The associations between whole grain and refined grain intakes and serum C-reactive protein

Low-grade inflammation, characterized by elevated inflammatory markers such as C-reactive protein (CRP), is a risk factor for chronic diseases including type 2 diabetes, coronary heart disease, certain cancers, dementia, depression, and all-cause mortality. Observational evidence links higher whole grain consumption to reduced risk of cardiometabolic diseases and mortality, and several studies suggest an inverse relationship between whole grain intake and low-grade inflammation. Findings for refined grains are inconsistent, with most studies reporting neutral or unfavorable effects. To clarify the role of grain type in systemic inflammation, this study examined cross-sectional associations between whole and refined grain intakes and serum high-sensitivity CRP (hs-CRP) among generally healthy older men and women in the KIHD cohort.

Prior observational studies generally show inverse associations between whole grain intake and inflammatory markers, including CRP and TNF-related markers, though some associations attenuate after adjustment for BMI or cereal fiber. Evidence for refined grains is mixed; some studies report no association with CRP but potential positive associations with other pro-inflammatory proteins (e.g., PAI-1). Randomized controlled trials have yielded conflicting findings: meta-analyses variously report no significant effect of whole grains on inflammatory biomarkers or modest reductions in hs-CRP and IL-6 compared with refined grains. Heterogeneity across studies may reflect differences in amounts and types of grains consumed, dietary assessment methods (food records vs FFQ vs 24-h recall), and baseline CRP levels in populations. Mechanistically, cereal fiber and other bioactive components of whole grains (phytochemicals, vitamins, minerals, unsaturated fats) may influence inflammation directly or via effects on adiposity, glycemic control, oxidative stress, and the gut microbiome.

Design and population: Cross-sectional analysis within the Kuopio Ischaemic Heart Disease Risk Factor Study (KIHD), a population-based cohort in eastern Finland. Participants were generally healthy men and women aged 53–73 years examined in 1999–2001. Of the invited cohort, 854 men and 920 postmenopausal women were examined; exclusions were applied for missing hs-CRP or diet data, hs-CRP >10 mg/L or leukocytes >11×10^9/L (acute inflammation), and diseases with inflammatory components (rheumatoid arthritis, colitis, diabetes, claudication, ischemic heart disease, cardiac insufficiency, stroke, cancer, or gallbladder/liver/pancreas diseases). Final analytic sample: 756 participants (391 men, 365 women). Measurements: Fasting blood samples were collected (8–10 AM); participants abstained from alcohol for 3 days and from smoking and eating for 12 h prior. Serum hs-CRP was measured using the IMMULITE chemiluminescent immunoassay (lower detection limit 0.1 mg/L; functional sensitivity 0.2 mg/L). Additional measurements included serum lipids, glucose, blood pressure, medical history, medications, smoking, alcohol, education (years), income, physical activity (KIHD 12-month leisure-time PA questionnaire), and anthropometrics (BMI calculated from measured height and weight). Dietary assessment: 4-day food records (three weekdays and one weekend day) with portion-size photo aids; records were reviewed with a nutritionist. Intakes were computed using NUTRICA 2.5 (Finnish food composition database). Whole grain defined per HEALTHGRAIN (whole kernel). Whole grain intake included whole grain ingredients in mixed dishes/recipes. Refined grain intake was total grain intake minus whole grain. Butter included butter as such and butter–vegetable oil mixes. Statistical analysis: ANCOVA and linear regression assessed associations of whole and refined grain intakes with hs-CRP. Grain intakes analyzed in quartiles and continuously; linear trends tested using median values of quartiles as continuous variables. Four adjustment models: Model 1 (age, gender, energy intake, examination year); Model 2 adds BMI, pack-years of smoking, leisure-time physical activity, education (years), alcohol (g/week); Model 3 adds dietary variables (fat quality ratio [SFA+TFA:PUFA+MUFA], fruits/vegetables/berries, red meat, dairy, fish, butter, vegetable oil margarine, eggs); Model 4 adds fiber from grains. Missing covariates imputed to cohort means (income n=48, pack-years n=9, alcohol n=1). Interactions by gender and BMI categories (<25, 25–<30, ≥30 kg/m^2) evaluated using stratified analyses and multiplicative interaction terms. Normality checked via histograms; two-sided alpha=0.05. Analyses conducted with SPSS 25.

