Nutritional Perceptions and Practices in Patients with Breast Cancer: A Qualitative Study in Northern Tanzania

Breast cancer is the most common cancer diagnosis in women worldwide [1]. Low-middle-income countries (LMIC) have a rising burden consisting of over half of the new diagnoses and deaths related to breast cancer [1, 2]. African countries have the lowest incidence rates of breast cancer, but the highest rates of breast cancer related mortality [2, 3]. The World Health Organization (WHO) acknowledges the disproportionate burden of disease in LMICs and recommends a multi-faceted, resource-appropriate approach to address it [1]. Treatment modalities of breast cancer include radiation therapy, chemotherapy, and surgery. In certain regions of Sub-Saharan Africa, treatment options are determined by access to resources such as a radiation oncologist or chemotherapy medications [4]. It is reported that 48-75% of female breast cancer patients in Sub-Saharan Africa undergo surgical treatment with modified radical mastectomy being the most common [4, 5].

Breast cancer risk is impacted by both modifiable (obesity and alcohol use) and non-modifiable factors (age, genetics, and hormonal exposure) which can be targeted through modalities including timely screening, education, and increased awareness [2, 6-9]. Currently, there is lack of strong evidence linking a specific food group or item and breast cancer risk [8]. However, an overall healthy diet focusing on whole grains, vegetables, fruit, and lean protein is recommended to reduce the risk of developing co-morbidities which could impact morbidity and mortality in patients with cancer [8, 10, 11]. Physical activity, body weight, and diet contribute significantly to cancer morbidity in LMICs although evidence as to its pathogenesis is still being studied [11]. The complexity of dietary decision making is rooted in both internal (knowledge, age, body weight) and external (socioeconomic status, education level, perceived needs) factors [12]. The multifactorial nature of nutrition only further emphasizes the WHO’s demand for systems level changes to address the inequities in breast cancer outcomes [1].

Countries in Sub-Saharan Africa are undergoing an epidemiological transition from communicable to non-communicable diseases, in part related to adoption of a Westernized diet [13-17]. Patients with cancer have been described as “nutritionally vulnerable” given the catabolic disease state and potential side effects of treatment [12]. Furthermore, the importance of nutrition in surgical patients cannot be understated as malnutrition is associated with longer hospital stays, increased risk of complications, worsened overall survival rates, and higher costs [18-21]. These outcomes further emphasize the importance of investigating the impact of nutrition in surgical oncology including breast cancer. This study explores the role of cancer nutritional beliefs and behaviors among breast cancer patients in northern Tanzania.

Study Setting

The study was conducted at the Cancer Care Centre (CCC) of the Kilimanjaro Christian Medical Centre (KCMC). KCMC is a tertiary teaching hospital located in Moshi, Kilimanjaro region, Tanzania. It is the only site offering comprehensive medical and surgical treatment for cancer patients in the Kilimanjaro region, a catchment area of 15 million people in both Tanzania and neighboring Kenya.

Patients are referred to the cancer center after receiving their diagnosis. The oncology department treats of over 100 new breast cancer patients each year. The breast cancer surgeries are done by general surgeons. The most common breast surgery performed is the modified radical mastectomy. Over the past four years, an average of 43 breast surgeries have been performed each year.

Study Design

This work was part of a larger mixed-methods study on understanding factors influencing late diagnosis of breast cancer among health care providers, patients, and traditional healers in Northern Tanzania.  Qualitative in-depth semi-structured interviews were conducted with breast cancer patients at CCC based at KCMC.

Recruitment of the Study participants

A purposeful sampling method was used to recruit the participants with histologically confirmed breast cancer, aged 18 years or older. The study was conducted from November 2020 through June 2021. Sample size was determined based on data saturation whereby recurrent patterns became evident in the participant descriptions [22-24].

In-depth Interviews

After obtaining informed consent, the semi-structured interviews were conducted in a private space at the CCC. The principal investigator conducted the in-depth interviews, while a sociologist acted as research assistant by taking notes during the interview. The interview guide consisted of open-ended questions about knowledge and beliefs on breast cancer etiology (Appendix A). All the interviews were audio tape-recorded. Field notes were taken during the in-depth interviews including observations and non-verbal cues from the participants. Interviews lasted between 30 and 60 minutes.

