Dietary and Lifestyle Factors and Risk of Non-Hodgkin’s Lymphoma in Oman

Background: The incidence of various types of cancers including the non-Hodgkin’s lymphoma (NHL) has increased during the recent years. Diet and lifestyle factors have been reported to play an important role in the etiology of NHL. However, no such data are available from the Middle Eastern countries, including Oman. Materials and Methods: Forty-three histologically confirmed cases of non-Hodgkin’s lymphoma (NHL) diagnosed at the Sultan Qaboos University Hospital (SQUH) and the Royal Hospital (RH), Muscat, Oman and forty-three age and gender matched controls were the subjects of this study. Frequency matching was used to select the control population. Information on social and demographic data as well as the dietary intake was collected by personal interviews, using a 117-items semi-quantitative food frequency questionnaire. Results: A non-significant increased risk of NHL was observed with higher body mass index (BMI) (OR=1.20, 95%CI: 0.45, 2.93), whereas a significantly decreased risk of NHL was associated with a higher educational level (OR=0.12, 95%CI: 0.03, 0.53). A significantly increased risk was observed for higher intake of energy (OR=2.67, 95%CI: 0.94, 7.57), protein (OR=1.49, 95%CI: 0.54, 4.10) and carbohydrates (OR=5.32, 95%CI: 1.78, 15.86). Higher consumption of daily servings from cereals (OR=3.25, 95%CI: 0.87, 12.09) and meat groups (OR=1.55, 95%CI: 0.58, 4.15) were also found to be associated with risk of NHL, whereas a significantly reduced risk was associated with higher consumption of vegetables (OR=0.24, 95%CI: 0.07, 0.82). The consumption of fruits, milk and dairy products however showed no significant association with the risk of developing NHL. Conclusion: The results suggest that obesity, high caloric intake, higher consumption of carbohydrate and protein are associated with increased risk of NHL, whereas a significantly reduced risk was observed with higher intake of vegetables.


Introduction
The incidence of various types of cancers including the non-Hodgkin's lymphoma (NHL) has increased during the recent years.In some countries the incidence of NHL has risen faster than the incidence of all other types of cancers (Chiu and Weisenburger, 2003;Zheng et al., 2004;Kabat et al., 2012).NHL is the 3 rd most common cancer in the Gulf Cooperation Council (GCC) states and its rate is higher in men than in women (Khoja and Zahrani, 2010;Salim et al., 2011;Tadmouri and Nair, 2012).Although the awareness on the biology and treatment of NHL has increased, however still little is known about its etiology.The increase in the incidence has been attributed to immunodeficiency, various infections, familial aggregation, blood transfusion, genetic susceptibility, chemical exposure to pesticides and solvents, hair-dyeing agents, nitrates in drinking water, as well as the dietary and
Food is the largest single antigenic challenge to the human immune system and operates on a regular and perpetual basis.Although it is not possible to exactly quantify the overall risks, it has been estimated that about 35 percent of all cancer deaths may be related to dietary factors (Doll and Peto, 1981;Anand et al., 2008).Culturally specific dietary patterns and a healthy lifestyle are likely to play a major role in determining cancer risk.There is compelling evidence that dietary patterns, food, nutrients, and other dietary constituents may be closely associated with the risk for several types of cancers (Davis, 1992;Donaldson, 2004;Ananad et al., 2008).Some studies observed positive associations between the intake of protein, fried red meat, dairy products, total fat, saturated fat and trans-saturated fat and NHL risk whereas the others reported no significant association (Donaldson, 2004;Daniel et al., 2012).The data on the association between meat and other nutrients intake and risk of NHL however, remains inconsistent.Diet high in protein and fat is considered to increase the risk of NHL as it may alter the immune-competence, which is an important and established risk factor for NHL (Zhang et al., 1999;Zheng et al., 2004;Milner, 2006;Lee et al., 2008;Kabat et al., 2012).On the other hand, a reduced risk has been reported with higher intakes of dietary fiber, whole grain foods and several fruits and vegetables.This has been attributed primarily to the presence of biologically active phytochemicals and antioxidant nutrients, which have the ability to quench free radicals and inhibit nitrosation reactions (Frankenfeld.et al., 2008;Thompson et al., 2010;Holtan et al., 2011;Ollberding et al., 2012).
Sultanate of Oman lies in the Southeastern part of Arabian Peninsula and represents a fine mix of ancient society and modern lifestyle.A steady increase in the per capita income, during the past years has improved the socio-economic status of the people resulting in drastic changes in lifestyle and food consumption patterns replacing the traditional foods with more Westernstyle ready-made foods (Musaiger et al., 2011;2012).According to the Ministry of Health of Oman, more than 900 new cancer cases are diagnosed in the Sultanate and NHL is the 3rd most common form of cancers (Ministry of Health, 2012).No information is currently available on the effect of dietary and lifestyle factors on the prevalence of NHL in Oman.The present case-control study was therefore conducted to identify various dietary and lifestyle factors associated with the risk of developing NHL in Omani population.