- Mean intakes: whole grains 136 g/day (SD 80), refined grains 84 g/day (SD 46) overall; men consumed more of both than women. Mean hs-CRP: 1.77 mg/L (SD 1.7) in men and 1.96 mg/L (SD 1.8) in women. - Whole grains: Higher intake associated with lower hs-CRP. Per 50 g/day higher whole grain intake, hs-CRP was lower by 0.12 mg/L (95% CI 0.02–0.21 lower) after adjustment for lifestyle and dietary factors (Model 3). In Model 1, the difference between highest vs lowest quartile was −0.76 mg/L (95% CI −1.18 to −0.34). Associations attenuated with further adjustment, and became non-significant after adding fiber from grains (Model 4). - Refined grains: Not associated with hs-CRP in Model 1, but higher intake associated with higher hs-CRP after multivariable adjustment. Per 50 g/day higher refined grain intake, hs-CRP was higher by 0.23 mg/L (95% CI 0.08–0.38) in Model 3; slightly attenuated with grain fiber adjustment (Model 4, 0.20 mg/L; 95% CI 0.05–0.35). - Fiber: Grain fiber intake was inversely associated with hs-CRP; per 5 g/day higher grain fiber, hs-CRP was 0.11 mg/L lower (95% CI 0.01–0.21) in multivariable models (Model 3). Total fiber showed similar inverse associations. - Effect modification: No statistically significant interactions by gender or BMI. Point estimates suggested a stronger inverse association of whole grains with hs-CRP among women and a stronger positive association of refined grains with hs-CRP among men. Associations were broadly similar across BMI categories.

The study addressed whether types of grain intake relate to systemic low-grade inflammation in older adults. Findings indicate that higher whole grain intake relates to lower hs-CRP, while higher refined grain intake relates to higher hs-CRP after accounting for lifestyle and dietary factors. Adjustment for grain fiber substantially attenuated the whole grain association, implicating cereal fiber as a key contributor. Results align with several observational studies showing inverse relationships between whole grains or cereal fiber and inflammatory markers, though RCT evidence is mixed, likely due to differences in intervention doses, grain types, baseline inflammation, and assessment methods. Potential mechanisms include effects of cereal fiber and whole grain bioactives on oxidative stress, adiposity, glycemic control, and modulation of the gut microbiome, which may dampen systemic inflammatory signaling. BMI did not modify associations in this cohort, suggesting effects not solely mediated by adiposity. The positive association observed for refined grains, independent of many lifestyle and dietary covariates, supports the view that grain processing and fiber removal may contribute to higher inflammatory status.

Higher whole grain intake was associated with lower hs-CRP and higher refined grain intake with higher hs-CRP among generally healthy older Finnish adults. The attenuation of the whole grain association after adjustment for grain-derived fiber suggests cereal fiber may partly explain the inverse relationship. These results support public health recommendations to prefer whole grain products. Future research should elucidate biological pathways linking grain components to inflammation and include well-designed intervention trials to determine causal effects and the roles of specific grain types and fibers.

- Cross-sectional observational design precludes causal inference. - Only one inflammatory biomarker (hs-CRP) was measured; no consensus exists on the optimal biomarker panel for low-grade inflammation. - Study population consisted largely of elderly Caucasians from a single geographic area in eastern Finland, limiting generalizability. - Potential residual confounding due to healthy or unhealthy dietary/lifestyle patterns associated with whole or refined grain intake, despite extensive adjustment. - Lack of detailed breakdown of specific grain food sources and their proportional contributions in this dataset.