to examine cancer-specific mortality in the present analysis. In particular, given the follow-up time of 5 to 9 years available in this study, less fatal cancer types such as breast and prostate cancer, may have been under-represented in the analysis. However, we plan future studies that include more years of follow-up and a wider range of cancer types to support cancer-specific mortality analysis.