Materials and Methods
In a case-control study approach, we evaluated the multiple potential causes of NHL.The study was conducted at the out-patient departments of Oncology Clinic, Sultan Qaboos University Hospital (SQUH) and Royal Hospital (RH), Muscat, Oman.Histologically confirmed cases of NHL classified according to the International Classification of Diseases, (WHO, 1977), diagnosed at SQUH and RH, Muscat were recruited for this study.Population based controls were recruited by random access method from the Muscat area and frequency matching was done to achieve a similar age and sex distribution to that of all cancer cases.Cases and control were frequency matched by age in 5-years intervals by adjusting the number of controls randomly selected in each age stratum.Forty-three cases and 43 age and gender matched controls were included in this study.The cases and controls were asked to complete a comprehensive questionnaire that included both demographic information and 117-items food frequency questionaire, in personal interviews.The study questionnaire was tested for validitiy, reliability and reproducibility before conducting the study.All study subjects were recruited on voluntary basis and a written informed consent was obtained.The study was approved by the Medical Research Ethics Committee of Sultan Qaboos University, Muscat, Oman.

Demographic and anthropometric data
The variables related to demographic, anthropometric and social characteristics such as gender, age, marital status, occupation, educational level, smoking, tobacco chewing, main sources of drinking water and intake of mineral and vitamin supplements, were included in the questionnaire.The anthropometric measurements included, weight, height and body mass index.

Semi-quantitative food frequency questionnaire (FFQ)
A comprehensive 117-items semi-quantitative food frequency questionnaire (FFQ) was designed to collect dietary information about the amount and frequency of consumption for various food items commonly consumed in Oman (Table 1).The FFQ was developed based on the already developed and validated food frequency questionnaires according to Block et al. (1990) and Willet (1998) keeping in view the Omani food consumption patterns.For each food item, a portion size was specified using the common household serving units/utensils (table spoon, tea spoon, cup 180mL, cup 240mL), which were shown to the volunteers to comprehend their intake of foods.Each subject was personally interviewed to complete the questionnaire characterizing his/her usual diet intake stating how often, on the average, they consumed the specific amount of each food during a period of one year prior to being interviewed with the understanding that this is highly correlated with more distant past.There were nine possible responses, including never, a few times per year, 1-2 times/month, 1-2 times/week, 3-4 times/week, 5-6 times/week, 1-3 times /day, 4-5 times/day, and more than or equal to (≥6) times/day.The data was collected on the consumption frequency and usual portion sizes for 117 food items in one year (before the diagnosis of NHL).The different food groups included in the questionnaire were: breads, vegetables, fruits, meat and meat products, milk and milk products, desserts, beverages, sandwiches, and traditional Omani dishes.These food items were included based on their contribution to the total intake of energy by the population groups and account for over 90% of total Omani population's intake.Each question had several choices and the interviewing researcher marked the best appropriate choice according to the reply of the subject.

