Alaini et al.

BMC Public Health 2019, 19(Suppl 4):546

https://doi.org/10.1186/s12889-019-6872-4

RESEARCH Open Access

Diet optimization using linear programming

to develop low cost cancer prevention food
plan for selected adults in Kuala Lumpur,
Malaysia
Reham Alaini1, Roslee Rajikan1* and Siti Masitah Elias2

Abstract
Background: Poor dietary habits have been identified as one of the cancer risks factors in various epidemiological
studies. Consumption of healthy and balance diet is crucial to reduce cancer risk. Cancer prevention food plan
should consist of all the right amounts of macronutrients and micronutrients. Although dietary habits could be
changed, affordability of healthy foods has been a major concern, as the price of healthy foods are more expensive
the unhealthy counterparts.
Methods: Therefore, using linear programming, this study is aimed to develop a healthy and balanced menu with
minimal cost in accordance to individual needs that could in return help to prevent cancer. A cross sectional study
involving 100 adults from a local university in Kuala Lumpur was conducted in 3 phases. The first phase is the data
collection for the subjects, which includes their socio demographic, anthropometry and diet recall. The second
phase was the creation of a balanced diet model at a minimum cost. The third and final phase was the finalization
of the cancer prevention menu. Optimal and balanced menus were produced based on respective guidelines of
WCRF/AICR (World Cancer Research Fund/ American Institute for Cancer Research) 2007, MDG (Malaysian Dietary
Guidelines) 2010 and RNI (Recommended Nutrient Intake) 2017, with minimum cost.
Results: Based on the diet recall, most of subjects did not achieve the recommended micronutrient intake for fiber,
calcium, potassium, iron, B12, folate, vitamin A, vitamin E, vitamin K, and beta-carotene. While, the intake of sugar
(51 ± 19.8 g), (13% ± 2%) and sodium (2585 ± 544 g) was more than recommended. From the optimization model,
three menus, which met the dietary guidelines for cancer prevention by WCRF/AICR 2007, MDG 2010 and RNI 2017,
with minimum cost of RM7.8, RM9.2 and RM9.7 per day were created.
Conclusion: Linear programming can be used to translate nutritional requirements based on selected Dietary
Guidelines to achieve a healthy, well-balanced menu for cancer prevention at minimal cost. Furthermore, the
models could help to shape consumer food choice decision to prevent cancer especially for those in low income
group where high cost for health food has been the main deterrent for healthy eating.
Keywords: Balanced diet, Cancer prevention, Linear programming

* Correspondence: [email protected]
1
Dietetics Programme, Faculty of Health Sciences, Universiti Kebangsaan
Malaysia, 50300 Kuala Lumpur, Malaysia
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 2 of 8

