Associations of Novel Dietary and Lifestyle Inflammation Scores With Incident Colorectal Cancer in the NIH-AARP Diet and Health Study


 
 
 Chronically higher inflammation, likely contributed to by dietary and lifestyle exposures, may increase risk for colorectal cancer (CRC). To address this, we investigated associations of novel dietary (DIS) and lifestyle (LIS) inflammation scores with incident CRC in the prospective National Institutes of Health–American Association of Retired Persons Diet and Health Study (N = 453 465).
 
 
 
 The components of our previously developed and externally validated 19-component DIS and 4-component LIS were weighted based on their strengths of associations with a panel of circulating inflammation biomarker concentrations in a diverse subset (N = 639) of participants in the REasons for Geographic and Racial Differences in Stroke Study cohort. We calculated the components and applied their weights in the National Institutes of Health-American Association of Retired Persons cohort at baseline, summed the weighted components (higher scores reflect a higher balance of proinflammatory exposures), and investigated associations of the scores with incident CRC using Cox proportional hazards regression. All statistical tests were two-sided.
 
 
 
 Over a mean 13.5 years of follow-up, 10 336 participants were diagnosed with CRC. Among those in the highest relative to the lowest DIS and LIS quintiles, the multivariable-adjusted hazards ratios (HRs) and their 95% confidence intervals (CIs) were HR = 1.27 (95% CI = 1.19 to 1.35; Ptrend < .001) and 1.38 (95% CI = 1.30 to 1.48; Ptrend < .001), respectively. The associations were stronger among men and for colon cancers. The hazards ratio for those in the highest relative to the lowest joint DIS and LIS quintile was HR = 1.83 (95% CI = 1.68 to 1.99; Pinteraction < .001).
 
 
 
 Aggregates of proinflammatory dietary and lifestyle exposures may be associated with higher risk for CRC.



Supplementary Methods: Calculating the DIS and LIS in the NIH-AARP Diet and Health Study
In NIH-AARP, we standardized each food group and the supplement score, by sex, to a mean of 0 and standard deviation of 1.0 based on the baseline distribution among all participants; i.e., − where denotes the study population mean, denotes the study population standard deviation, and X denotes the participant's intake.
For the LIS, baseline smoking status was categorized as 'current' or 'former and never'.
Baseline body mass index (BMI) was categorized as normal (18.5 -24.99 kg/m2), overweight (25 -29.99 kg/m2), or obese (BMI ≥ 30 kg/m2). Baseline heavy alcohol consumption for men and women was defined as > 2 or > 1 drinks/day, respectively; moderate consumption was defined as individuals consuming alcohol in less than these amounts. For physical activity, we categorized participants as those who did not or rarely exercised, exercised 1 -2 times/week, or exercised ≥ 3 times/wk.

