Prospective evaluation of C-reactive protein, smoking and lung cancer death in the Third National Health and Nutrition Examination Survey

  • Authors:
    • Marisa A. Bittoni
    • Brian C. Focht
    • Steven K. Clinton
    • Janet Buckworth
    • Randall E. Harris
  • View Affiliations

  • Published online on: August 31, 2015     https://doi.org/10.3892/ijo.2015.3141
  • Pages: 1537-1544
Metrics: Total Views: 0 (Spandidos Publications: | PMC Statistics: )
Total PDF Downloads: 0 (Spandidos Publications: | PMC Statistics: )


Abstract

Chronic inflammation plays an important role in lung carcinogenesis. Few prospective studies have examined associations between lung cancer, serum C-reactive protein (CRP), a measure of systemic inflammation, and inflammatory lifestyle factors, such as smoking and obesity. This study prospectively examined the relationship between CRP and lung cancer death and its interrelationships with several lifestyle factors. Baseline data on smoking and other lifestyle variables were collected for 8,950 participants in the Third National Health and Nutrition Examination Survey (NHANES III: 1988-1994). Baseline CRP levels were measured in serum samples by nephelometry. Mortality status was ascertained through probabilistic record matching using the National Death Index through 2006. Cox proportional hazard regression models were used to estimate hazard ratios (HRs) for CRP and lung cancer death, with adjustment for smoking and other variables. During 18 years of follow-up, 219 individuals died from lung cancer. Multivariate regression models revealed a dose-response effect for elevated CRP and risk of lung cancer death when adjusting for age, gender, BMI and smoking. Compared to individuals with CRP <3 mg/l, lung cancer death was significantly associated with elevated levels of CRP: HR=1.63 (95% CI=1.15-2.26) for 3-7 mg/l and HR=2.44 (95% CI=1.81‑3.45) for CRP >7 mg/l, P-trend <0.0001). The risk of lung cancer death for smokers increased 9-fold in adjusted models (P<0.0001). When stratified by gender and smoking status the effects of CRP were similar for smokers and males but did not reach statistical significance for females and non-smokers. This study supports a dose-dependent relationship between lung cancer death and CRP for males and smokers, but additional efforts are needed to better elucidate these relationships in women and non-smokers. The results suggest that CRP may emerge as a valuable tool in identifying high-risk subgroups of smokers for lung cancer prevention strategies.

Introduction

Lung cancer is the leading cause of cancer mortality in both men and women, accounting for ~28% of all cancer deaths. Approximately 221,200 new lung cancer cases and 158,040 deaths have been estimated to occur in 2015 (1). While the 1-year relative survival rate of lung cancer increased from 37% in 1975–1979 to 43% in 2003–2009, which is primarily due to improved surgical techniques and therapies, the 5-year survival rate remains very low at only 17% (averaged across all stages) (1). The continued low survival rate is primarily due to the lack of effective early detection methods for lung cancer, and because the disease is usually in the advanced stages by the time it is diagnosed (2,3). Overall, lung cancer incidence and mortality rates in the United States have modestly decreased in recent decades due to reduced smoking levels, improved screening techniques, chemoprevention, and treatment methods for lung cancer (4,5).

Although tobacco is the most critical risk factor for lung cancer development, evidence suggests that other environmental exposures, such as radon, asbestos, and dietary and lifestyle variables may also contribute to risk (3,6). However, the evidence for many of these factors has not been conclusive (6).

Based on past evidence, we propose that many factors contributing to cancer risk may interact within a complex network of inflammatory processes to mediate lung carcinogenesis (710). In addition to tobacco, environmental/occupational exposures, combined with inherited susceptibility variants, may impact carcinogen metabolism or inflammatory processes to modulate risk among smokers (1115).

This study sought to examine the potential of a biomarker of systemic inflammation, serum C-reactive protein (CRP), to predict the risk of lung cancer death in a large prospective cohort. CRP, which is an acute-phase protein that displays a rapid and distinct rise in its plasma concentration in response to acute inflammation, infection and tissue damage, has been examined in studies as a marker of chronic inflammatory states (1618). Most critically, CRP has been reported to be elevated in epidemiologic studies, in some studies marking the presence of prevalent cancer (19), but also associated with an increased risk of future cancer in otherwise healthy individuals (16,2022).

There have only been a few large, prospective studies that have focused on CRP and its relationship to lung cancer risk or mortality (20,23,24), but other studies have been modest in statistical power (14,25,26). This study sought to prospectively examine the relationship between CRP and lung cancer death, as well as the interrelationships with other lifestyle factors. It was hypothesized that increased levels of CRP would exhibit a positive relationship with lung cancer death and that the ability of CRP to predict poor outcomes in men and women would be enhanced by controlling for smoking status, obesity and demographic factors.