Background Currently there is no study done in Malaysia to create

A balanced diet can be categorized with the concept of a balanced, optimal diet for cancer prevention using
diversity and simplicity that is related with the percep- linear programming at a minimum cost. Hence, this
tion of healthy eating [1]. Healthy diets are obtained by study aimed to construct a balanced, palatable and af-
taking foods that contain recommended dose of macro- fordable diet that helps to prevent cancer at the lowest
nutrients and micronutrients [2]. Dietary Guidelines can price possible to make it affordable and achievable for
be adapted to promote health and preventing diet-re- low income individuals.
lated chronic diseases including cardiovascular disease,
type 2 diabetes, some cancers, and obesity [3]. Through- Methods
out recorded history, wise choices of food and drink and This study was approved by the National University of
of habitual behavior have been recommended to protect Malaysia Medical Research Ethics Committee (UKM-
against cancer as well as other diseases, and to improve REC) (NN-2017-128). One hundred healthy adults aged
wellbeing. In Malaysia cancer is one of the major health 19 and above both male and female from a local
problems [4]. It is undeniably one of the most important university in Kuala Lumpur, staff and students, were ran-
non-communicable diseases in Malaysia and contributed domly selected to participate in this study. Dietary intake
to 13.56% of all deaths occurred in the Ministry of and socio-demographic data were collected between
Health Hospitals in 2015 [5]. Nutrition and related September–October 2017.
factors such as physical activity, obesity believed to con-
tribute crucially to cancer occurrence [6]. Data collection
Socioeconomic status of the population plays a critical A set of questionnaires was distributed to the subjects to
role in eating patterns and food choices. Several studies assess their socio-demographic profile. Basic information
have shown that energy-dense foods are foods that are such as age, gender, marital status, and education level,
commonly chosen by the low socioeconomic class due lifestyle of the participants such as smoking habits, his-
to their cheaper price [7] as financial resources are lim- tory of weight and height were obtained. Body mass
ited [8–11]. Foods of lower nutritional value and index (BMI) was calculated by using weight and height
lower-quality diets generally cost less per calorie and and classified based on WHO BMI classification into
tended to be selected by groups of lower socioeconomic underweight, normal, overweight and obesity in adults
status. A number of nutrient-dense foods were available [18]. Eating pattern for both male and female staff and
at low cost but were not always palatable or culturally students of a local university in Kuala Lumpur, were
acceptable to the low-income consumer [12]. Further- recorded and assessed by three-day food records, where
more, high price of healthy foods has been one of the the subjects had to document their food intake for two
main deterrents for healthy eating among the lower in- days in a weekday and one over the weekend. Based on
come group [13–15]. The encouragement toward healthy, the subject’s diet history information, a food list is pre-
balanced, affordable and palatable diet among the pared and the price for each food items was obtained
low-income group may improve their overall health status from the Ministry of Domestic Trade, Cooperative, and
and eventually reduce the prevalence of chronic diseases Consumerism (KPDNKK). The food prices were set in
in Malaysia including cancer. terms of price per serving size.
Linear programming can be used to formulate minimum
cost menus while making sure it met all the criteria of all
macronutrients and micronutrients that has been set by Statistical analysis
dietary guidelines [16]. It is used in diet problem-solving Dietary intake was initially analyzed using Nutritionist
techniques by creating a model that contains all the optimal Pro™ software version 4.0.0 and compared with Recom-
food, cost and quality of a diet. By using linear program- mended Nutrient Intake (RNI) 2017 for Malaysian [19].
ming, the majority of populations can benefit from an opti- The data was needed to calculate their energy intake and
mal diet at a minimum cost, which enables them to have to assess eating pattern in order to plan a diet that
an adequate daily nutrition within their financial potentials. emphasizes their preferences. Then both socio-demo-
Linear programming has been applied in the Pacific graphic and dietary recall data were analyzed using stat-
Northwest of the USA, which was the only study that istical products and service solution (SPSS) program
presents an application of mathematical optimization version 23.0. Descriptive analysis, which includes the
tools of dietary guidelines for cancer prevention. Six-spe- mean, percentage and standard deviation, was used to
cific food plans were generated that met both the key find the average with its standard deviation. All the food
2007 dietary recommendations for cancer prevention is- items in the food list prepared were analyzed using
sued by the WCRF/AICR 2007 and the DRIs set by the Nutritionist Pro™ software and market survey to deter-
Institute of Medicine [17]. mine the nutrient content i.e. energy, macronutrients
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 3 of 8