Estimation of nutrient intake
The information from the food frequency questionnaire was summarized into several nutrients and food groups.Estimates of total daily intake of energy, protein, carbohydrate and fat were calculated from the nutrient estimates assigned to each dietary item using a computer based Nutrient Analysis Program (NutriBase 6.0).The chemical composition of traditional Omani dishes was determined from either our own laboratory data on the composition of food or from the already published DOI:http://dx.doi.org/10.7314/APJCP.2013.14.2.841 Dietary and Lifestyle Factors and Risk of Non-Hodgkin's Lymphoma in Oman literature (Musaiger et al., 1998;Ministry of Health, 2006).In order to calculate the intake of nutrients (protein, carbohydrates, and fat) in grams, all the frequencies were converted into daily basis.The nutrients were calculated by multiplying the portion sizes (in grams) by the nutrient content per 100 grams of that food by the frequency intake per day.Finally the total daily intake of each nutrient was calculated by summing the nutrient values across all foods which contained that specific nutrient.All the respondents were subsequently grouped according to their intake level of each nutrient based on the frequency of consumption among both cases and controls.
In addition to this, the number of daily servings of food groups according to the food guide pyramid (Cereals, Vegetables, Fruits, Milk and Meat) was estimated.Serving sizes were defined either in usual natural units (e.g.teaspoon, tablespoon or cup) or as an average serving in grams.This was achieved by estimating the number of servings provided by each food item in FFQ.Some questionnaire items were allocated to two or more food groups; according to their nutrient contents (e.g.pizza was apportioned into both bread and cheese for one half serving each) and then the daily servings for each food group were calculated by multiplying the portion size by the number of serving of that food by the frequency intake per day.Finally the total daily intake of each nutrient was calculated by summing up the serving values across all food items which contributed any servings to that food group.Subsequently, all the respondents were grouped according to their intake level for each food group based on frequency of consumption among cases and controls.There is often an interest in how much of one's caloric intake comes from a particular nutrient.We therefore computed the percentage contribution of total calories the respondents consumed from that particular nutrient (protein, carbohydrate or fat).Both the cases and controls were categorized according to their nutrient densities from each macronutrient.

Statistical analysis of the data
A number of variables were created from the original data sets, which were defined and categorized into 3 quartiles in order to facilitate the interpretation of results and to assess the disease risk.The consumption of nutrients was examined in terms of absolute quantitative terms (in grams) and the percentage of total calories consumed by respondents was calculated based on macronutrients intake.The total daily intake of number of servings per day for each food group (cereals, vegetables, fruits, milk and milk products and meat and meat substitutes) were estimated according to food guide pyramid and were also categorized into 3 quartiles of consumption.The lowest  quartile of intake served as the reference group for all comparisons.
The statistical analysis of data was done using the Statistical Analysis Software Package (SPSS) version 16.The statistical model suggested by Willet and Stamfer (1986) was used to assess the effect of calories and nutrient intake.The adjustments for confounding variables were also considered in the final model.Multivariate analysis was used to assess the independent contribution of each nutrient of interest to disease status.The association between the food groups and risk of NHL was estimated using the odds ratios (OR) and 95% confidence intervals (95%CI) calculated from the maximum likelihood estimates using binary logistic regressions.The OR and 95%CI relating to non-dietary components and NHL were calculated using binary logistic regressions.