Data Analysis

All the audio tapes were transcribed verbatim in the language (Kiswahili) of the conducted interview, and transcripts were cross-checked for errors by reading of the transcripts together while audiotape played. The transcripts were then translated into English. The investigators read through each transcript to gain a sense of the participant’s perception at large. All the transcripts of the data were imported into NVivo software version 12 for computer-based data coding. Line by line thematic coding via grounded theory technique was done by three independent reviewers. Codes were compared based on their differences and similarities and summarized into categories and grouped into themes, which were analyzed using thematic analysis [25].

Quality Assurance

To enhance the credibility and reliability of the study, data was collected in different regions, involving rural and urban areas to compare information gathered in different contexts [26, 27]. Discussions were summarized with participants to ensure that they agreed with the content. Supervision of data collection, transcription, and translation were done by one PhD student and one social scientist and cross-checked by the supervisors who are fluent in both languages (Kiswahili and English). The study followed the consolidated criteria for reporting qualitative research (COREQ) guidelines for reporting qualitative research, including details of the research team, study methods, context of the study, findings, analysis, and interpretations [28].

Ethical Consideration

Ethical approval was granted from the local ethical reviewed board (ref: 2425), National review board (ref: NIMR/HQ/R.8a/Vol./X/3733) and from President’s Office, Regional Administration and Local Government of Tanzania (PO-RALG) (ref: AB.307/223/01).

Participant Characteristics

Twenty women with confirmed breast cancer participated in the study (Table 1). The average age was 56 years (range: 40-82), and fourteen participants (70%) had a diagnosis of stage IV breast cancer. Fourteen participants (70%) received surgical treatment of their breast cancer with four being completed at KCMC, nine at regional and district hospitals, and one by a traditional healer. Eight participants underwent partial mastectomy, one bilateral modified radical mastectomy, three unilateral modified radical mastectomy, and two unilateral simple mastectomy (Table 1). Relationship status varied among women, with seven (35%) being married. Fifteen participants (75%) identified as Christian. Eleven participants (55%) have a primary school education and three (15%) never attended school. Thirteen participants (65%) are unemployed.

Food Access

Food access is defined as having the legal, political, economic, and social resources to obtain adequate food [29]. Food access contributes to overall food security along with food availability, utilization, and stability [29]. Two (10%) participants emphasized that the cost of purchasing healthy food can be a significant barrier, especially for those with limited financial resources. This is evident from one participant who lives rurally stating the “money to buy fruits is not there (F, 57, Table 2)” and they “have to go to the neighbors to see if they know anyone with the fruit tree but whenever I cannot get them there is nothing I can do about it” (F, 57, Table 2).

Physical location of participants impacted their access to food, with some areas having limited availability of certain types of food. In general, participants commonly consumed maize porridge (ugali), bananas and stew (ndizi), green vegetables, meat, and fish. Seven participants (35%) noted getting their produce from their own garden or farm, whereas others needed to travel to the market. One individual who lives in an urban setting stated the following:

“…We are living on the zone of fish [Nyumba ya Mungu] so we eat more fish every day…It’s hard to get vegetables. [Vegetables] are not available as the place is desert…Per week can take vegetable once or not to get at all…” (F, 40).

Food Preparation

Two common methods to prepare food among participants included boiling and frying with reused oil (Table 3). Among participants who indicated reusing oil, the number of times it was reused ranged from two days to over a week. One participant (5%) mentioned using “the oil until it was black in color and it was used until it was finished” (Table 2). Several types of oil were used in food preparation including sunflower oil (Alizeti), soybean oil (Sunola), and olive oil (Safi) as well as animal oils (cow or goat oil).