and most micronutrients. Lastly, Excel Solver was uti- was 25 ± 6 years, where more than 59% were between 19
lized to produce the linear programming model. to 29 years old. More than two-third (79%) was single
Before running the program, the details of each and the rest (21%) were married. Based on the results,
macronutrient and micronutrient of food items, price all of the subjects completed at least secondary schooling
per serving size, were filled in Microsoft Excel. The next while 20% have completed a bachelor’s degree or higher.
step was setting up the constraints in the model such as
upper bound (UL) and lower bound (LB) for energy, Anthropometric measurements
macronutrients and micronutrients. Excel Solver was Anthropometric measurements were taken from the
later used to determine the optimal foods portion size subjects. After the weight and height of the subjects
that met all the dietary and nutrient recommendation of were obtained, Body Mass Index (BMI) was calculated
WCRF/AICR 2007 [20], MDG 2010 [21] and RNI 2017 for each subject. Majority of the study population (48%)
[19] with the lowest possible cost. From the suggested had normal BMIs between 18.5 kg/m2 and 24.9 kg/m2,
foods portion, a daily balanced menu was later planned which were almost similar to the prevalence of normal
The optimization model will be repeated several times body weight of healthy Malaysian adults of 45.6% (Na-
to produce two more suggested palatable menus with tional Health Morbidity Survey 2015) [23]. On the other
the lowest possible costs. hand, 33% of the subjects had BMIs between 25 kg/m2
and 29.9 kg/m2, thus belonging to the overweight classi-
Linear programming model development fication. Prevalence of overweight for healthy Malaysian
Cancer prevention diet models with the lowest cost were adults was 30%, which is almost similar to the subjects
planned. The formulation for Linear Programming is as (National Health Morbidity Survey 2015) [23]. The
follows: remaining 19% were classified as obese.
X
Minimize : z ¼ cx Dietary intake
X j j
Subject to : bi ≤ aij x j ≤bi and x j ≥ 0 All the macronutrients average intake of the subjects
was met as shown in Table 1. Similarly, according to
The portion size of food item j is represented as xj; aij MANS 2008, the proportions of calories derived from
denotes the amount of nutrient i in one portion of food macronutrients were within the recommendations for a
item j; cj was the cost of a portion of food item j; bi healthy diet, while intake of micronutrients such as iron,
denotes the largest or smallest acceptable quantity of nu- calcium and vitamin A was about 50% of RNI particu-
trient i. The constraints in the model for this study were larly in women [24].
WCRF/AICR 2007 [20], MDG 2010 [21] and RNI 2017 Fiber intake was only 7.5 ± 4.05 g, thus failing to meet
[19]. Palatability constraints were also included to ensure the prescribed recommendations (Table 1). Only 30% of
that the suggested menus were suited to the subjects’ the subjects meet the recommendations for fiber. The
common food pattern. average intake of non-starchy vegetables and fruits were
In this study, the cost of food items (z) is the objective below the recommended amount of WCRF/AICR 2007
function that we want to minimize. Subjects’ energy (2.4 versus 5 serving per day). In addition, the average
needs were calculated before the linear programming intake of unprocessed grains and legumes were also
program started. The ideal energy of the subject was below the recommendation (0.7 versus 3 serving).
calculated using the Mifflin St-jeor Formula [22]. The Malaysian Adult Nutrition Survey (MANS) 2014 [25] re-
minimum and maximum value of macronutrient and vealed that Malaysian adults on average do not consume
micronutrient were set based on WCRF/AICR 2007 sufficient fruits and vegetable in terms of frequency and
[20], MDG 2010 [21] and RNI 2017 [19]. Choosing food amount, therefore does not achieve the recommended
items from the dietary recall of the subjects and avoiding intake of fibers and other micronutrients. In this study,
the repetition or large portions of certain foods were no subject met the requirement for iron and folic acid.
also considered to ensure the palatability of the menu. Other micronutrients intakes such as calcium, vitamin
B3, B12, vitamin C, vitamin E, vitamins K were also
Results poor. Similarly, the MANS 2014 [25] also reported that
Socio-demographic data the intake of micronutrients in relation to RNI could be
A total of 100 subjects from staff and students of partici- described as low particularly for calcium and vitamin C
pated in this study, with the females constitute the ma- intake. Healthy Eating Index for Malaysians showed that
jority (65%) and the remaining balance of 35% was male. only a small percentage of Malaysian met dietary re-
The mean age of female subjects was 24 ± 5.5 years, quirements and found that majority of the respondents
where more than 50% of female were between the ages (80.7%) were at risk of poor diet quality [26]. As for zinc,
19 to 29 years. While, the mean age for male subjects selenium and phosphorous, all subjects achieved the
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 4 of 8