Supplementary Methods: Statistical Analyses
Prior to conducting the Cox proportional hazards regression, we assessed the proportional hazards assumption by calculating Martingale and Schoenfeld residuals, testing time-dependent covariates, and inspecting log(-log) survival curves for each variable in the model. Variables that violated the proportional hazards assumptions were included in the SAS STRATA statement in all models and are listed in the table footnotes. We ruled out multicollinearity considering a condition index ≥ 30 and a variance decomposition proportion ≥ 0.5 as evidence of multicollinearity.
Potential confounders were based on biological plausibility, previous literature on their associations with CRC, inflammation, or dietary and lifestyle exposures, and causal diagrams.
Covariates considered for all models included age, sex, race, education, marital status, comorbidities (self-reported gallbladder disease, heart disease, emphysema, or diabetes mellitus), hormone replacement therapy use (among women), family history of CRC in a first degree relative, self-reported history of colon polyps, and total energy intake. Covariates considered for the DIS models also included individual covariates for smoking status, BMI, alcohol intake, and physical activity since they are strong CRC risk factors with hypothesized contributions to colorectal carcinogenesis through inflammation plus other independent pro-or anti-carcinogenic pathways. Covariates considered for the LIS models included former smoking status since the LIS only includes current smoking at baseline as a component, and former tobacco smoking is also associated with CRC risk. LIS models also included the equally-weighted DIS, since the 19 DIS components individually are weak CRC risk factors, to account for the components inflammation plus other colorectal carcinogenic-related effects, and to reduce model size (including the score components individually or collectively in the equally-weighted dietary inflammation score yielded no substantial differences in our estimated LIS-CRC association estimated HRs). We also considered adjustment for regular aspirin and other NSAID use in the subset of the cohort that completed RFQs 6 months from baseline, and adjustment for colonoscopy screening over follow-up in the subset that completed follow-up questionnaires from 2004 -2005; however, adjustment for these covariates did not materially change the estimated DIS/LIS-CRC associations.
We also investigated potential effect modification by conducting separate analyses for the DIS and LIS within categories of age (< / ≥ 65 years), sex and hormone replacement therapy use (among women), race (white, black, or other), baseline comorbidity (yes/no), family history of CRC in a first degree relative (yes/no), and for the DIS, baseline smoking status (never, former, or current), BMI (normal, overweight, or obese), baseline alcohol intake (non-drinker, moderate drinker, or heavy drinker), and baseline physical activity (exercises never or rarely, 1 -3 times/week, ≥ 3 times/week). In a subset of the cohort that completed RFQs 6 months from their baseline questionnaire, we conducted analyses within strata of regular aspirin or other NSAID use (≥ once/week). In the subset that completed follow-up questionnaires from 2004 -2005, we excluded participants who were diagnosed with CRC or were otherwise censored prior to 2004, and conducted analyses within strata of time since their last colonoscopy during follow up (never, < 5 years ago, ≥ 5 years ago). We assessed effect modification by comparing the stratum-specific estimates and by calculating Wald test p-values for model interaction terms.
To test statistically for heterogeneity by CRC site, we conducted a case-only analysis using multivariable logistic regression. The dependent variable was the CRC subtype (i.e., colon, left colon, right colon, and rectum/rectosigmoid cancer) with rectum/rectosigmoid cancer as the referent group. The independent variables were the DIS and LIS modeled continuously, with the same covariates as in the Cox proportional hazards models plus person-time (years of follow-up).
We took the P-value for the continuous DIS and LIS to be the Pheterogeneity.

Supplementary Methods: Sensitivity Analyses, Expanded with Rationales
To assess the sensitivity of the associations to various considerations, we repeated the analyses with the following variations. First, to explore potential differences in inflammation-related vs.
total contributions of the aggregated score components to risk, we constructed equally-weighted DIS and LIS versions by assigning positive or negative equal weights to dietary/lifestyle components we hypothesized a priori to be pro-inflammatory or anti-inflammatory, respectively.
Second, to rule out a substantial influence of the the supplement score component in our DIS, we calculated a DIS without Supplementary micronutrients and assessed its association with CRC.
Third, we calculated a reverse-direction Healthy Eating Index-2015 (reverse HEI-2015; i.e., so a higher score would be higher risk) (6), and the empirical dietary inflammatory pattern (EDIP), as described by Tabung et al. (7), and investigated their associations with CRC. Fourth, we investigated associations of each individual lifestyle component with CRC. Fifth, we excluded individuals who died or were diagnosed with CRC within two years from baseline to rule out spurious effects on risk estimates by participants with an undiagnosed CRC or morbid illness that may have affected their long-term diets or lifestyles at the time of questionnaire completion. Contain variety of potent antioxidants (e.g., -carotene, folacin, magnesium, calcium, glucosinolates, isothiocyanates, lutein, and indoles); contain flavonoids and polyphenols, which activate the transcription factor, Nrf2, which plays a key role in cellular protection against oxidative stress and inflammation (9,10,19,(11)(12)(13)(14)(15)(16)(17)(18) Tomatoes Contain -carotene, vitamin C, and lycopene, the latter of which is a potent singlet oxygen quencher and one of the most powerful antioxidants among the natural carotenoids (20)(21)(22)(23) Apples and berries Contain flavonoids (e.g., anthocyanins, quercetin, and phenolic acids) that suppress proinflammatory cytokine production and are powerful antioxidants; potentially increase postprandial plasma antioxidant capacity (24-26) Deep yellow or orange vegetables and fruit Contain pro-vitamin A carotenoids (e.g., -carotene and α-carotene), which have a conjugated double-bond structure making them strong antioxidants (27) Other fruits and real fruit juices Contain antioxidants (e.g., flavonoids, such as hesperidin, naringenin, neohesperidin, limonene, vitamin C, -cryptoxanthin, plant sterols, salicylates, naringin, nobelitin, and narirutin) with similar mechanisms to those described above (13,(28)(29)(30)(31)(32)(33)(34)(35) Other vegetables Contain antioxidants and polyphenols with similar mechanisms to those described above Legumes Contain folacin, iron, isoflavones, protein, vitamin B6, and have a high antioxidant capacity; rich in fiber, which is associated with beneficial alterations to the gut microbiota, reducing immune response in the gut (12,36,37) Fish Contain Ω-3 fatty acids, which compete with pro-inflammatory Ω-6 fatty acids by synthesizing eicosanoids and suppress the capacity of monocytes to synthesize IL-1β and TNF-α (38-40)

