Evaluation of the impact of chemotherapy on eating habits of women with non-metastatic breast cancer

Breast cancer is the second most frequent cancer worldwide and the most common in women. Risk factors include hormonal status, race, genetics, as well as life style. Thus a decrease in incidence of this disease should be possible by life style alterations such as an increase in physical activity, maintaining body weight within the limits specified by body mass index (BMI) calculations, a moderate consumption of alcohol, and an adequate diet.1 2 Overweight or obesity at diagnosis can be a factor indicating a worse prognosis, with recurrence and comorbidities.3

Nutritional education can be an important factor in maintaining an adequate energy balance. A restriction of 10-40% of the total caloric intake has been shown to decrease cellular proliferation and increase apoptosis; however, this is not considered a preventative strategy, on its own. A decrease in alcohol intake and smoking, a good diet, and regular physical activity also seem to be important for the prevention of cancer, as these factors help control hormonal levels, and reduce the production of free radicals.4 5

Nutritional requirements are different before, during and after cancer treatment. Often, patient receives conflicting information, which linked with taboos, aversions, preferences and beliefs, lead to an increase or decrease in food intake, and may result in dietary disequilibrium.3 Nutritional alterations such as a direct effect of medication on metabolism or side effects involving the gastrointestinal tract take place during chemotherapy. Appetite may also be affected by the stress induce by a cancer diagnosis and chemotherapy.5 Dietetic supervision to avoid weight loss is recommended during chemotherapy, as well as to avoid interruptions in food intake with radiotherapy.6 It is also a useful method to prevent and control associated side effects.7

After treatment, disease free patients constitute a risk group to develop diseases such as diabetes, osteoporosis, and cardiovascular problems, as well as second primary tumors. As one of the measures to prevent these disorders, this group should be encouraged to adopt a more healthful lifestyle.8

There are a number of studies, which investigated the relationship among nutrients, nutrition, and prevention of breast cancer. However, very few studies have focused on alteration or maintenance of dietary habits combined with sociodemographic variables, BMI, gastrointestinal (GI) symptoms, and altered food groups before and after treatment. The objective of this study was to analyze changes in food intake during chemotherapy in patients with breast cancer. To do this, patients responded to a food frequency questionnaire (FFQ) about their intake from the various food groups (Appendix).

This was a prospective study with breast cancer patients, 18 years of age or older, who were seen in the Clinical Oncology Ambulatory Care, from November 2009 until October 2010. Patients included in the study were those with non-metastatic disease, in adjuvant or neoadjuvant treatments, with curative intent with the following chemotherapy regime: anthracyclines (4-8 cycles of adriamycin and cyclophosphamide (AC) with a 21 day break between cycles. Exclusion criteria were the following: patients on parental or nasogastric tube nutrition; presence of ascites; mechanical obstruction of the GI tract; previous surgery involving the esophagus, stomach or intestines, concomitant radiotherapy. There was no nutritional advice given at this stage; the study was then undertaken according to a protocol developed by nutrition department of the hospital.

Before their first chemotherapy session, patients answered the questions about personal details and life style, as well as the FFQ designed specifically for the Brazilian population.9 The questions on this form took between 20 to 40 minutes to answer. Following this, height and weight were measured for calculation of BMI.10 A nutritionist documented the diagnostic information (tumor, stage, associated diseases, and chemotherapy treatment protocol) using details from the patient's chart.

Thirty days after the finish of the current chemotherapy, on the return appointment to the clinical oncology department, or for some other procedure, the research nutritionist documented details related to side effects of treatment and weighed the patients. The FFQ was applied again, for later correlation of any changes in BMI with food intake.

A total of 41 patients with nonmetastatic breast cancer answered to the first questionnaire; however, 15 were excluded (4 did not respond to the second questionnaire, 9 responded after the study had closed, and 2 did not finish treatment).

The second questionnaire was given 28-38 days after the latest chemotherapy treatment (one patient replied in 28 days, another in 35, and another in 38 days). The median age of the patients was 53 years (29-69 years). A total of 57.7% of women were married and 50% had completed university level training. Surgery was performed on 70%. Only one patient was treated on the Brazilian National Health Service (Sistema Único de Saúde - SUS), while all the others (96.2%) had private health insurance plans. Disease staging was as follows: 6 patients (23%) were stage I, 12 patients (46%) were stage II and 8 (31%) were stage III. Hypertension was found in 38.5% and cardiovascular disease in 15.4% (Table 1). Side effects occurred in 22 (85%) patients, either singly or together, such as nausea (17 patients), alteration in taste (11 patients) loss of appetite in 7, vomiting (6 patients) mucositis in 3 patients, diarrhea in 2 patients and pain in one patient.