Table 1 The average energy/macronutrients/micronutrients intake with the recommendations and percentage of subjects achieving
recommendations
Energy/Macronutrient / Average Intake ± Percentage of Intake from Recommendation % Of subjects achieving
Micronutrient SP Energy Recommendation
Energy (kcal) 1580 ± 300 70.5
Protein (g) 56 ± 18.5 14.7% ± 4.6% 10–20%3 78
3
Carbohydrates (g) 203 ± 45.8 53.5% ± 11.2% 50–65% 66
Fat (g) 48 ± 17.09 28.5% ± 9.5 25–30%3 59
Cholesterol (mg) 184 ± 91 – < 300 mg/day 3
89
Saturated Fat (g) 15 ± 6.8 9% ± 1% < 10%1 79
3
Monounsaturated (g) 11 ± 5 6.5% ± 2.5% 12–15% 2
Polyunsaturated (g) 9±4 5.3% ± 2.5% 3–7%3 33
Fibre (g) 7.5 ± 4 – 25–30 1
0
Sodium (mg) 2585 ± 544 – < 20001 20
Potassium 1181 ± 483 – ≥ 4700 3
0
Vitamin A 954 ± 812 – 600–30003 80
Beta-Carotene 1061 ± 813 – 2500 1
6
Vitamin E (mg) 4 ± 2.3 – 7.5–10003 10
Vitamin K (μg) 32 ± 20 – 55–1000 3
15
Vitamin C (mg) 36 ± 25 – 70–20003 8
Vitamin B12 (μg) 2 ± 1.5 – ≥4 3
20
Folate (μg) 124 ± 76.6 – 400–10003 2
Iron (mg) 13 ± 5.2 – 29–45 3
0
Calcium 418 ± 226 – 1000–25003 0
Thiamine (mg) 1 ± 0.2 – ≥1.1 3
33
Riboflavin (mg) 1 ± 0.9 – ≥ 1.13 45
Niacin (mg) 11 ± 4 – 14–35 3
26
Zinc (mg) 5.5 ± 1.61 – 4.7–353 30
Magnesium (mg) 124 ± 70.8 – 320–500 3
0
Copper (mg) 0.6 ± 0.5 – 0.9–23 10
Selenium (μg) 35 ± 18.4 – 24–400 3
66
Phosphorus (mg) 822 ± 322 – 700–30003 67
Sugar (g) 51 ± 19.8 – 3
< 10 (50 g) 44
Source: 1WCRF/AICR 2007, 2MDG 2010, 3RNI 2017

recommended requirements set by the RNI 2017. While Malaysian adults. Red meat consumption is associated with
for potassium, the average intake of a subject was the formation of N-nitroso compounds. This increases the
1181 ± 480 mg, and none of the subjects met the recom- level of nitrogenous residues in the colon and is associated
mendation requirement of the RNI 2017, which is 4700 with the formation of DNA adducts in colon cells. High in-
mg. The average intake for sugar and sodium were (51 ± take of red meat may result in more absorption of haem
19.8 g) and (2585 ± 544 g) respectively, which exceeds iron, greater oxidative stress and potential for DNA dam-
the recommendations. age. Beside, red meat is high in animal fat and is energy
The average intake of red and processed meat is less than dense food. All these factors contribute for considering red
300 g per week, meeting the recommendations of WCRF/ and processed meat as a cause of colorectal cancer [27].
AICRF 2007. However, the intake of processed food and
salt was higher than the recommended amount. Likewise, Development of cancer prevention model and menu with
MANS 2014 [25] showed that consumption of “processed minimum cost
foods” added with salt and condiments had increased and Linear programming has been used to formulate nutri-
appeared among the top ten most consumed foods by tionally optimal dietary patterns, to examine the
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 5 of 8