Demographic and social characteristics
The characteristics of patients with NHL and controls are shown in Table 2.All the reported associations are relative to the lowest quartile.There was no significant difference in the age between the cases and controls.The educational level of controls was significantly better than that of cases.Although no consistent increase or decrease was observed, there was a significant decreasing trend in the risk of NHL with increasing level of education (OR=1.00,0.21, 0.27, 0.12).Our results are in line with the findings of Hermann et al. (2010), who also could not confirm an overall consistent association between educational level and risk of Hodgkin's lymphoma (HL) and NHL.Although the risk of the NHL was equal among single and married respondents in both the groups (OR=1.00), a non-significant increased risk of NHL was observed with higher BMI level (OR=1.00,1.84, 1.20).Significant differences were also observed in the smoking status (P=0.02).There were more smokers (21%) among cases (NHL patients) as compared to controls (5%).No significant difference was observed in tobacco chewing (P=0.31).The source of drinking water as well as the intake of vitamins and minerals supplement also did not show any significant differences (Table 3).Overall the results revealed a significantly increased risk of NHL with increasing BMI.However, the risk of NHL decreased significantly with increasing level of education.No significant differences were observed for gender, marital status, tobacco chewing, vitamins and minerals supplement intake and source of drinking water.
Conflicting results have been reported in the literature on the relationship of obesity to NHL risk.The increased risk of NHL associated with overweight and obesity observed in our study is consistent with the reports, which indicated that obese men and women (BMI ≥30 Kg/m 2 ) had a significantly increased risk of NHL (Pan et al., 2005;Skibola, 2007;Reynolds et al., 2011;Kabat et al., 2012).People with lower BMI were found to have consistently lower risk of NHL (Kelemen et al., 2006).The mechanism for the link between NHL risk and obesity is not clear.One proposed hypothesis is that early life influences on   growth and immune function and obesity may impair the immune function and therefore may increase the risk of NHL later in life (Lee et al., 2008;Kabat et al., 2012).Another possible mechanism is that obesity can cause changes in the metabolism of endogenous hormones, including sex steroids, insulin and insulin-like growth factors, which could distort the normal balance between cell proliferation, differentiation, and apoptosis (Bianchini et al., 2002).Leptin also plays an important role in the regulation of food intake, energy expenditure, and the control of body weight (Pan et al., 2005).Based on the data from case-control and prospective cohort studies there is increasing evidence that obesity may increase the risk of NHL.Our results are consistent with these findings.
Total energy intake and risk of NHL Significant differences were observed in the total energy intake between the cases and controls (Table 4).Although there was no specific trend between an increase in energy intake and risk of NHL, the higher intake of energy was found to be associated with an increased risk of NHL.The quartiles of caloric intake suggested an increased risk of NHL in the second quartile followed by a decrease in OR in the 3 rd quartile as shown in Table 4.The percentage of energy (kcal/day) contribution from various macronutrients is shown in Figure 1.No significant differences were observed in the percentages of daily energy intake from protein and fat between the cases and the controls (p=0.69,0.86, 0.07 for protein, fat and carbohydrate respectively).A positive effect of total caloric intake on disease has been suggested previously (Purdue et al., 2004;Pan et al., 2005).A populationbased case-control study from eight Canadian provinces found that high total energy intake showed a relatively strong association with risk of NHL (Purdue et al., 2004).However, some other studies did not observe such association (Chang et al., 2006;Talamini et al., 2006).Obesity, energy homeostasis and physical inactivity have been reported to influence the immune function and therefore may influence the risk of NHL (Pischon et al., 2008;Kobat et al., 2012).The results of our study showed a relatively positive association between the total energy intake and risk of NHL and are in line with these findings.

