br Table br Multivariable logistic analysis to identify demo
Multivariable logistic analysis to identify demographic and health behaviors associated with high CRP levels (N = 2753).
Characteristics ORs (95%CI)*
Overall Women Men
Age group Reference Reference Reference
Marital Status Reference Reference Reference
Education Reference Reference Reference
< High school
Place of birth Reference Reference Reference
Language acculturation Reference Reference Reference
Cigarette smoking status Reference Reference Reference
Alcohol drinking Reference Reference Reference
Body mass index Reference Reference Reference
Physical activity Reference Reference Reference
*Age category, sex, marital status, education, place of birth, language acculturation, cigarette smoking status, alcohol drinking status, BMI category, and physical activity were adjusted in the logistic regression as appropriate.
Then, among those Doxorubicin born in the Mexico, we explored whether years of living in U.S. and age of immigration in U.S. might affect CRP levels. With years of living in U.S. increased, median levels of CRP was gra-dually increased from 3.00 to 3.77 (Table 3). Such increase was parti-cularly evident among men (P for trend < 0.001). In further multi-variable analysis, compared to those living in U.S. for less than 10 years, those living in U.S. for 10 to 20, 20 to 30, 30 to 40, and at least 40 years had increased likelihoods of having high levels of CRP. And a dose-response trend of increasing likelihoods was observed across the BMI category (P for trend = 0.018). Such relationship was observed in both men and women, but the trend was more evident among men (P for trend = 0.040). No similar association was observed for age of immigration in U.S. Finally, we investigated whether serum CRP levels were associated with incidents of cancer using multivariable Cox regression analysis and the results are summarized in Table 4. Among the 2753 study partici-pants, a total of 177 cases were identified during the follow-up as having experienced cancer. The major cancers included cancers in breast (N = 53), colorectal (N = 18), uterine (N = 15), cervical (N = 14), prostate (N = 12) and lung (N = 11). When CRP levels were dichotomized into two groups (high or low) using median CRP levels (3.20), high CRP levels were not significantly associated with cancer risk. However, in further quartile analysis using 25%, 50%, and 75% values of CRP levels as cutoff points, those in the highest quartile (≥ 6.40) had an increased cancer risk when compared to those in the lowest quartile of CRP levels (< 1.32) (HR = 1.88, 95%CI = 1.12, 3.13). In addition, a statistically significant dose–response trend was observed (P = 0.035). Similarly, when CRP levels were separated into 4 groups using clinical CRP levels , those with highest CRP levels
Median levels of CRP and risk of high CRP by years of living in U.S. among MAC study participants born in Mexico.
Characteristics CRP, median (range) P-value CRP, median (range) P-value CRP, median (range) P-value
Risk of high CRP levels
P for trend
*Adjusted by age category, gender, marital status, education level, language acculturation, cigarette smoking status, alcohol drinking status, BMI category, age of immigration, and physical activity levels as appropriate.
(> 10) had 1.86 fold increased risk of cancer (HR = 1.96, 95%CI = 1.01, 3.41) when compared with those with lowest CRP level (< 1). We also examined the impact of the time between cancer diag-nosis and CRP measures at baseline and incidents of cancer, using 5 years as the cutoff point. The significant association between CRP and cancer risk was only evident among participants whose cancer was diagnosed beyond 5 years of blood collection, but not among those whose cancer was diagnosed earlier.
The current study assessed the associations between serum CRP levels with demographic characteristics and health behaviors, as well as for the first time prospectively examined the relationship between pre-diagnostic serum CRP levels and all cancer risk among Mexican Americans. We found that median levels of serum CRP significantly differed by demographic characteristics and health behaviors. In quartile analysis, a significant dose response trend was observed be-tween increasing CRP levels and overall cancer risk (P < 0.05). We further found that the risk was more evident among those whose cancer was diagnosed beyond 5 years of blood collection.
We found that women had higher median levels of serum CRP than men did. Such observation is in consensus with the findings from pre-vious studies [18,19]. For example, in the Multiethnic Study of Ather-osclerosis (MESA) cohort, CRP levels were higher in women compared with men despite accounting for BMI and other common confounding variables . Our observation that median CRP levels increased with age group is consistent with the literature reports too [20,21]. Inter-estingly, in a recent publication, CRP could promote ageing via a Smad3‐dependent p21/p27 mechanism and through impairing cell re-generation by causing the G1 cell cycle arrest . In the National Social Life, health, and Aging Project (NSHAP), education levels were inversely associated with CRP levels , which is consistent with our findings. However, in the multivariable analysis, the education was not associated with high CRP levels. In NSHAP, marriage was found a protective factor against elevated CRP for men . In our study, si-milar but non-significant association was observed for men. In the multivariable analysis, men or married or living together had 1.42 fold increased likelihood of having high CRP levels (OR = 1.42, 95%CI: 0.88, 2.29).