Table 2 Comparison of the three models produced with WCRF/AICR, MDG, RNI and palatability constrains
Constrain LB UB Model I Model II Model III
WCRF/AICR 2007
Non-Starchy Vegetables and fruits (serving/day) 5 – 13 14.5 11.5
Cooked Vegetables – – 6 3.5 6
Salad – – 1 2 –
Spices – – 4 6 3.5
Fruits – – 2 3 2
Red and Processed meat (serving/week) – 3.5 0 0 0
Unprocessed grains and legumes (serving/day) 1 – 2 2 5
Fibre (g) 25 35 32.3 26 26.3
Sodium (g) 500 2000 563 1631 1853
MDG 2010
Cereals and Grains (serving) 6 8 6.5 6 6
Meat/Poultry (serving) 1 2 1 1 2
Fish (serving) 1 3 1 1 2.5
Legumes (serving) 0.5 1 1.5 1 1
Palatability
Vegetable oil (serving) 1 2 1
Palm Oil (serving) 1 2 2 2
RNI 2017
Energy (kcal) 1600 2000 1802 1685 1889
Protein (g) 42 100 69.5 58 102
Carbohydrates (g) 212 300 224 218 236
Fat (g) 38 68 62 59 56
Cholesterol (mg) 60 200 135.7 116 217
Saturated Fat (g) 10 22 16.5 10.5 19.9
Monounsaturated (g) 22 33 23 19.8 23.6
Polyunsaturated (g) 10 15 15 13.7 13.5
Potassium (mg) 4700 10,000 4729 4696 4704
Vitamin A (RE) 600 3000 4826 1913 1409
Beta-Carotene 2500 17,000 10,430 10,516 6365
Vitamin E (mg) 7.5 1000 12 10.2 13.2
Vitamin K (μg) 55 1000 55 71.6 99.7
Vitamin C (mg) 70 1000 549 233 274
Vitamin B12 (μg) 4 23 11.5 7.3 6.6
Folate (μg) 400 1000 413 450 475
Calcium (mg) 1000 2500 1000 1143 1113
Iron (mg) 29 45 29.5 30 29
Thiamine (mg) 1.1 500 10 2.2 1.8
Riboflavin (mg) 1.1 25 3.4 2.3 3.2
Niacin (mg) 14 35 32 29 28.5
Zinc (mg) 4.7 35 6.1 10.6 17.7
Magnesium (mg) 320 500 227 342 298
Copper (mg) 0.9 2 1.4 1.1 1.5
Selenium (μg) 24 400 47 43 23
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 6 of 8

Table 2 Comparison of the three models produced with WCRF/AICR, MDG, RNI and palatability constrains (Continued)
Constrain LB UB Model I Model II Model III
Phosphorus (mg) 700 4000 1370 1507 1608
Sugar (g) 0 50 47 48 45