Materials and methods

The data obtained for this study were from the National Health and Nutrition Examination Survey III (NHANES III), which was conducted by the National Center for Health Statistics (NCHS) (27). The NHANES III study used a complex, multi-stage, stratified sample of civilian, non-institutionalized persons aged two months or greater. NHANES III was conducted from October 1988–1994 in two phases. In NHANES III, 39,695 persons were selected over the six years, and, of those, 33,994 (86%) were interviewed in their homes. All interviewed persons were invited to mobile examination centers (MECs) for a medical examination. A detailed description of design specifications and methods can be found elsewhere (27).

For the present study, data from NHANES III participants were used to prospectively examine the association between CRP and lung cancer death. The criteria for inclusion in this study were: a) age ≥40 years and b) study participants were cancer-free when they began the study (i.e., they had never been told by a doctor that they had cancer, with the exception of non-melanoma skin cancer). This prospective study followed cancer-free individuals from the time of entry into the NHANES III study until death or December 1, 2006.

Measures
Household Adult Questionnaire

The NHANES III Household Adult Questionnaire included all data collected during the household interviews for adults aged ≥17 years. Demographic and lifestyle data, such as age, sex, race, education, smoking and history of cancer, were obtained from the Household Adult Questionnaire. Interviews were conducted by field staff, who received intensive initial training and formal retraining, which continued throughout the study to ensure high skill levels. The data collection system was automated in phase 2, during which interviews were conducted using CAPI (computer-assisted personal interview). Details on survey instruments and forms, training manuals and data collection procedures are published elsewhere (27).

MEC physical examination

Blood and urine specimens were obtained at the MEC within one month of the interview, where several lab tests and measurements were performed, including anthropometric body measurements, X-rays and electrocardiography. While some of the blood and urine analyses were performed in the MEC laboratory, most were conducted by contract laboratories (28). For those who could not visit the MEC, a limited home examination was conducted.

CRP analysis

CRP was measured in serum samples by nephelometry and stored at −70°C within 2 months of collection. CRP levels were analyzed using a fully automated Behring Nephelometer Analyzer System (Behring Diagnostics, Inc., Somerville, NJ, USA). Additional details about the specific methods for quantifying CRP are provided elsewhere (21).

Ascertainment of mortality through National Death Index Linkage

The outcome measure for this prospective study was death from lung cancer, which was ascertained through a record linkage process with mortality data from the National Death Index (NDI). The NHANES III database was linked with NDI mortality records through December 31, 2006, which resulted in a follow-up time for mortality that ranged from 6 to 12 years (29,30). This linkage was performed through probabilistic matching using social security number, first, middle and last name, year/day of birth, and additional fields described elsewhere (30). Mortality status could also be ascertained if the death certificate was received directly from NCHS and matched with an NHANES III survey record based on a participant's name and other linkable information, such as occupation (31). The underlying cause of death was established from ICD-9 codes through 1998 and ICD-10 codes for 1999–2000, but the final cause of death was determined by the ICD-10 codes after adjusting for changes between the coding systems. At the end of the follow-up period, 3,384 adult NHANES III participants were determined to be deceased through the NDI linkage process (30).

Statistical analysis
Descriptive statistics

Descriptive statistics were displayed for all variables, with frequencies and percentages shown for discrete variables and means and standard deviations (SDs) displayed for continuous variables. Tests of statistical significance were also shown, with Chi-square tests performed for discrete variables and t-tests performed for continuous variables. The level of statistical significance was α=0.05.

Regression analysis

Cox proportional hazards regression analysis was performed to assess the relationship between CRP levels and the study outcome of time to lung cancer death. Effects were quantified by estimating hazard ratios (HRs) and associated P-values from univariate and multivariate regression models. Regression models were developed to examine the relationship between CRP and lung cancer death, as well as covariate factors, with a dichotomous indicator variable denoting whether the person died (dead=1, not dead=0, i.e., right censored). CRP was classified into 3 levels: <3 mg/l, 3–7 mg/l and >7 mg/l, based on cutpoints previously defined in the literature (32,33). Graphical checks of the data and covariates revealed that the proportional hazards assumption was met. Fig. 2 shows that the hazards by CRP levels were proportional.

The effect of demographic and other risk factors on the relationship between inflammation and lung cancer death was assessed from the following factors: age (years), gender (male, female), race (white, African American/other), education (<high school, ≥high school), smoking status (ever smoked ≥100 cigarettes, never smoked), pack-years (number of packs smoked per day times number of years smoked), BMI (kg/m2), and waist-hip ratio (WHR). All analyses were conducted using SAS 9.3.

Results

Participants

Out of 10,735 participants who met the eligibility criteria for inclusion (age ≥40 and cancer-free at baseline), a total of 1,785 participants did not have a measurement for CRP, and were, therefore, excluded, providing a total of 8,950 participants for analysis in this study (Fig. 1). There were 219 deaths from lung cancer over the follow-up period of 218 months. The total follow-up time for lung cancer deaths was 199 months and for the remaining cohort was 218 months.