Perceived Causes of Cancer Related to Diet

Animal Products

Upon exploring explanations for causes of cancer, five participants (25%) mentioned animal products including meat and food cooked with animal oil (Table 3). One participant stated that animal oil has been advertised as causing cancer on media platforms. Participants associated animal oil with unnatural components such as drugs, toxins, or chemicals present in food. Two participants (10%) specifically mentioned how animals are injected with medications, and they believed that eating this meat permits toxins to enter the body. One example of this concern is stated below:

“…Now these meats have brought to us cancers and that what we now eat… These animals currently are given drugs do you know which drugs they are given? Before the cows were not injected drugs. Currently, cows, chicken are injected drugs now when you eat, you also eat and those medicines and it enters in your circulation…” (F, 70, Table 2).

Toxins or Chemicals in Food

In addition to toxins associated with eating meat, participants stated that “vegetables from the market are grown with chemicals/poison” (F, 59, Table 2). In total, six participants (30%) expressed concern over the cancer-causing potential of chemicals and toxins in food (Table 3). This realization prompted one individual to start growing food at home without the use of harmful chemicals instead of relying on the market for fruits and vegetables.

Cooking Oil and Fried Food

Another perceived factor in causing cancer was eating fried food with one participant mentioning “…you find that cooking oil is not good, our fathers used to take natural foods…” (F, 48, Table 2). Another participant stated “…oil which have been reused for several purposes, thereafter, used for cooking can cause cancer…” (F, 60, Table 2). Overall, five participants (25%) raised the concern of cooking oil or fried foods as it related to causing cancer (Table 3).

Alcohol Use

Twelve participants (60%) indicated drinking Mbege and/or beer (Table 3), usually up to one liter per day. Mbege is a beer traditionally brewed by the Chagga ethnic group of Tanzania, and is made from fermented bananas and sorghum or millet [30, 31]. Most of the women who reported alcohol intake were social drinkers, but one participant indicated excessive alcohol consumption mentioning that “when I decided to drink, I could drink even 10 bottles” (F, 64). Five participants (25%) perceived alcohol use to be a cause of cancer as elucidated in the following quote (Table 3):

“…Another thing drinks like these hard drinks, as our local brew (Mbege) it was prepared naturally, our parents were drinking, a person boils his banana and keep, he boils his millet and keep, a person takes in the morning, I have never heard a person get a problem, these beers that we are taking we don’t know what they were made of, you don’t know its contents you wonder a person takes three beers and he is drunk, so it means it could be one thing contributing…” (F, 48, Table 2).

Farming and Coffee Plantations

Six participants (30%) reported farming for a living and growing coffee or having spent their childhood growing up on a coffee plantation (Table 3). Of these individuals, four (67%) discussed the use of pesticides, either by themselves or others, on the farm. One participant also explained how this could be a cause contributing to cancer:

“…It seems that cancer is caused by the life we live, perhaps for example agricultural work – the fields we have grown a lot and using these coffee poisons were probably poisonous to our body, like for me my family had coffee plantations, although I did not spray but, after spray we use to get in and pick coffee, so we use to get the smell, a lot…” (F, 60, Table 2).

Smoking and Body Weight

When asked what they think causes cancer, very few participants cited smoking and being overweight or obese as non-nutrition related perceptions. Although none of the patients were smokers themselves, two participants (10%) perceived smoking to be a cause of cancer. The following illustrates the role of the medical community in shaping an individual’s mindset:

“…Frankly I don’t know yet…Even in those days the doctor was questioning me about my eating habits because I had become overweight… Being overweight, the foods that we eat contribute to this problem. Drinking alcohol and smoking cigarettes contributes to someone getting cancer…” (F, 82, Table 2).

Modifications in Diet following Cancer Diagnosis

Change in Beverage Consumption

Following their cancer diagnosis, four participants (20%) reported decreasing their alcohol intake (Table 3). Two participants (10%) also mentioned that they stopped consuming soda. One patient discussed her reasons to stop drinking both beer and soda, exemplifying willingness for behavior change following proper nutrition education:

“…They say that beer has too much alcohol and so I would only use soda and juice… now I just drink fruit juice after discovering that I have cancer; that is when they started educating us, saying that sodas are not good since they have gases and also that they make poison…” (F,41, Table 2).

