Joint Effects of Dietary Trace Metals and DNA Repair Capacity in Lung Cancer Risk

The role of dietary trace metals in lung cancer development remains unknown, but are of interest because trace metals affect DNA stability (1-3). We have previously shown that suboptimal DNA repair capacity (DRC) is associated with increased lung cancer risk in ever and never smokers (4). Because dietary trace metals have the potential to modulate DRC, the diet-DRC pathway may influence lung cancer risk.

Dietary trace metals are important for maintenance of DNA integrity. Zinc (Zn; ref. 1) and copper (Cu; ref. 2) are cofactors for enzymes, including Cu-Zn superoxide dismutase (CuZnSOD) and DNA repair proteins. Zn catalyzes the production of cysteine-rich metallothionein, a potent scavenger of hydroxyl radicals (1) and is involved in regulation of DNA transcription through Zn finger proteins (5). Zn deficiency may render p53 to adopt a “mutant-like” conformation that alters the ability of the cell to respond to DNA damage (1). Whereas Cu catalyzes the formation of reactive oxygen species, oxidative damage has been linked to chronic Cu overload (2). Selenium (Se) is a key component of antioxidant enzymes, including glutathione peroxidase, which removes hydrogen peroxide generated by free radicals (3). Iron (Fe) is required for cellular function, but Fe overload may generate reactive oxygen species and damage DNA (6). Calcium (Ca) is also an important mediator of reactive oxygen species because oxidative stress causes cytoplasmic influx of Ca²⁺ from the extracellular environment (7). Increasing levels of Ca²⁺ in the cytoplasm lead to influx into the mitochondria and nuclei, which could affect DNA stability and signal transduction (7).

Because dietary trace metals could affect DRC, knowledge about their independent and joint associations may lead to a better understanding of lung cancer etiology and prevention. In a recent publication, we showed that dietary intakes of Cu and Zn were inversely associated with lung cancer risk, but not Se (8). We now extend this research to evaluate the joint associations between dietary trace metals and functional data on DRC in lung cancer risk.

Our study included 1,139 newly diagnosed, histologically confirmed lung cancer cases and 1,210 healthy controls with DRC data, a subset of an ongoing lung cancer study (8). Cases were recruited before treatment at M. D. Anderson Cancer Center (Houston, TX). There was no age, gender, ethnic, or stage restrictions. Controls were recruited from the largest multispecialty physician group in Houston, and frequency matched to cases by age (±5 years), sex, ethnicity, and smoking status (current, former, and never; ref. 9). The overall response rate among cases and controls was ∼75%. This research was approved by the M. D. Anderson Cancer Center Institutional Review Board.

Our previous article described in detail the collection of epidemiologic data and dietary trace metal assessment (8). In brief, participants were interviewed in person about demographic and lifestyle characteristics. Smokers of at least 100 cigarettes in their lifetime were classified as ever-smokers; among whom, former smokers had quit smoking at least 1 year before diagnosis (cases) or before interview (controls). For cases, body mass index was based on self-reported height and weight 1 year before diagnosis.

Dietary data were collected from a modified (135-item) National Cancer Institute food frequency questionnaire, shown to be valid and reliable across populations (10). Participants were asked about their diet during the year before diagnosis (cases) or study enrollment (controls).

The host-cell reactivation assay was used to measure nucleotide excision repair capacity, as previously described (4). Cultured lymphocytes were transfected with nonreplicating plasmids harboring a chloramphenicol acetyltransferase (CAT) gene, a bacterial drug-resistant gene. Plasmids were treated with benzo(a)pyrene diol epoxide, an ultimate tobacco carcinogen, before transfection, to introduce benzo(a)pyrene diol epoxide–DNA adducts as the substrate of cellular repair. Only stimulated lymphocytes will uptake the plasmids and exhibit nucleotide excision repair activity that removes the adducts in the plasmids. The DRC value (%) is the CAT expression of cells transfected with plasmids divided by that of cells transfected with untreated plasmids ×100%.