Demographic, lifestyle and anthropometric characteristics

Demographic characteristics of participants at baseline are displayed in Table I. Over half of the participants in this study were female (51.9%), but two-thirds of the lung cancer deaths occurred in males (P<0.001). Over 70% of participants in the study cohort were white and almost one-fourth were black. Over half of participants had a high school education or greater, with over half of lung cancer deaths not having a high school education. The mean age at baseline for the entire cohort was 61 years, but those who died of lung cancer were older (64.7 versus 61.0 years, P<0.001).

Table I

Demographic and clinical characteristics of participants.

Table I

Demographic and clinical characteristics of participants.

VariableLung cancer deaths (cases) (N=219)
n (%)
Non-cases (N=8,731)
n (%)
Total (N=8,950)
n (%)
Genderb
 Male144 (65.7)4,161 (47.7)4,305 (48.1)
 Female75 (34.3)4,570 (52.3)4,645 (51.9)
Race
 White154 (70.3)6,402 (73.3)6,556 (73.3)
 African American62 (28.3)2,091 (24.0)2,153 (24.1)
 Other3 ( 1.4)238 (2.7)241 (2.7)
Education
 <High school113 (52.1)4138 (47.7)4,251 (47.5)
 ≥High school104 (47.9)4,529 (52.3)4,699 (52.5)
Ageb
 40–49 years20 (9.1)2,378 (27.2)2,398 (26.8)
 50–59 years39 (17.8)1,676 (19.2)1,715 (19.2)
 60–69 years84 (38.4)2,024 (23.2)2,108 (23.6)
 70–79 years61 (27.9)1,497 (17.2)1,558 (17.4)
 ≥80 years15 (6.9)1,156 (13.2)1,171 (13.1)
 Mean ± SD64.7±10.061.0±14.061.0±14.0
Body mass index (BMI)b
 <18.5 underweight7 (3.2)190 (2.2)197 (2.2)
 18.5–24.9 normal100 (45.7)2,740 (31.4)2,840 (31.7)
 25.0–29.9 oveweight74 (33.8)3,366 (38.6)3,440 (38.4)
 >30 obese38 (17.4)2,435 (27.9)2,473 (27.6)
 BMI (kg/m2) Mean ± SDc25.7±4.627.6±5.627.6±5.6
 Waist-hip ratio Mean ± SDa0.96±0.10.95±0.10.95±0.1
Smoking statusc
 Ever smoked ≥100 cig202 (92.2)4,705 (53.9)4,907 (54.8)
  Former84 (38.4)2,845 (32.6)2,929 (32.7)
  Current18 (53.8)1,860 (21.3)1,978 (22.1)
 Never smoked17 (7.8)4,026 (46.1)4,043 (45.2)
Pack-years (Mean ± SD)a24.3±23.418.5±20.618.8±20.8
 CRP levelsb
 CRP<3 mg/l108 (49.3)5,297 (60.7)5,405 (60.4)
 CRP 3–7 mg/l50 (22.8)1,747 (20.0)1,797 (20.1)
 CRP >7 mg/l61 (27.9)1,687 (19.3)1,748 (19.5)
 Mean ± SDa6.4±7.95.4±9.05.4±9.0

a P<0.05;

b P<0.001;

c P<0.0001.

Anthropometric and lifestyle risk factors are also shown in Table I. Upon examining BMI by established categories that define obesity (34), it was evident that those who died of lung cancer were significantly less likely to be classified as overweight or obese compared to the remaining cohort (P<0.001). The mean BMI at enrollment was slightly lower for those who had lung cancer deaths versus the remaining cohort (25.7 and 27.6, respectively, P<0.0001). The mean WHR was slightly higher for lung cancer deaths versus those who did not die of lung cancer (0.96 and 0.95, respectively; P=0.04).

While almost 55% of all participants had ever smoked >100 cigarettes in their lifetime, a large majority (92.2%) of those who died of lung cancer had ever smoked (P<0.0001). Of those, almost 54% were current smokers at baseline compared to only 21% of the remaining cohort. Males were also more likely to smoke than females (63 versus 37%, respectively, not shown). Those who died of lung cancer also had a significantly higher number of mean pack-years smoked (24.3 versus 18.5; P<0.05).

Mean CRP levels were significantly higher at enrollment for those who subsequently died of lung cancer versus the remaining cohort (P<0.05). When examining CRP by the previously employed cutoff values of low, moderate and high CRP levels (<3, 3–7, >7 mg/l), those who died of lung cancer had higher overall CRP levels. Over half (50.7%) of all lung cancer deaths had a CRP level of 3 or greater compared to only 39.3% the remaining cohort, which was statistically significant (P<0.001).

The overall survival graph by CRP levels is shown in Fig. 2. The overall survival probabilities decreased from 1.0 at study start to ~0.97, 0.96 and 0.94, respectively, for the lowest to highest CRP levels at the end of follow-up. Therefore, the group with the lowest CRP levels had a higher probability of survival by the end of the study.