Food Preparation

Two participants (10%) noted shifting to boiling food instead of frying as they had done in the past, as explained from the perspective of one patient (Table 3):

“…At the beginning, out of ignorance we liked to fry our food but later on we changed, and we boil our food and even with green vegetables, I cut my onions, green pepper, carrots and I mix them together before boiling them. If it is bananas, after boiling the meat or if it is fish I cut everything, I want to mix with it and boil it together. That is how I like it nowadays…” (F, 82, Table 2).

Vegetable Consumption

Eight participants (40%) included green vegetables as part of their diets (Table 3). Vegetables were fried or boiled to eat alone or incorporated into their main meals. Commonly cited vegetables include potatoes, beans, legumes, and spinach. Three participants (15%), including those who did not prefer the taste of green vegetables previously, increased vegetable consumption after diagnosis (Table 3).

“…Green vegetables [are] what we are supposed to eat and it’s what we are eating now… Before that we weren’t using that often, because we didn’t know the meaning of it, we remained using meat… We didn’t know a lot about green vegetables…” (F, 70, Table 2).

This is one of the few studies exploring patient perceptions and behaviors relating to nutrition and perceived breast cancer risk. It shed light on both daily dietary habits and perceptions of nutrition-related causes of cancer, which may explain dietary changes after receiving their breast cancer diagnosis. Animal products, food toxins and chemicals, food preparation, and alcohol use were mentioned as causes of breast cancer relating to nutrition by many participants. Few participants also identified body weight and tobacco use, known risk factors for cancer, as potential causes of breast cancer. Following a cancer diagnosis, participants commented on dietary changes such as increasing vegetable consumption, decreasing alcoholic beverage and soda consumption, and altering food preparation. These findings demonstrate misconceptions in this population surrounding breast cancer causation but also exemplify patients’ willingness to make behavioral changes after receiving a breast cancer diagnosis.

Our study demonstrated a misconception of the relationship between nutrition and cancer with women identifying certain food groups, additives, or preparation methods to be the cause of breast cancer. The Academy of Nutrition and Dietetics (AND) acknowledges the role of an overall healthy diet in the prevention of cancer as well as diet both during and after cancer treatment; however, conflicting evidence exists regarding individual food items’ impacts on the risk of breast cancer development [8, 10, 32, 33]. Although not specific to breast cancer, AND’s recommendations mirror in some ways the causes the participants identified, and changes participants made after receiving their diagnosis. For example, AND recommends limiting processed foods high in fat and sugar while our participants identified food additives, alcohol, soda, and frying foods as causes of breast cancer [10]. Cooking practices may contribute to carcinogenesis through the exposure to mutagenic compounds in certain food preparations including deep-frying [34]. Some participants continued to reuse their frying oil theoretically compounding the production of mutagenic compounds; whereas other participants recognized the potential risk associated with frying foods and shifted to boiling them.

Patients’ perceptions of risk have been identified as a clinical challenge contributing to breast cancer prevention as their perceptions motivate health protective behaviors [35-38]. Participants in this study altered their beverage consumption, food preparation, and vegetable consumption after receiving a diagnosis of breast cancer. These changes aligned with their perceived cause of cancer which exemplify attempts at adopting a healthier diet. A mixed methods review including international articles emphasized that patients are most willing to engage in health protective behaviors within the health care setting, whereas lifestyle changes were less willingly adopted [39]. Our study challenges these findings with most of our participants making dietary changes after receiving a cancer diagnosis. On the other hand, a recent study demonstrates patient’s willingness to optimize nutritional behaviors following the diagnosis of cancer similar to our findings [12].

The majority of our participants identified a component of nutrition as a cause of cancer but did not differentiate between risk and causation as it relates to nutrition. Alcohol remains to be the only food group with an association for cancer development consistently supported in the literature [8]. In addition to diet, it is well established that the use of tobacco products increases one’s risk of cardiovascular disease, lung cancer, and other non-communicable diseases [40-42]. However, our population held a misunderstanding related to breast cancer causation rather than identifying alcohol, being overweight or obese, and tobacco use as risk factors for breast cancer development. The lack of differentiation between a risk factor and a cause is not unique to our participants, but a common misunderstanding in the general population [43, 44].