At the first interview before treatment, 69% of women (n=18) were overweight or obese. At the end of the chemotherapy cycles, 73% had a BMI greater than 25 kg/m². This represented a weight increase of 7%, although there was no statistically significant difference in weight before and after treatment. Figures 1 and 2 illustrate the BMI at the two evaluations. Only 13 patients (32%) of those interviewed took regular physical exercise (average of 5 hours per week) before their cancer diagnosis. The other 68% did not do any regular exercise. Of the 26 participants, 9 (35%) took physical exercise. Three patients (11%) were smokers with an average consumption of 14 cigarettes per day. Four patients reported occasional alcohol ingestion of 2.9 times per week.

At the end of treatment, 58% (15 patients) were given nutritional advice. Some of these women received information on other services (21%), although most of them were advised at this institution (78%).

The answers to the FFQ gave us information referring to food intake of the previous 30 days. Each patient chose one of nine categories of intake frequency (monthly to daily) for each of the 98 food items on the list. These replies were converted into daily intake amounts for each item, as seen in Table 2. Table 3 highlights those foods, which had a significant difference in intake before and after treatment.

In terms of food items consumed more than once per month, there was a numeric change in all items. Table 3 presents a summary of the foods with altered intake, and Table 4 compared the results in terms of food groups for the questionnaires done before and after treatment. There was no influence of socio demographic variables on food consumption from any food group (Mann-Whitney test, details not shown).

Table 5 and Table 6 show a significant relationship between the increase of bread and cereals with nausea, and also an increase in consumption of meat, fish and eggs with a loss of appetite (Mann-Whitney test). There was no significant association found for the other food groups. The number of patients was too small to evaluate pain (n=1), diarrhea (n=2 or 8%), and mucositis (n=3 or 11 % of patients).

The FFQ has been used worldwide for epidemiologic studies and should be adapted for specific populations.11 The FFQ chosen for this study had been based on food information sheets filled in over 24 hours by 200 adults older than 25 years, living in the Metropolitan region of São Paulo, during the first six months of 2002.12

Our study sample was homogenous: women with an average age of 53, who were or are married (76.9%) with 11 to 15 years of formal education (88.4%), and private health care coverage (96.2%). There was no association between any sociodemographic variable and a change in the food group consumption. Differently of us, Patterson et al. (2003)13 found that patients between the ages of 35 to 59 years modified their diets more than those people over 60 years of age. Salminen et al. (2004)14 showed that the probability of a positive modification in dietary habits was greater in the group with a higher level of education, younger age and with a longer time from diagnosis.

A significant statistical difference was found in the increased ingestion of whole milk. On the second questionnaire, 15.4% of patients who previously never consumed this product began to use this item 2-3 times per month. Patients (3.8%) who consumed items in this group once per week, increased their consumption to 2-4 times per week and those which consumed these 5-7 times per week began to consume these once per day (Table 3).

Stroganoff (red meat) also was consumed more frequently. During chemotherapy, women are encouraged by their health professionals and care givers to diversify their intake of food to avoid a monotonous diet. This could have been a creative and nutritious option to vary the menu. Table 3 also shows an increase in consumption of beef liver: 38.5% of women began to eat this meat. 30.8% of those who initially ate this 2-3 times per month began to eat it from once a week up to once a day.

Tuna and shrimp also were consumed more frequently. Fish (fresh or tinned) are rich in omega 6 and omega 3 fatty acids to heart. In the first questionnaire, 92.3% of women said they did not eat shrimp, and in the second questionnaire, this number had decreased by 23.1%, with the consumption frequency being reported as 2-3 times per month (19.2%).

The significant increase in these foods and in the all the protein groups could be attributed to the awareness of the importance of adequate quantities of protein an iron in the diet. Meat foods are avoided due to beliefs that certain food groups should be eliminated during treatment, even though these food groups are not associated with mortality.5 Hanf and Gonder (2005) 15 quoted a possible increase in risk of breast cancer in women with a high consumption of well grilled or baked red meat. However, they emphasize that even in a meta-analysis, which included over 25,000 cases of breast cancer, there was no significant correlation between breast cancer and red meat consumption, or with milk or milk products.

