Association between C-reactive protein and risk of overall and 18 site-specific cancers in a Japanese case-cohort

Abstract

Background: Evidence of the association between chronic low-grade inflammation, as reflected by C-reactive protein (CRP) measurements, and cancer risk is equivocal. Specifically, few studies have examined this in uncommon cancers and Asian populations.

Methods: We utilised a case-cohort design consisting of multi-types of cancer (N = 3608), and a random subcohort (N = 4432) in a Japanese large population-based study, with a median follow-up time of 15.6 years, and measured baseline plasma CRP using high sensitivity assay. The hazard ratios (HRs) were estimated using weighted Cox proportional hazards methods.

Results: The multivariable-adjusted HR (95% confidence interval) for the top quartile of CRP was 1.28 (1.11‒1.48) (P_trend < 0.001) for overall cancer compared to the bottom quartile of CRP. Among site-specific cancers, higher CRP levels were associated with an increased risk of colorectal, lung, breast, biliary tract, and kidney cancer, and leukaemia. These positive associations remained among participants after >3 years' follow-up. Furthermore, subgroup analyses for overall cancer robustly showed a positive association with CRP levels, regardless of sex and obesity.

Conclusion: Our consistent findings suggested that chronic low-grade inflammation measured by CRP is associated with the risk of cancer.

Fig. 1. Hazard ratios (HRs) for total and site-specific cancer according to tertiles of C-reactive protein (CRP) by sex.

Tertiles based on the sex-specific distribution of subcohort (<0.3, 0.3 to <0.77, ≥0.77 for men and <0.22, 0.22 to <0.58, ≥0.58 for women). HR for the highest tertile of CRP, compared to the lowest tertiles of CRP originated from weighted Cox regression model stratified by area (9 areas), adjusted for age (continuous), BMI (<21, 21 to <25, 25 to <30, or ≥30 kg/m2), smoking (never, pack-years of ever or current smokers: <20, 20 to <40, or ≥40 cigarettes/day for men, never, ever for women), alcohol drinking (nondrinker, occasional, current drinker:1 to <150, 150 to <300, 300 to <450, or ≥450 g of ethanol per week for men, nondrinker, occasional, regular for women), physical activity (<1 or ≥1 days/week), history of diabetes, history of chronic disease and family history of cancer. P interaction originated from the model included the interaction term.