Regression analysis

The results of the Cox proportional hazards regression analyses are shown in Tables II and III. Table II shows that individuals with CRP levels 3–7 mg/l had a 1.63 times higher risk of lung cancer death compared to the reference group (CRP <3 mg/l), which was statistically significant (95% CI=1.15–2.26). Individuals with CRP >7 mg/l had 2.44 times the risk of dying from lung cancer, which was also statistically significant (95% CI=1.81–3.45). A dose-response effect was observed for higher CRP levels and lung cancer death, even when adjusting for other factors such as smoking (P-trend <0.0001). Notably, the HRs for CRP and lung cancer were adjusted for the dominant lung cancer risk factor of smoking, for which the HR was 9.83, which revealed a greatly increased risk of lung cancer death (95% CI=5.65–15.59).

Table II

Cox proportional hazards regression overall model of CRP and lung cancer.

Table II

Cox proportional hazards regression overall model of CRP and lung cancer.

MultivariateCases
n (%)
Controls
n (%)
HR (95% CI)
CRP <3 mg/la108 (49.3)5,297 (60.7)-
CRP 3–7 mg/l50 (22.8)1,747 (20.0)1.63 (1.15–2.26)
CRP >7 mg/l67 (27.9)1,687 (19.3)2.44 (1.81–3.45)
Smoker (ever vs never)202 (92.2)4,705 (53.9)9.83 (5.65–15.59)
BMI (kg/m2)219 (100.0)8,731 (100.0)0.92 (0.89–0.95)

a P-trend <0.0001.

b Models were also adjusted for age (HR=1.05; 95% CI=1.04–1.06) and gender (male vs female) (HR=1.37; 95% CI=1.03–1.82).

Table III

Cox proportional hazards regression stratified models of CRP and lung cancer death by smoking status and gender.a

Table III

Cox proportional hazards regression stratified models of CRP and lung cancer death by smoking status and gender.a

Ever smokersCases
n (%)
Controls
n (%)
HR (95% CI)b
CRP <3 mg/l99 (49.0)2,801 (59.5)-
CRP 3–7 mg/l44 (21.8)970 (20.6)1.54 (1.08–2.20)
CRP >7 mg/l59 (29.2)934 (19.9)2.58 (1.85–3.61)
BMI (kg/m2)200 (100.0)4,690 (100.0)0.91 (0.89–0.95)

Never smokersCases
n (%)
Controls
n (%)
HR (95% CI)

CRP <3 mg/lc9 (52.9)2,496 (62.0)-
CRP 3–7 mg/l6 (35.3)777 (19.3)2.46 (0.87–7.26)
CRP >7 mg/l2 (11.8)753 (18.7)-
BMI (kg/m2)17 (100.0)4,011 (100.0)0.96 (0.99–1.00)

MalesCases
n (%)
Controls
n (%)
HR (95% CI)b

CRP <3 mg/l68 (47.2)2,740 (65.9)-
CRP 3–7 mg/l34 (23.6)779 (8.7)1.89 (1.26–2.90)
CRP >7 mg/l42 (29.2)642 (15.5)3.04 (2.25–4.92)
Smoker (ever vs never)136 (94.4)2,963 (71.2)6.39 (3.10–13.00)
BMI (kg/m2)142 (100.0)4,154 (100.0)0.89 (0.86–0.93)

FemalesCases
n (%)
Controls
n (%)
HR (95% CI)

CRP <3 mg/l40 (56.0)2,557 (56.0)-
CRP 3–7 mg/l16 (21.3)968 (21.2)1.26 (0.62–1.99)
CRP >7 mg/l19 (25.3)1,045 (22.9)1.40 (0.66–2.21)
Smoker (ever vs never)66 (88.0)1,742 (38.1)12.51 (6.25–25.52)
BMI (kg/m2)75 (100.0)4,547 (100.0)0.96 (0.92–1.01)

a HR, hazard ratio; CRP, C-reactive protein; BMI, body mass index; models were also adjusted for age and gender.

b P-trend <0.0001 for males and smokers.

c CRP in never smokers was dichotomized as CRP <3 mg/l and CRP ≥3 mg/l.

Due to the powerful effect of smoking in this study and the fact that the majority of men (63%) were smokers, the models were stratified according to smoking and gender. Table III shows the multivariate model by smoking status (ever versus never smoked). For ever smokers, the HRs for CRP revealed a statistically significant dose-response effect (P-trend <0.0001), with higher levels of CRP indicating a higher risk of lung cancer death, similar to the results in Table II (HR=1.54 and 2.58 for CRP 3–7 mg/l and CRP >7 mg/l, respectively). For never smokers, individuals with CRP >3 mg/l (CRP was dichotomized due to the small sample size), were at increased risk (HR=2.46, 95% CI=0.87–7.26), but the estimate did not reach statistical significance.

The results for males were similar to those of smokers, although the HRs were higher (1.89 and 3.04 for CRP 3–7 mg/l and CRP >7 mg/l, respectively), with a dose-response effect (P-trend <0.0001). The HR estimates for females, however, were not significant at either level of CRP. Interestingly, the HR for female smokers was statistically significant (HR=12.51; 95% CI=6.25–25.52) and was double that of males (HR=6.39; 95% CI=3.10–13.00).