Quartiles of dietary Zn, Cu, Se, Fe, and Ca were created from the distribution of intake in controls. Multiple logistic regression analysis was done to calculate odds ratios (OR) and 95% confidence intervals (95% CI) for the dietary trace metals, and their joint associations with DRC in lung cancer risk. The models were adjusted for age, gender, race, smoking status, pack-years smoked, family history of cancer, body mass index, education, income, and total calories (model 1). Among controls, correlations between intakes of Zn, Cu, Se, Fe, and Ca and total calories were significant (P < 0.001). Although caloric intake did not differ in cases and controls, we included it in the model because of its correlation with the trace metals. Also, intakes of Cu, Zn, Se, Fe, and Ca were analyzed individually (model 1), and in a model adjusting for the other trace metals (model 2). In joint effects analyses (trace metals + DRC), subjects with high dietary Zn, Cu, Se, Fe, and Ca (greater than median split for energy-residual adjusted Zn, Cu, Se, Fe, and Ca intake in controls) and proficient DRC (greater than median split for DRC in controls) were the reference group because this group would be expected to have the lowest risk. Potential interactions between dietary trace metals and other lung cancer risk factors were tested on a multiplicative scale in the main-effects models. Statistical analyses were done with SAS (version 8.0; SAS Institute, Inc.). All tests were two-sided, and a P value of <0.05 was considered significant.

Among 1,139 cases and 1,210 controls with DRC data, the mean ages were 61.1 and 60.2 years (P = 0.03), within the 5-year matching criterion. Cases compared with controls had fewer never and former smokers, but more current smokers. Total caloric intake and vitamin/mineral supplement use did not differ between cases and controls. Cases had significantly lower dietary Zn (11.26 versus 11.80 mg/d), Cu (1.42 versus 1.55 mg/d), and Fe (15.15 versus 15.91 mg/d; but not Se and Ca) and lower DRC (8.26 versus 8.93%) than controls.

Like our previous overall analysis (n = 1,676 cases; n = 1,676 controls), the current study revealed significant protection against lung cancer ranging from 39% to 60% for dietary Cu, 16% to 32% for Zn, 19% to 34% for Fe, but not for Se or Ca. The direction and magnitude of these associations were similar when dietary Cu, Zn, or Fe intakes were reciprocally adjusted for the other trace metals. Further, more efficient DRC was associated with protection, with OR of 0.69 (0.54-0.87) and 0.53 (0.41-0.68), respectively, for the third and fourth quartiles compared with the lowest quartile of DRC. Also, mean DRC levels in the controls did not differ by quartile of intake of Zn, Cu, Se, Fe, and Ca (data not shown).

Compared with subjects with high dietary Cu + proficient DRC, there was a monotonic 33%, 66%, 154% increased risk of lung cancer for those with high Cu + suboptimal DRC, low Cu + proficient DRC, and low Cu + suboptimal DRC (P_trend < 0.0001), respectively (Table 1). These associations were similar in men and women, but older (>60 years), lean (body mass index <25) subjects, those with late-stage disease, alcohol users, and with a family history of cancer were at higher risk (Table 1). In the low Cu + suboptimal DRC group, the risk estimates were similar for current and never smokers (Table 1).

Compared with subjects with high dietary Zn + proficient DRC, the low Zn + suboptimal DRC group had the highest risk (Table 2). The risk was more pronounced in men than women. Among current smokers, we found a monotonic 2.2-, 2.26-, and 2.76-fold increased risk for lung cancer for subjects with high dietary Zn + suboptimal DRC, low dietary Zn + proficient DRC, and low dietary Zn + suboptimal DRC (P_trend <0.0001). The associations were more pronounced in heavy smokers, and those with a family history of cancer (Table 2), but were similar by body mass index (<25 or >25), number of cigarettes smoked per day (<20 or >20), and supplement use and alcohol intake (data not shown).

Compared with subjects with high dietary iron + proficient DRC, those with low iron + suboptimal DRC were at the highest risk for lung cancer, with more pronounced effects in men; older (>60 years); or lean (body mass index <25) subjects; and those with longer duration of smoking (>31 years), late-stage disease, and a family history of cancer (Table 3). No significant associations appeared in the joint analysis for dietary Se or Ca and DRC (data not shown).