Consumption of protein, fat, carbohydrate and risk of NHL
The mean daily consumption of protein, fat and carbohydrate for cases and controls is presented in Tables 5, 6 and 7 respectively.No significant differences (P=0.21) were observed in the protein intake.Although the risk estimate for the highest quartile of protein intake (≥90 g/ day) was 1.49, only significantly elevated OR of 2.74 was observed for the second quartile of protein consumption (65-90 g/day).However, when the total protein intake was examined in quartiles as the percentage energy contribution from protein, no association was observed with risk of NHL (1.00, 0.67, and 1.11).The increased NHL risk associated with higher protein intake observed in this study is consistent with the data reported in some previous studies, which suggested that high intake of protein can influence the metabolic pathways of immune system and may increase the risk of NHL (Talamini et al., 2006, Mozaheb et al., 2012).Although the evidence from animal studies supports an etiologic role of high protein consumption, the results from epidemiological studies are inconsistent.A population based case-control study conducted by Ward et al. (1994) found that animal protein was not significantly associated with the risk of NHL, whereas Chiu et al. (1996) found an elevated risk associated with higher consumption of animal protein.The increased risk associated with higher protein intake appeared to be mainly related to animal protein sources and not with vegetable protein consumption (Zheng et al., 2004).Other studies however found no association between animal or vegetable protein intake and NHL risk (Ward et al., 1994, Zhang et al., 1999).Diets high in animal protein may lead to prolonged (chronic) antigenic stimulation which could induce a state of immune unresponsiveness or oral tolerance and possibly may elevate the lymphoma risk.On the other hand, a population-based case-control study conducted in eight Canadian provinces found no association between the protein intake and risk of NHL (Purdue et al., 2004).A similar study from Sweden also reported similar results  Percentages of energy (kcal/day) from macronutrients (Chang et al., 2006).
The mean daily fat intake between the cases (80.4 g/day) and controls (62.2 g/day) differed significantly (p=0.02).However, the mean daily fat intake did not show a consistent association with NHL risk.An increase in NHL risk (OR=3.00)was observed in the second quartile for daily fat intake, followed by a decrease in risk in the third quartile (OR=0.63),which was much lower than that in first quartile (Table 6).Similarly, the percentage energy contribution from fat in different quartiles did not show a consistent increase or decrease in risk estimates (OR=1.00,1.36, 1.09) (Table 7).No apparent association could therefore be established between the amount of fat intake and percentage of calories contributed from fat with NHL risk.Our results however do not support the data that higher consumption of total fat, saturated fat and trans-saturated fat is associated with higher NHL risk (Purdue et al., 2004).The discrepancy in the results may be because of a smaller sample size.
Significant differences were observed in carbohydrate intake between the cases (354.7 g/day) and controls (286.1 g/day).Carbohydrate intakes in the second and third quartiles were both associated with significantly elevated OR of 5.50 and 5.32 respectively.Overall the total carbohydrate consumption in terms of absolute intake showed an elevated NHL risk (Table 7).The ORs for the percentage energy contribution from carbohydrate in three different categories of consumptions were 1.00, 0.76 and 2.33 respectively.The data from various epidemiologic studies presents an inconsistent picture of the association between carbohydrate intake and overall risk of NHL, ranging from significantly increased risk (Zheng et al., 2004) to no association (Chang et al., 2006).Our results however suggest that high intake of carbohydrates, in particular from refined grains; can increase the risk of NHL.
Intake of number of servings from various food groups and risk of NHL Significant differences were observed in the daily intake of number of servings from the cereals and vegetables groups between the cases and controls.No significant differences were however observed in the daily intake of number of servings from fruits, milk and milk products and meat and meat substitute groups (Table 8) The OR describing the association between the individual food groups and risk of NHL are presented in Table 8.There was a dramatic increase in NHL risk associated with second quartile of cereal consumption, which decreased in the third quartile (OR=1.00,9.10, 3.25).Overall, higher intake of number of daily serving from cereals, in particular from refined grains, was found to be positively associated with risk of NHL.Likewise, no consistent increase or decrease was found to be associated in all three quartiles of consumption of meat and vegetables groups.However, higher total meat consumption was generally found to be associated with a significantly higher risk of NHL (OR=1.00,2.38, 1.55), whereas a decreased risk of NHL was associated with higher daily intake of number of servings from vegetables group (OR=1.00,0.09, 0.24).No apparent association was observed between the quartiles for the daily intake of number of servings from fruits and milk and milk products with NHL risk (OR=1.00,0.95, 1.88 for fruit and OR=1.00, 0.51, 0.81 for milk and milk products).Our results support the hypothesis that higher consumption of cereals, in particular the refined grains, is associated with increased risk of NHL (Talamini et al., 2006).Our results are also consistent with the data from other case-control studies, which showed that the consumption of breads and cereal products was positively associated with NHL risk (Ward et al., 1994;Zheng et al., 2004).The results from some other studies however did not report any such association (Purdue et al., 2004;Chang et al., 2006).