Our patient population demonstrated the willingness and ability to make lifestyle modifications related to nutrition in response to their cancer diagnosis. Our findings align with an international study including 4 countries from North America and Europe which described similar post-diagnosis dietary changes [12]. The time of cancer diagnosis is described as a “teachable moment” or an opportunity to increase health literacy related to both cancer and nutrition [12]. However, ideally education regarding cancer risk reduction would be received prior to diagnosis to have an impact. In high income countries, multiple mass media campaigns have been employed to provide education on healthy lifestyle behaviors ranging from alcohol intake, nutrition practices, and solar UV exposure and concluded to be promising in their study population [45-47]. The question remains as to whether media campaigns would be efficacious in LMICs; notably one participant reported receiving nutrition information by means of the television and radio. Education programs related to nutrition and cancer risk reduction in sub-Saharan Africa are limited in the literature. However, nutrition interventions targeting an overall healthy diet have been widely documented specifically in school settings [48-50]. A systematic review in Sub-Saharan Africa highlighted a crucial point that nutrition related decision making is multi-factorial [51]. Although school-based nutrition education increases knowledge it does not always translate into eating patterns given the role of external factors including socioeconomics, food access, and policy [12, 51]. More studies are needed in the Sub-Saharan African population to establish a multi-factorial approach to changes in nutrition practices specifically as it relates to cancer risk reduction.

It is also important to acknowledge study limitations. By nature of our study design, we are limited by recall and survivorship bias as in-depth interviews are our only source of data. Although our number of participants was determined based on data saturation, there still exists a limitation in generalizability. Therefore, the study must be interpreted in the appropriate lens as our population consists of women from the Northern regions of Tanzania, and the results cannot be generalized to other parts of the country or continent due to the diverse ethnic populations which reside in various regions. Given the retrospective nature of our study design, we are also limited on establishing associations between nutrition and surgery. An opportunity exists to explore the impact of nutrition on surgical decision making as well as surgical recovery. Lastly, all our participants were female, so further research which includes male patients with breast cancer is needed.

Implications for practice

Our study uniquely discovered that many patients independently modified their diet towards healthier choices after breast cancer diagnosis highlighting the opportunity to leverage this self-motivation to mitigate breast cancer risk factors. However, findings also demonstrated misconceptions among participants around nutrition and breast cancer risk, emphasizing the need for evidence-based patient education. Incorporating nutrition education would ensure a more comprehensive approach to breast cancer prevention and treatment while optimizing patient outcomes, especially in low-resource settings.

In conclusion, our results revealed that participants displayed a high level of concern about the role of nutritional factors in influencing their breast cancer development. Most importantly, their beliefs were strong enough to motivate many to change their behaviors, ultimately adopting a healthier lifestyle. A misconception existed among our study population regarding the causation of cancer based in nutrition. However, many participants correctly identified risk factors for cancer, including alcohol and tobacco use and being overweight or obese. Culturally sensitive education and community engagement is needed to ensure appropriate changes are being made on an individual level following diagnosis of breast cancer in the northern Tanzanian community. These educational interventions will harness the willingness of participants to make lifestyle modifications and potentially reduce the burden of breast cancer in this area.

To all the study participants who accepted to take part in our study. Dr. Glory Makupa (Clinical Oncologist), Kilimanjaro Christian Medical Centre, Kilimanjaro Clinical Research Institute (KCRI) and KCMC University for supporting EM, The University of Alabama (UAB), Global Health Working Group, Martin-Luther-University Halle-Wittenberg (Grant numbers [81295996]) program, German Federal Ministry of Education and Research 01KA2220B and Science for Africa Foundation to the Developing Excellence in Leadership, Training and Science in Africa (DELTAS Africa) program [Del-22-008] with support from Wellcome Trust and the UK Foreign, Commonwealth & Development Office and is part of the EDCPT 2 programme supported by the European Union for partly supporting this work.