relationship between diet cost and diet quality in West- limit of the constraint values. However, other nutrients
ern countries and to develop food-based dietary guide- such as carbohydrate (CHO), fat, vitamin A and fiber
lines in developing countries where residents need to reached the upper limit of the maximum acceptable
achieve nutritional requirement with their limited in- value of constraints.
come of diet [28]. Similarly, a Malaysian study done by The food list selected comprised mainly on fruits and
Rajikan et al. developed a healthy and palatable diet for vegetables with the highest serving, as complex mixture
low income women at the minimum cost based on of phytochemicals present in whole vegetables and fruits
Malaysian Dietary Guidelines 2010 and Recommended may have additive and synergistic effects responsible for
Nutrient Intake 2005 via linear programming [29]. anti-cancer activities [6]. Beside, these food items are
Optimization models provide an elegant mathematical low-energy density and high in fibers, which will provide
solution that can help to determine that a set of dietary sufficient fibers to meet the recommendations of 25-g
guidelines is achieved by Malaysian population sub- fibers by WCRF/AICR and meeting the new recommen-
groups. There were three models produced by linear dation of RNI 2017 for potassium.
programming. Table 2 shows all nutrients constrains and From the suggested food list of the models, it is under-
the food groups of the three different models produced stood that each model consisted of at least two servings
by LP based on the dietary guidelines of WCRF/AICR of whole and unprocessed grains such as brown rice,
2007, MDG 2010 and RNI 2017. The three models pro- oat, lentils, and whole meal bread, thus ensuring high
duced, fulfilled the upper and the lower limits of the fiber and nutrient contents. The food list for each model
constrains including macronutrient and micronutrient also provides at least two servings of fruits and more
recommendations set by WCRF/AICR 2007 and RNI than nine servings of vegetables, although it resulted in
2017, the serving size of the food groups based on slight variation of the existing diets. In consequence, an
WCRF/AICR 2007 and MDG 2010. The palatability optimal cancer prevention menu is developed based on
factor was also considered by including servings from a model that was produced using a linear programming
vegetable oil and palm oil. method where it meets the requirements of constraints
Looking at the three LP models as shown in Table 2, based on the dietary guidelines of WCRF/AICR 2007,
iron, potassium and calcium only reached the lower MDG 2010, and RNI 2017.
Table 3 Developments of three cancer prevention menus with minimal cost
Meal Menu 1 Menu 2 Menu 3
Breakfast Fried Kuey Teow Laksa Penang with sardine (1 ½ Hailam Noodle with chicken (1 cup)
(1 cup) cup) Banana (1 piece)
Papaya (1 slice) Pineapple (1 cup)
Morning snack Bubur Kacang Hijau dan Oat cookies Cream Crackers
Keledek with sweet potatoes (1 ½ cup) (2 pieces) (3 pieces)
Coffee plain (1 cup)
Lunch Steamed Brown Rice (3scoops) Steamed Basmati Rice (3 scoops) Basmati Rice (3scoops)
Fried Indian Mackerel (1 medium piece) Soy Sauce Chicken (1 medium Grilled Catfish + Air Asam (medium piece)
Spinach Soup with tomato (1cup) piece) Sambal Lady Finger (1 cup)
Ulam Raja (1 cup)
Steamed cabbage
(1 cup)
Afternoon snack Capati (1 piece) Banana Oat Smoothie with milk Peanut Butter Sandwich (1 sandwich)
Dhal Gravy with turmeric and carrots (1 ½ (1 cup) Chocolate Drink (1 cup)
cup)
Dinner Steamed Brown Rice (3 scoops) Fried Rice with sardine (3 scoops) Basmati rice (3 scoops)
Grilled Chicken Breast (2 pieces) Ulam Raja (1 cup) Masak Lemak with tempeh and vegetables (1
Fried Pucuk Paku Stir fried Red spinach cup)
(1 cup) Orange/carrot juice Mango (1 piece)
Guava (1 piece) (1 cup)
Food Cost per day RM 7.8 RM 9.2 RM 9.7
(RM)
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 7 of 8