In this large case-control study, we confirm (8) an inverse association between dietary Cu and Zn (but not Se) and lung cancer risk. We also found that increased total Fe intake was associated with reduced lung cancer risk (P_trend = 0.0002). We now show intriguing joint effects from decreased DRC and low intakes of selected trace metals.

Cu, Zn, Se, Fe, and Ca are involved in reactive oxygen species metabolism (1-3, 6, 11) and are cofactors for enzymes needed for DNA stability and signal transduction pathways (1-3). Because these functions of Cu, Zn, Se, Fe, and Ca are implicated in the initiation and progression of lung cancer, we postulated that the joint associations between dietary minerals and DRC may improve our understanding of lung cancer etiology and prevention. To our knowledge, there are no published reports of the joint associations between dietary trace metal intakes and DRC with lung cancer risk.

In three studies of dietary Zn intake and lung cancer risk, two reported an inverse association (12, 13) and one reported a positive association (14). Two studies of plasma/serum Zn and lung cancer revealed inconsistent results (15, 16). There are no published studies of Cu intake and lung cancer risk except our data (8). Published studies of either serum or hair Cu levels and lung cancer had small samples and inconsistent results (17, 18). It is possible that in an environment with heightened oxidative stress, and low Zn and Cu intakes, DNA damage and inadequate DRC may contribute to cancer. In the low Cu + DRC or low Zn + DRC groups, current smokers had the highest risk. Cigarette smoking is an established risk factor for lung cancer and is an important source of reactive oxygen species in the lungs; thus, individuals with low Cu and Zn intakes and suboptimal DRC may be unable to combat carcinogenic effects from cigarettes.

We found that Fe intake was inversely associated with lung cancer risk and those with low Fe + suboptimal DRC were at the highest risk. In a previous study, total Fe intake increased lung cancer risk, whereas heme iron decreased risk (12). We could not distinguish between heme and nonheme iron intakes.

The null findings for dietary Se-DRC and lung cancer risk were unexpected. There is inconsistent evidence that Se intake may reduce lung cancer risk (14, 19, 20). Our null finding is probably not due to an inadequate range of Se intake, because Se intake in our control population is comparable with national estimates (21). In an earlier study, Ca intake was positively associated with risk (12). We found a nonsignificant 17% increased risk in the highest quartile of Ca intake, but no joint associations for Ca and DRC.

The data from this study, as with all case-control studies, may be subject to recall bias. However, cases were recruited at diagnosis and asked about their diet during the year before diagnosis, whereas controls reported their diet the year before the study. Although the food frequency questionnaire is practical for large epidemiology studies, it is well known that use of the food frequency questionnaire is associated with measurement errors (22). To improve the accuracy of dietary reporting, interviewers were trained in food frequency questionnaire administration; and incomplete responses were requeried by staff nutritionists. Another source of error could arise from nutrient values for dietary trace metals in the food composition database. The DIETSYS+Plus database has a wide cross-section of foods, including ethnic foods, but analytic values for trace metal composition of foods in the Houston area are not available for comparison purposes. Dietary trace metals may vary by region, and soil content, contributing to variability of the nutrient composition in foods. Despite these limitations, the food frequency questionnaire has been shown to reliably assess dietary intake to classify individuals by quartile of intake (23).

Our controls consumed comparable daily amounts of dietary Zn, Cu, Se, Fe, and Ca to those reported by the 1999 to 2000 National Health and Nutrition Examination Survey (NHANES) for the U.S. population (21). On average, our controls consumed 10.4 mg/d dietary Zn compared with 11.4 mg/d in NHANES (1999-2000); 1.3 mg/d dietary Cu compared with 1.2 mg/d in NHANES (1999-2000); and 91.7 μg/d dietary Se compared with 103.1 μg/d in NHANES data; Fe intake in our controls was 15.9 versus 15.2 mg/d; and Ca intake was 872 mg/d compared with 863 mg/d in NHANES. We recognize that our study would be strengthened by more objective serum measurements of trace metals.

Dietary trace metals in lung cancer risk remain an understudied area of research. These intriguing joint associations between dietary trace metals and DRC need to be confirmed in prospective research and biological mechanisms underlying the associations should be elucidated in more detail.