With respect to meat, fish and egg intake, we found a significant relation between the increase in consumption of these items and loss of appetite (Table 5).

Consumption of lettuce and escarole was decreased significantly. In the first questionnaire, 23.1% of women ate these greens 2-3 times per day and during treatment this declined to 2-4 times per week (34.6%). This could be explained by the fact that some patients were advised to eat raw foods during chemotherapy.

Consumption of papaya increased by 7.7% (p=0.032). On the first questionnaire the most frequent consumption was reported to be 2-4 times per week (34.¨), and on the second, this had increased to 5-7 times per week as is shown in Table 3.

There was a significant increase in the consumption of cashew and acerola (Barbados cherry), fruit juices (Table 3). Before treatment, 69.2% did not drink cashew fruit juice and its consumption was 2-3 times per month up to 2-4 times per week. After treatment, frequency was increased from 2-3 times per month to up to once per day. 15.4% of patients did not modify their consumption at all or less than once per month for other frequencies. Acerola fruit juice also underwent change; 19.2% of women who never drank this juice began to drink it 2-4 times per week. Both these fruits are high in vitamin C (acerola 1,500mg and cashew 200mg). The significant increase in these fruit juices could be due to the presence of vitamin C during treatment.

In the first questionnaire, 57.7% of patients did not eat nuts such as peanuts and cashews, or ate them less than once per month. After the second questionnaire, the group which never ate nuts had decreased by 42.3%, and 11.5% had begun to eat nuts once per week.

The specific dietary changes found in this study agree with the results of Salminen et al. (2004),14 who studied dietary modifications in 345 women with breast cancer. One third of their group (p=0.033) modified their diets. Higher educational levels, younger age and a longer time with the diagnosis were all significantly associated with a probability of change. The main alterations were a reduction in consumption of animal fat, sugar and red meat, and an increase in intake of fruits and vegetables.

In our study, more women were also found to be eating macaroni and pasta, which are sources of carbohydrate. The biggest change was found in the groups that never consumed these items, in which frequency increased to as much as 5-7 times per week.

Table 4 shows a significant increase in amount of meat, fish and eggs consumed (p=0.020) and in the group for sweets savories and cakes with icing (p=0.019), which suggests an increase in total caloric intake. This is likely due to the fact that the last group contain simple carbohydrates and, principally, saturated fats, which could explain the weight gain observed overall.

The increase in the group consisting of sweets, savories and iced baked goods represents an increase in simple and complex carbohydrate consumption. The increase in consumption of this group could be associated with an increase in weight during treatment.16

Group V, composed of bread, cereals, root vegetables and beans, was associated positively with the occurrence of nausea. Almost all items in this group were eaten more frequently. We think that the presence of nausea contributed to the increase in consumption in this group.

In group II (meat, fish and eggs) 73% of patients increased the frequency of consumption of these items on the second evaluation: in group IV (fruits and natural juices) and V (bread, cereals, root vegetables, and beans), VII (sweets, savories, and iced baked delicacies), 65% of patients increased their consumption frequency. For group VII, 15% of patients maintained their previous levels, while in groups IV and V this percentage was 4% and 8%, respectively. Our sample was small, but in spite of this, these were still significant differences in the two groups already discussed.

Table 6 shows the maintenance of consumption of beer and an increase in the consumption of wine and other alcoholic drinks. 54% of those interviewed increased their alcoholic beverage intake during treatment. An association between alcohol ingestion and the risk of breast cancer was confirmed by Gonzalez and Riboli (2010).17 In that study, in European women, they found an increase in risk of 3% for a recent increase of 10g per day of alcohol.

Our study group of 26 cases was very homogeneous in terms of treatment and pathological details, even with the limitations of the small sample size, diversity of tumor type and age; the results presented here demonstrate that there was a change in the eating habits of patients undergoing chemotherapy for breast cancer, as evidenced by the FFQ. We show that more attention should be paid to overweight and obesity during treatment, and patients, in fact, need to be guided about the initiation and/or maintenance of healthy lifestyle habits. This could help stave off a second primary tumor and other chronic diseases, as well as contribute to a better quality of life.

APPENDIX - Food Frequency Questionnaire (FFA)