The results for BMI showed a reduced risk of lung cancer death (HR=0.92; 95% CI=0.89–0.95) in the overall model (Table II). In the stratified models (Table III), the results for BMI revealed the lowest HR for males, which was statistically significant (HR=0.89; 95%CI=0.86–0.93). The BMI for ever smokers was slightly higher but still statistically significant (HR=0.91; 95%CI=0.89–0.95). The BMI HRs for females and never smokers were not statistically significant.

Discussion

Our prospective study examined the interrelationships between CRP, lifestyle risk factors and risk of lung cancer death. CRP is an acute-phase protein that is produced in the liver and whose main biologic function is to recognize pathogens and damaged cells in the host, as well as to mediate their elimination by recruiting the complement system and phagocytic cells (17). CRP is suggested as a prognostic factor for the risk of several cancers (35), including lung cancer (14,20,24) and chronic lung disease (18,36). We hypothesized that CRP concentrations may provide an integrative biomarker of inflammatory responses of the host to tobacco and other environmental factors, which may predict those at greater risk of death from lung cancer.

We observed a significant 2.4-fold positive association between baseline CRP concentrations and subsequent risk of lung cancer death. Furthermore, a significant dose-dependent relationship emerged. Our study supports earlier findings indicating a positive association between CRP and lung cancer. In a prospective study by Allin et al (24), a 2-fold increased risk of lung cancer was observed among incident cases with a similar dose-response relationship. Chaturvedi et al (20) and Shiels et al (23) also observed a 2- to 3-fold increased risk of CRP and incidence of lung cancer with a dose-response effect. A meta-analysis of 6 prospective studies of CRP and incident lung cancer found a lower risk ratio of 1.32, but the authors indicated that considerable heterogeneity across studies may have been related to differential effects in the different populations, and there may also have been residual confounding factors (19). Of note is that previous studies did not assess other potential modulating factors of CRP, such as obesity and other lifestyle factors.

The interrelationships between tobacco use and CRP were investigated in the regression analysis. Past studies show that lung cancer risk is clearly related to smoking (37,38), which is consistent with the result of our multivariate model showing a 9-fold increased risk of lung cancer for those who had ever smoked versus never smoked. Allin et al (24) did not report the HRs for smoking and Chaturvedi et al (20) and Shiels et al (23) controlled for smoking through matching. When stratified by smoking, the HR remained similar to that of smokers, but did not reach statistical significance, which may have been due to the smaller sample of non-smokers. This finding warrants further investigation, however, since others have shown that there may be different mechanisms in smokers versus non-smokers (39).

When the multivariate model was stratified by gender, the relationship between CRP and lung cancer death showed differential effects. While a strong positive relationship between CRP and lung cancer death remained for males, the effect was weakened and not significant for females. Interestingly, the HR for smoking in females was twice that of males (12.51 versus 6.39). CRP, however, was not shown to be statistically significant for females, which suggests that lung cancer in females in this study may have been more related to carcinogens in tobacco smoke than inflammation (as measured by CRP), whereas the results for males showed that lung cancer appeared to be related to both smoking and inflammation. Khera et al (40) investigated CRP differences between males and females and observed that females tended to have higher CRP levels than males. They proposed that different cutoff levels may be useful in studies assessing CRP and health outcomes based on gender, but future research is needed. We conclude that the interrelationships between tobacco, smoking, markers of inflammation and risk of cancer in men and women may differ and warrant further investigation.

Obesity emerged as a significant factor with regards to lung cancer death in this study, demonstrating a protective effect for males and smokers but not always for females and non-smokers. One reason for this occurrence may be that smoking is associated with reduced BMI and males were more likely to smoke than females in this study. Past studies assessing obesity and lung cancer have primarily shown a protective effect, but many studies have not stratified between males and females (39). This issue should, therefore, be further investigated in future studies. Another difference in the stratified models was that the HR for BMI was not statistically significant for non-smokers, and only approached significance for females. It is known that smokers tend to be leaner than non-smokers, which may explain why BMI was no longer significantly related to lung cancer death in non-smokers. Male smokers also tend to be leaner than female smokers, which may explain why the protective BMI did not quite reach statistical significance for females. Others have noted that smoking may mask the effect of BMI (41). Also, since males were more likely to smoke in this study, and smoking is usually correlated with other inflammatory factors, such as increased CRP and decreased BMI, further studies are needed to specifically examine the differences between males and females, and to determine which mechanisms may be causing these differential results. Possible assessments include environmental exposures, histologic differences, and biological or genetic factors (39).

The primary strength of this study was the novelty of examining the CRP-lung cancer relationship and associations with lifestyle factors in the NHANES cohort, which is a large national probability sample. Our results are, therefore, more likely to be generalizable to the US population. NHANES was a well-planned and executed study that included a detailed plan, well-trained staff and implemented many quality control methods in the data collection phase in order to ensure high quality data and a consistent data collection process. CRP and other study data were obtained from blood draws that were analyzed by an independent lab; anthropometric data were obtained through physical exams conducted at baseline versus self-report.