Poultry
Inversely associated with inflammation markers (41); contain low amounts of saturated fat (42); contain l-arginine, which improves endothelium-dependent dilation (precursor of the endogenous vasodilator nitric oxide) and decreases platelet aggregation and monocyte adhesion (12) Red and organ meats Contain heme iron, which increases the bioavailability of iron, which in turn increases oxidative stress; contain Ω-6 fatty acids, which increase oxidative stress through free radical production and are converted to arachidonic acid which stimulates expression of IL-1β and TNF-α in monocytes, and IL-6 and IL-8 in endothelial cells (43)(44)(45); contain saturated fats that mimic lipopolysaccharide, a pro-inflammatory stimulant, in the gut, and increase cytotoxic, pro-oxidant, and pro-inflammatory bile acids in the colon (43,46) Processed meats Contain heme iron, higher saturated fat contents, Ω-6 fatty acids (see above), and additives, such as nitrites, with suspected pro-inflammatory properties (41,47) Added sugars Sparse in nutrients; induce postprandial hyperglycemia, which act as stressful stimuli through subsequent repeated mild postprandial hypoglycemia (48) and reduce nitric oxide availability (plays role in regulation of inflammatory response (49)); elevate pro-inflammatory free fatty acid levels (40); produce oxidative stress through oxidation of membrane lipids, proteins, lipoproteins, and DNA (50) High-fat dairy Contains calcium, which binds bile acids and free fatty acids, decreasing oxidative damage in the gut; dairy fat contains fatty acids with potential inflammation-reducing properties, such as CLA, cis-and trans-palmitoleic acid, butyric acid, phytanic acid, and alpha-linolenic acid (51)(52)(53) Low-fat dairy Similar mechanisms to high-fat dairy (see above), with lower fat content Coffee and tea Tea contains flavonoids and antioxidants (e.g., epicatechin and quercetin) (54); coffee contains phytochemicals and antioxidants, such as javamide; both coffee and tea contain varying amounts of caffeine which inhibit secretion of IL-1β induced by adenine and N4-acetylcytidine (36,55)