Table 3 shows the three menus produced, according to antioxidant drinks were suggested such as carrot and or-
raw food items at the lowest possible cost based on ange juice. Furthermore, it is evident that the diet models
WCRF/AICR, MDG, RNI and palatability constraints by do not include any processed meat, fast food, or sugary
using LP. The production of every menu is different drinks; where lean proteins were the only protein source.
from another as it follows the list of food ingredients Looking at the fat content in the three models, we can
selected according to the LP models. For each model, see that it emphasizes on less saturated fats and trans-fat
every list of ingredients included in the model will have by reducing the consumption of fat, which is mainly
a slight difference by removing food items that have achieved by appropriate cooking methods. Based on the
been selected in the previous menus or placing limits on menu that has been set up, almost all models use a min-
the same food from a model to the next model so that imal of 3 tablespoons of oil. Therefore, the menu is
quantities are different or not selected by the next provided with many ways of cooking such as steaming,
model. Therefore, the price is expected to increase from baking or grilling. In a study conducted by Asmaa et al.
menu 1 to menu 3 as the models have stricter require- [33], the use of healthy cooking methods such as steam-
ment and the cheapest nutritionally dense foods have ing can reduce fat content in food.
been chosen in the previous model. The development of However, there were few limitations in this study. The
these menus had different energy levels within the lower subjects in this study may have not been representative
and the upper limit set for kcal (1600–2000 kcal). Model because they were not randomly sampled from the
1 is 1802 kcal with the lowest price RM 7.8 (USD 1.97), general Malaysian population, rather, they were only lim-
model 2 is 1680 with RM9.2 (USD 2.35) and model 3 is ited to a local university staff and students. A larger
1889 Kcal with the highest price RM 9.7(USD 2.48). number of subjects were from different economic and
social background and thus more lists of food items
Discussion should be included in the model to increase the variety
Diets that give more emphasis to those plant foods that of food choices in future studies.
are high in nutrients, high in dietary fiber, and low in en-
ergy density, (non- starchy vegetables & fruits) probably
protect against some cancers such as: mouth, pharynx, Conclusion
esophagus, stomach colorectal, lung, pancreas and pros- In general, the use of linear programming is a very
tate [20]. These foods are high in antioxidants (caroten- effective tool in producing a balanced diet and can easily
oids, beta-carotene, lycopene and Allium) such as, pink interpret dietary recommendations into a nutritional
sweet potatoes, papaya, tomato, onions, garlic, mango, model that is based on local market prices. It formulated
carrots and fiber, which are low in energy density, and the current guidelines for cancer prevention by creating
so, promote healthy weight. a balanced and optimal diet for cancer prevention at
In addition, we can observe that the menu also em- minimum cost with more specific details and accuracy.
phasizes on the intake of cruciferous vegetables, such as In addition, because this research focuses on the specific
broccoli, mustard leaves, cabbage and cauliflower which nutrients needed at minimal cost, the menus produced
are associated with the reduction in the risk of several are ideal for people who want to maintain healthy eating
types of cancer [30]. The traditional food tempeh, which habits but experience financial difficulties.
is rich in phytoestrogens, is also included in the menu as
seen in menu 3, as it is found to exhibit a plethora of Abbreviations
different anti-cancer effects, including inhibiting prolif- AICR: American Institute for Cancer Research; BMI: Body Mass Index;
DNA: Deoxyribonucleic Acid; DRI: Dietary Recommended Intake;
eration [31]. Developing cancer prevention diet requires KPDNKK: Ministry of Domestic Trade, Cooperative, and Consumerism;
little modification from the existing diet mainly by in- LB: Lower bound; LP: Linear Programming; MANS: Malaysian Adult Nutrition
creasing the vegetables and fruit serving. Survey; MDG: Malaysian Dietary Guidelines; NHMS: National Health Morbidity
Survey; RNI: Recommended Nutrient Intake; UKMREC: Universiti Kebangsaan
The recommended menu used only a dash of salt in Malaysia (National University of Malaysia) Medical Research Ethics
seasoning the dishes and this causes the resulting menu Committee; UL: Upper bound; WCRF: World Cancer Research Fund
to have a low sodium amount of less than 2000 mg.
Studies showed that salt and salt-preserved foods are Acknowledgments
probably a cause of stomach cancer [32]. The other The authors would like to thank all of the subjects for their cooperation and
alternatives are to use natural flavoring to replace salt support toward this research project.
are turmeric, onions, garlic, chili and mustard leaves that
contain lower sodium content. Funding
The menu also restricted the use of added sugar and the The authors acknowledged the financial assistance for publication received
from the Research University Grant awarded by the Ministry of Health to the
intake of sugary drinks, except for sugar that is naturally National University of Malaysia specifically for the Consortium of B40
found in fruits and vegetables. Instead, healthy high Research (CB40R) under the auspice of B40 Grand Challenges (IDE 2018–01).
Alaini et al. BMC Public Health 2019, 19(Suppl 4):546 Page 8 of 8