There were also several limitations in this study. Several variables, including demographic/lifestyle data, were based on participant recall. The surveys were primarily done in person, however, using computer-assisted techniques, with standard cards, pictures and other probing methods used to help participants recall each data item as accurately as possible. Self-reported smoking in NHANES was previously shown to accurately reflect smoking status (42). The lack of high-sensitivity assays to measure CRP limited the lower threshold of CRP that could be detected and, therefore, limited the number of strata we could use for the analysis. Our study outcome of lung cancer death versus lung cancer incidence was another limitation, but since lung cancer has a short survival time, our results are believed to be reasonably similar to what lung cancer incidence data would have produced. Also, the outcome ascertainment for lung cancer death was not histologically confirmed, but was obtained through a death certificate linkage process. This could have led to misclassification since individuals with lung cancer may have died of other causes. As a result, we may have missed some of the lung cancer cases. Death certificates also do not reflect histology, so we could not conduct any histology analyses.

In conclusion, several important results emerged from this study, which can be used to make recommendations for reducing the risk of lung cancer death. First, this study supports the concept that a higher concentration of CRP, particularly in those who smoke, may identify a group at particularly high risk of developing lung cancer. Most critically, our study supports that the CRP-lung cancer relationship is quantitatively dose-dependent and CRP may serve as a biomarker in prevention studies, particularly for men exposed to tobacco smoke. Yet additional efforts are necessary to better elucidate the relationships in women and non-smokers. While effective interventions to quit smoking are greatly needed, CRP may emerge as a valuable tool to identify particularly high-risk subgroups of smokers for lung cancer prevention strategies.

References

1 

American Cancer Society. Cancer Facts and Figures 2015. American Cancer Society; Atlanta: 2015, http://www.cancer.org/acs/groups/content/@editorial/documents/document/acspc-044552.pdf.

2 

Boffetta P and Trichopoulos D: Cancer of the lung, larynx and pleura. Cancer Epidemiology. Adami H, Hunter D and Trichopolous D: Oxford University Press; New York, NY: 2008

3 

Dela Cruz CS, Tanoue LT and Matthay RA: Lung cancer: Epidemiology, etiology, and prevention. Clin Chest Med. 32:605–644. 2011. View Article : Google Scholar : PubMed/NCBI

4 

Szabo E, Mao JT, Lam S, Reid ME and Keith RL: Chemoprevention of lung cancer: Diagnosis and management of lung cancer. 3rd edition. American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. 2013, http://www.guideline.gov/content.aspx?id=46162.

5 

Anthonisen NR, Skeans MA, Wise RA, Manfreda J, Kanner RE and Connett JE: Lung Health Study Research Group: The effects of a smoking cessation intervention on 14.5-year mortality: A randomized clinical trial. Ann Intern Med. 142:233–239. 2005. View Article : Google Scholar : PubMed/NCBI

6 

World Cancer Research Fund/American Institute for Cancer Research. Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. AICR; Washington, DC: 2007, http://www.dietandcancerreport.org/cancer_resource_center/downloads/Second_Expert_Report_full.pdf.

7 

Hanahan D and Weinberg RA: The hallmarks of cancer. Cell. 100:57–70. 2000. View Article : Google Scholar : PubMed/NCBI

8 

Hanahan D and Weinberg RA: Hallmarks of cancer: The next generation. Cell. 144:646–674. 2011. View Article : Google Scholar : PubMed/NCBI

9 

Mantovani A, Allavena P, Sica A and Balkwill F: Cancer-related inflammation. Nature. 454:436–444. 2008. View Article : Google Scholar : PubMed/NCBI

10 

Balkwill F: Tumour necrosis factor and cancer. Nat Rev Cancer. 9:361–371. 2009. View Article : Google Scholar : PubMed/NCBI

11 

Thorgeirsson TE, Geller F, Sulem P, Rafnar T, Wiste A, Magnusson KP, Manolescu A, Thorleifsson G, Stefansson H, Ingason A, et al: A variant associated with nicotine dependence, lung cancer and peripheral arterial disease. Nature. 452:638–642. 2008. View Article : Google Scholar : PubMed/NCBI

12 

Bach PB, Kattan MW, Thornquist MD, Kris MG, Tate RC, Barnett MJ, Hsieh LJ and Begg CB: Variations in lung cancer risk among smokers. J Natl Cancer Inst. 95:470–478. 2003. View Article : Google Scholar : PubMed/NCBI

13 

Amos CI, Wu X, Broderick P, Gorlov IP, Gu J, Eisen T, Dong Q, Zhang Q, Gu X, Vijayakrishnan J, et al: Genome-wide association scan of tag SNPs identifies a susceptibility locus for lung cancer at 15q25.1. Nat Genet. 40:616–622. 2008. View Article : Google Scholar : PubMed/NCBI