Refined grains and starchy vegetables
Some processed grains contain emulsifiers, which potentially break down mucin in the gut leading to inflammation (58); and induce hyperglycemia (mechanisms described similar to those described above in 'Added Sugars') Supplement score † Comprises micro-nutrients, minerals, and vitamins solely from supplement intakes, some with similar mechanisms to those described above (e.g., iron as pro-oxidant, vitamins A, C, and E as antioxidants) Abbreviations: BMI, body mass index; CLA, conjugated linoleic acids; DIS, dietary inflammation score; hsCRP, high-sensitivity C-reactive protein; IL, interleukin; LIS, lifestyle inflammation score; METs, metabolic equivalents of task; Nrf2, Nuclear factor-erythroid 2 (NF-E2)-related factor 2; NSAID, nonsteroidal anti-inflammatory drug; PA1, plasminogen activator inhibitor-1; TNF, tumor necrosis factor * Weights are  coefficients from multivariable linear regression models conducted in a subset of the REGARDS cohort study (n = 639), and represent the average change in an inflammation biomarker score (sum of z-scores for circulating hsCRP, IL-6, IL-8, and IL-10 [the latter with a negative sign]) concentrations per one standard deviation increase in a dietary component or the presence of lifestyle component. Covariates in the final model to develop the weights included: age, sex, race (Black or White), education (high school graduate or less vs. some college or more), region (stroke belt, stroke buckle, or other region in the US), a comorbidity score (comprises a history of cancer, heart disease, diabetes mellitus, or chronic kidney disease), hormone replacement therapy (among women), total energy intake (kcal/day), season of baseline interview (Spring, Summer, Fall, or Winter), and regular use of aspirin, other non-steroidal anti-inflammatory drugs or lipid-lowering medications (≥ twice/wk); and all the dietary/lifestyle components in the DIS and LIS. For the NIH-AARP study, all dietary components were standardized based on their sex-specific distributions in the analytic cohort at baseline, and all lifestyle components were dummy variables. † All vitamin and mineral supplement intakes measured (from multivitamin/mineral and individual supplements) were ranked into quantiles of intake and assigned a value of 0 (low or no intake), 1, or 2 (highest intake) for hypothesized anti-inflammatory supplements (e.g., vitamin E), and 0 (low or no intake), -1, or -2 (highest intake) for hypothesized pro-inflammatory supplements (e.g., iron) Supplementary HRT, hormone replacement therapy; LIS, lifestyle inflammation score; NSAID, non-steroidal anti-inflammatory drug; NIH-AARP, National Institutes of Health-American Association for Retired Persons Diet and Health Study * Inflammation scores constructed as described in the text and Table 1; a higher score reflects a higher balance of pro-inflammatory exposures † From interaction term in the full Cox proportional hazards regression model, calculated using the Wald test ‡ Covariates in the DIS Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, smoking (current, former, or never), body mass index (in kg/m2; continuous), alcohol intake (non-drinker, moderate-drinker, or heavy-drinker), physical activity level (exercises not at all or rarely, 1 -2, or ≥ 3 times/wk), and total energy intake (kcal/day); history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, sex, and BMI were included in the SAS STRATA statement || Self-reported heart disease, diabetes mellitus, gallstone or gallbladder disease, or emphysema at baseline ¶ In a first degree relative # Heavy drinker defined as > 1 drink/day for women and > 2 drinks/day drinks for men; moderate drinker defined as 1 drink/day for women, 1 -2 drinks/day for men ** Aspirin/other NSAID use was ascertained in a subset of the baseline cohort that completed follow-up and risk factor questionnaires (N = 284,211 and N = 283,295, respectively) † † Colonoscopy history was assessed in remaining baseline cohort members from 2004 -2005; CRC cases diagnosed prior to 01/01/2004 were excluded from colonoscopy history stratification ‡ ‡ Covariates in the LIS Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, total energy intake (kcal/day), former smoker (yes/no), and the equally-weighted DIS; history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, and sex were included in the SAS STRATA statement Supplementary  Table 1); higher scores indicate a higher balance of pro-versus anti-inflammatory exposures † Covariates in the DIS-equal weight Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, smoking (current, former, or never), body mass index (in kg/m2; continuous), alcohol intake (non-drinker, moderate-drinker, or heavy-drinker), physical activity level (exercises not at all or rarely, 1 -2, or ≥ 3 times/wk), and total energy intake (kcal/day); history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, sex, and BMI were included in the SAS STRATA statement ‡ Covariates in the LIS-weight Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, total energy intake (kcal/day), former smoker (yes/no), and the equally-weighted DIS; history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, and sex were included in the SAS STRATA statement Supplementary inflammation score; HR, hazard ratio; LIS, lifestyle inflammation score; NIH-AARP, National Institutes of Health-American Association for Retired Persons Diet and Health Study * Inflammation score constructed as described in the text and Table 1; a higher score reflects a higher balance of pro-inflammatory exposures † Covariates in the DIS Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, smoking (current, former, or never), body mass index (in kg/m2; continuous), alcohol intake (non-drinker, moderate-drinker, or heavy-drinker), physical activity level (exercises not at all or rarely, 1 -2, or ≥ 3 times/wk), total energy intake (kcal/day), supplemental micronutrient score (calculated as described in Table 1); history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, sex, and BMI were included in the SAS STRATA statement Supplementary  (3), but the scoring was reversed such that a lower score was considered potentially higher risk; the EDIP was constructed as described by Tabung et al.(60) based on servings of intake † Covariates in the HEI Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, smoking (current, former, or never), body mass index (in kg/m2; continuous), alcohol intake (non-drinker, moderate-drinker, or heavy-drinker), physical activity level (exercises not at all or rarely, 1 -2, or ≥ 3 times/wk), and total energy intake (kcal/day); history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, sex, and BMI were included in the SAS STRATA statement ‡ Covariates in the EDIP Cox proportional hazards models included those described in footnote 'b', except for alcohol intake Supplementary  White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, total energy intake (kcal/day), former smoker (yes/no), and the equally-weighted DIS; history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, and sex were included in the SAS STRATA statement † Normal BMI: 18.  Table 1; a higher score reflects a higher balance of pro-inflammatory exposures † Covariates in the DIS Cox proportional hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, smoking (current, former, or never), body mass index (in kg/m2; continuous), alcohol intake (non-drinker, moderate-drinker, or heavy-drinker), physical activity level (exercises not at all or rarely, 1 -2, or ≥ 3 times/wk), and total energy intake (kcal/day); history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, sex, and BMI were included in the SAS STRATA statement ‡ Covariates in the LIS Cox proportional Hazards models were: age at entry (continuous), sex, race (Black, White, or other), education (less than high school and high school graduate, some college, or college graduate or higher), marital status (married or non-married), heart disease or history of stroke at baseline (yes/no), diabetes mellitus at baseline (yes/no), emphysema at baseline (yes/no), gallstone or gallbladder disease at baseline (yes/no), current hormone replacement therapy use (among women), family history of colorectal cancer in a first degree relative, history of colon polyp, total energy intake (kcal/day), former smoker (yes/no), and the equally-weighted DIS; history of CRC in a first degree relative, self-reported heart disease diagnosis, age at entry, and sex were included in the SAS STRATA statement Supplementary Alcohol, -carotene, caffeine, dietary fiber, folic acid, magnesium, thiamin, riboflavin, niacin, zinc, monounsaturated fats, polyunsaturated fats, -3 fatty acids, -6 fatty acids, selenium, isoflavones, flavan-3ol, flavones, flavanols, flavanones, anthocyanins, green or black tea, garlic, onion, turmeric, thyme & oregano, pepper, rosemary, eugenol, ginger, saffron, and vitamins A, B-6, C, D, & E Beer, wine, tea, coffee, darkyellow vegetables, leafy green vegetables, snacks, fruit juice, and pizza Leafy greens, tomatoes, apples and berries, deep yellow or orange vegetables and fruit, other fruits and real fruit juices, other vegetables, legumes, fish, poultry, high-and low-fat dairy, coffee and tea, nuts, supplement score Moderate alcohol intake, moderate or heavy physical activity Pro-inflammatory components Vitamin B-12, iron, trans fat, carbohydrates, cholesterol, total energy intake, protein, saturated fat, and total fat Processed meats, red meat, organ meat, fish (other than dark-meat fish), other vegetables (e.g., celery, mushrooms, green peppers, etc.), refined grains, high-energy beverages, lowenergy beverages, and tomatoes  Table 1