Our results indicated that higher total meat consumption was found to be associated with significantly higher risk of NHL.These results are in line with the previously reported data (Chiu et al., 1996;Zheng et al., 2004;Chang et al., 2006).Association with NHL risk was however inconsistently observed across different types of meat (De-Stefani et al., 1998).The consumption of cured meat, salami, and sausages was associated with a modest increase in NHL risk (Davis, 1992).A positive association has also been reported between the high intake of beef, pork, lamb meat and liver with increased NHL risk (Francheschi et al., 1989;Zhang et al., 1999).The results of the present study are consistent with these findings.Some other studies however, found no associations with red meat intake and NHL risk (Talamini et al., 2006;Skibola, 2007;Daniel et al., 2012).Such inconsistencies in the results may be attributed to differences in study design, geographic location, or smaller sample size.The results from the studies on fish consumption and reduced risk of NHL are also inconsistent and have reported only a weak but non-significant inverse association between the fish consumption (≥4 servings of fish per week) and risk of NHL (Matsuo et al., 2001;Purdue et al., 2004;Skibola, 2007).Other studies however did not support an association between fish consumption and reduced risk of NHL (Ward et al., 1994).Overall, the evidence is inconclusive but suggests that an inverse association exits between the fish consumption and NHL risk.
The results of the present study showed an inverse association between the vegetable consumption and risk of NHL.Our results are consistent with the reported data suggesting that a higher intake of vegetables, in particular the dark green leafy vegetables, may reduce the risk of NHL (Ward et al., 1994;Talamini et al., 2006;Chang et al., 2006;Kelemen et al., 2006;Skibola, 2007;Chiu et al., 2011;Chen et al., 2012).Vegetables are considered as a rich source of carotenoids, vitamins C, E, folate and dietary fiber, as well as numerous other phytochemicals that may inhibit the carcinogenesis by neutralizing the reactive oxygen species generated endogenously or supplied exogenously.Thus, they may reduce the oxidative DNA damage and mutations, and enhance immune responses (Zhang et al., 2000;Hemsdorff et al., 2012;Lee et al., 2012).The increase in the magnitude of cancer over the last 33 years in Kuwait has been attributed to changes in lifestyle such as dietary habits, physical activity and obesity (Elbasmi et al., 2010).It has been suggested that the primary focus in the prevention of cancer should be given to reduce the obesity (Willet, 2010) and increased consumption of fruits and vegetable will not only help in reducing the rate of obesity but will also help in reducing the risk of many cancers (Aune et al., 2012;Reiss et al., 2012).Our results are consistent with these findings and showed that higher daily intake of vegetables was associated with a decreased risk of NHL.

Study limitations
There are however some limitations of the study, for example the methodological issues such as the use of a semi-quantitative FFQ, which is not a perfect tool for measuring the dietary intakes.However, it is the most practical and well-validated method available for such type of studies.Secondly, the study did not examine the heterogeneity across the NHL subtypes with respect to dietary associations.Thirdly, the small sample size (n=86, 43 cases and 43 controls) limited the statistical power of the analysis to detect some weak associations.The findings for individual dietary factors should therefore be interpreted with caution.

Discussion
In conclusions, Overall, this study revealed a number of risk factors among which obesity, high caloric and protein intakes were found to be associated with elevated risk of NHL Higher consumption of cereals, in particular the refined grains and meat were also found to be associated with increased risk of NHL, whereas a significantly reduced risk was observed with higher consumption of vegetables.This is the first study to report about an association between diet and NHL risk in Omani population.The results add valuable information on the subject and also lead to a better understanding of dietary consumption patterns in Omani population and their implications in overall health and disease.Considering how little is known about the etiology of NHL in Oman, further well-designed randomized control trials with larger cohorts should be conducted to validate these preliminary results and to quantify the effects of various nutritional and metabolic biomarkers in the context of entire diet and its possible association to NHL and other types of cancers.