14 

Siemes C, Visser LE, Coebergh JW, Splinter TA, Witteman JC, Uitterlinden AG, Hofman A, Pols HA and Stricker BH: C-reactive protein levels, variation in the C-reactive protein gene, and cancer risk: The Rotterdam Study. J Clin Oncol. 24:5216–5222. 2006. View Article : Google Scholar : PubMed/NCBI

15 

Huang YT, Lin X, Liu Y, Chirieac LR, McGovern R, Wain J, Heist R, Skaug V, Zienolddiny S, Haugen A, et al: Cigarette smoking increases copy number alterations in nonsmall-cell lung cancer. Proc Natl Acad Sci USA. 108:16345–16350. 2011. View Article : Google Scholar : PubMed/NCBI

16 

Allin KH and Nordestgaard BG: Elevated C-reactive protein in the diagnosis, prognosis, and cause of cancer. Crit Rev Clin Lab Sci. 48:155–170. 2011. View Article : Google Scholar : PubMed/NCBI

17 

Volonakis J: Human C-reactive protein: expression, structure, and function. Mol Immunol. 38:189–197. 2001. View Article : Google Scholar

18 

Vagulienė N, Žemaitis M, Miliauskas S, Urbonienė D, Šitkauskienė B and Sakalauskas R: Comparison of C-reactive protein levels in patients with lung cancer and chronic obstructive pulmonary disease. Medicina (Kaunas). 47:421–427. 2011.

19 

Heikkila K, Ebrahim S and Lawlor DA: A systematic review of the association between circulating concentrations of C-reactive protein and cancer. J Epidemiol Community Health. 61:824–833. 2007. View Article : Google Scholar

20 

Chaturvedi AK, Caporaso NE, Katki HA, Wong HL, Chatterjee N, Pine SR, Chanock SJ, Goedert JJ and Engels EA: C-reactive protein and risk of lung cancer. J Clin Oncol. 28:2719–2726. 2010. View Article : Google Scholar : PubMed/NCBI

21 

Gunter EW and McQuillan G: Quality control in planning and operating the laboratory component for the Third National Health and Nutrition Examination Survey. J Nutr. 120(Suppl 11): 1451–1454. 1990.PubMed/NCBI

22 

Tsilidis KK, Branchini C, Guallar E, Helzlsouer KJ, Erlinger TP and Platz EA: C-reactive protein and colorectal cancer risk: A systematic review of prospective studies. Int J Cancer. 123:1133–1140. 2008. View Article : Google Scholar : PubMed/NCBI

23 

Shiels MS, Pfeiffer RM, Hildesheim A, Engels EA, Kemp TJ, Park JH, Katki HA, Koshiol J, Shelton G, Caporaso NE, et al: Circulating inflammation markers and prospective risk for lung cancer. J Natl Cancer Inst. 105:1871–1880. 2013. View Article : Google Scholar : PubMed/NCBI

24 

Allin KH, Bojesen SE and Nordestgaard BG: Baseline C-reactive protein is associated with incident cancer and survival in patients with cancer. J Clin Oncol. 27:2217–2224. 2009. View Article : Google Scholar : PubMed/NCBI

25 

Il'yasova D, Colbert LH, Harris TB, Newman AB, Bauer DC, Satterfield S and Kritchevsky SB: Circulating levels of inflammatory markers and cancer risk in the health aging and body composition cohort. Cancer Epidemiol Biomarkers Prev. 14:2413–2418. 2005. View Article : Google Scholar : PubMed/NCBI

26 

Trichopoulos D, Psaltopoulou T, Orfanos P, Trichopoulou A and Boffetta P: Plasma C-reactive protein and risk of cancer: A prospective study from Greece. Cancer Epidemiol Biomarkers Prev. 15:381–384. 2006. View Article : Google Scholar : PubMed/NCBI

27 

NCHS Plan & Operation of the Third National Health and Nutrition Examination Survey, 1988–94. Vital and Health Statistics. 32:1–416. 1992.

28 

U.S. Department of Health and Human Services. Third National Health and Nutrition Examination Survey - NHANES III Examination Data File. Public Use Data File Documentation Number 76200. Centers for Disease Control and Prevention; Hyattsville, MD: 1996, http://wonder.cdc.gov/wonder/sci_data/surveys/hanes/hanes3/type_txt/exam1.asp.

29 

Madans JH and Hunter EL: Improving and integrating data systems for public health surveillance. J Public Health Manag Pract. 2:42–44. 1996. View Article : Google Scholar

30 

NCHS (National Center for Health Statistics). NHANES III Mortality File: Matching Methodology. 2005, ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/datalinkage/linked_mortality/archived_files/mort_2004_release/matching_methodology_nhanes3_final.pdfftpsimpleftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/datalinkage/linked_mortality/archived_files/mort_2004_release/matching_methodology_nhanes3_final.pdf.

31 

NCHS National Health and Nutrition Examination Survey III Linked Mortality File. http://www.cdc.gov/nchs/data_access/data_linkage/mortality/nhanes3_linkage.htm. Accessed October 15, 20112006

32 

Imayama I, Ulrich CM, Alfano CM, Wang C, Xiao L, Wener MH, Campbell KL, Duggan C, Foster-Schubert KE, Kong A, et al: Effects of a caloric restriction weight loss diet and exercise on inflammatory biomarkers in overweight/obese postmenopausal women: A randomized controlled trial. Cancer Res. 72:2314–2326. 2012. View Article : Google Scholar : PubMed/NCBI

33 

Pearson TA, Mensah GA, Alexander RW, Anderson JL, Cannon RO III, Criqui M, Fadl YY, Fortmann SP, Hong Y, Myers GL, et al: Centers for Disease Control and Prevention; American Heart Association: Markers of inflammation and cardiovascular disease: application to clinical and public health practice: A statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation. 107:499–511. 2003. View Article : Google Scholar : PubMed/NCBI

34 

World Health Organization. Physical status: the use and interpretation of anthopometry. Report of a WHO expert committee. (WHO Technical Report Series 854). World Health Organization; Geneva: 1995, http://www.who.int/childgrowth/publications/physical_status/en/.

35 

Guo YZ, Pan L, Du CJ, Ren DQ and Xie XM: Association between C-reactive protein and risk of cancer: A meta-analysis of prospective cohort studies. Asian Pac J Cancer Prev. 14:243–248. 2013. View Article : Google Scholar : PubMed/NCBI

36 

López-Campos JL, Calero C, Rojano B, López-Porras M, Sáenz-Coronilla J, Blanco AI, Sánchez-López V, Tobar D, Montes-Worboys A and Arellano E: C-reactive protein and serum amyloid a overexpression in lung tissues of chronic obstructive pulmonary disease patients: A case-control study. Int J Med Sci. 10:938–947. 2013. View Article : Google Scholar : PubMed/NCBI

37 

Wynder EL and Graham EA: Tobacco smoking as a possible etiologic factor in bronchiogenic carcinoma; a study of 684 proved cases. J Am Med Assoc. 143:329–336. 1950. View Article : Google Scholar : PubMed/NCBI

38 

Howlader N, Noone AM, Krapcho M, et al: SEER Cancer Statistics Review, 1975–2008. National Cancer Institute; Bethesda, MD: 2010, http://seer.cancer.gov/archive/csr/1975_2008/.

39 

Smith L, Brinton LA, Spitz MR, Lam TK, Park Y, Hollenbeck AR, Freedman ND and Gierach GL: Body mass index and risk of lung cancer among never, former, and current smokers. J Natl Cancer Inst. 104:778–789. 2012. View Article : Google Scholar : PubMed/NCBI

40 

Khera A, McGuire DK, Murphy SA, Stanek HG, Das SR, Vongpatanasin W, Wians FH Jr, Grundy SM and de Lemos JA: Race and gender differences in C-reactive protein levels. J Am Coll Cardiol. 46:464–469. 2005. View Article : Google Scholar : PubMed/NCBI

41 

Yang Y, Dong J, Sun K, Zhao L, Zhao F, Wang L and Jiao Y: Obesity and incidence of lung cancer: A meta-analysis. Int J Cancer. 132:1162–1169. 2013. View Article : Google Scholar

42 

Caraballo RS, Giovino GA, Pechacek TF and Mowery PD: Factors associated with discrepancies between self-reports on cigarette smoking and measured serum cotinine levels among persons aged 17 years or older: Third National Health and Nutrition Examination Survey, 1988–1994. Am J Epidemiol. 153:807–814. 2001. View Article : Google Scholar : PubMed/NCBI

Related Articles

Journal Cover

October-2015
Volume 47 Issue 4

Print ISSN: 1019-6439
Online ISSN:1791-2423

Sign up for eToc alerts

Recommend to Library

Copy and paste a formatted citation
x
Spandidos Publications style
Bittoni MA, Focht BC, Clinton SK, Buckworth J and Harris RE: Prospective evaluation of C-reactive protein, smoking and lung cancer death in the Third National Health and Nutrition Examination Survey. Int J Oncol 47: 1537-1544, 2015
APA
Bittoni, M.A., Focht, B.C., Clinton, S.K., Buckworth, J., & Harris, R.E. (2015). Prospective evaluation of C-reactive protein, smoking and lung cancer death in the Third National Health and Nutrition Examination Survey. International Journal of Oncology, 47, 1537-1544. https://doi.org/10.3892/ijo.2015.3141
MLA
Bittoni, M. A., Focht, B. C., Clinton, S. K., Buckworth, J., Harris, R. E."Prospective evaluation of C-reactive protein, smoking and lung cancer death in the Third National Health and Nutrition Examination Survey". International Journal of Oncology 47.4 (2015): 1537-1544.
Chicago
Bittoni, M. A., Focht, B. C., Clinton, S. K., Buckworth, J., Harris, R. E."Prospective evaluation of C-reactive protein, smoking and lung cancer death in the Third National Health and Nutrition Examination Survey". International Journal of Oncology 47, no. 4 (2015): 1537-1544. https://doi.org/10.3892/ijo.2015.3141