Availability of data and materials 10. Drewnowski A, Darmon N, Briend A. Replacing fats and sweets with
The datasets used and/or analyzed during the current study are not publicly vegetables and fruits—a question of cost. Am J Public Health. 2004;94(9):
available due to limited consent of the study participants but are available 1555–9.
from the corresponding author on reasonable request. 11. Dowler E. Budgeting for food on a low income in the UK: the case of lone-
parent families. Food Policy. 1997;22(5):405–17.
About this supplement 12. Darmon N, Drewnowski A. Contribution of food prices and diet cost to
This article has been published as part of BMC Public Health Volume 19 socioeconomic disparities in diet quality and health: a systematic review
Supplement 4, 2019: Health and Nutritional Issues Among Low Income and analysis. Nutr Res. 2015;73(10):643–60.
Population in Malaysia. The full contents of the supplement are available 13. de Mestral C, Stringhini S, Marques-Vidal P. Barriers to healthy eating in
online at https://bmcpublichealth.biomedcentral.com/articles/supplements/ Switzerland: a nationwide study. Clin Nutr. 2016;35(6):1490–8.
volume-19-supplement-4. 14. Lennernäs M, Fjellström C, Becker W, Giachetti I, Schmitt A, De Winter A,
Kearney M. Influences on food choice perceived to be important by
nationally-representative samples of adults in the European Union. Eur J
Authors’ contributions
Clin Nutr. 1997:51.
RR and SME conceived of the presented idea. RA did the data collection,
15. Glanz K, Basil M, Maibach E, Goldberg J, Snyder D. Why Americans eat what
performed the computations, analysed the data and drafted the manuscript.
they do: taste, nutrition, cost, convenience, and weight control concerns as
RA and SME verified the analytical methods and supervised the findings of
influences on food consumption. J Am Diet Assoc. 1998;98(10):1118–26.
this work. All authors provided critical feedback and helped shape the
16. Pasic M, Catovic A, Bijelonja I, Crnovrsanin S. Linear programming local cost nutrition
research, analysis and manuscript. All authors have read and approved the
optimization model: Annals of DAAAM & Proceedings; 2011. p. 0389-90. www.
manuscript.
daaam.info/Downloads/pdf/proceedings/proceedings_2011/0389_Pasic.pdf.
17. Masset G, Monsivais P, Maillot M, Darmon N, Drewnowski A. Diet
Ethics approval and consent to participate optimization methods can help translate dietary guidelines into a cancer
This study was approved by the Universiti Kebangsaan Malaysia (National prevention food plan. J Nutr. 2009;139(8):1541–8.
University of Malaysia) Medical Research Ethics Committee (UKMREC) (NN- 18. WHO EC. Appropriate body-mass index for Asian populations and its
2017-128). All participants provided written consent. implications for policy and intervention strategies. Lancet (London,
England). 2004;363(9403):157.
Consent for publication 19. National Coordinating Committee on Food and Nutrition (NCCFN).
Not applicable Recommended nutrient intakes for Malaysia: Ministry of Health Malaysia; 2017.
20. World Cancer Research Fund / American Institute for Cancer Research.
Food, nutrition, physical activity, and the prevention of Cancer: a global
Competing interests
perspective. Washington DC: AICR; 2007.
The authors declare that they have no competing interests.
21. National Coordinating Committee on Food and Nutrition (NCCFN).
Malaysian dietary guidelines 2010: Ministry of Health Malaysia; 2010. dg.
Publisher’s Note sncoc.org/upload/affix/20140818104029708.pdf.
Springer Nature remains neutral with regard to jurisdictional claims in 22. Rao ZY, Wu XT, Liang BM, Wang MY, Hu W. Comparison of five equations
published maps and institutional affiliations. for estimating resting energy expenditure in Chinese young, normal weight
healthy adults. Eur J Pharm Med. 2012;17(1):26.
Author details 23. Institute for Public Health. National health and morbidity survey 2015. Non-
1
Dietetics Programme, Faculty of Health Sciences, Universiti Kebangsaan communicable diseases, risk factors & other health problems. Vol. II. Institute
Malaysia, 50300 Kuala Lumpur, Malaysia. 2Financial Mathematics Programme, for Public Health Report No. NMRR-14-1064-21877. 2015.
Faculty of Science and Technology, Universiti Sains Islam Malaysia, 71800 24. Mirnalini JK, Zalilah M, Safiah M, Tahir A, Siti MH, Siti DR, et al. Energy and
Nilai, Negeri Sembilan, Malaysia. nutrient intakes: findings from the Malaysian adult nutrition survey (MANS).
Malays J Nutr. 2008;14(1):1–24.
Published: 13 June 2019 25. Institute for Public Health. National health and morbidity survey 2014.
Malaysian adult nurition survey. Vol. I. Institute for Public Health Report No.
NMRR-12-815-13100. 2014. iku.moh.gov.my.
References 26. Rezali FW, Chin YS, Mohd Shariff Z, Yusof M, Nisak B, Sanker K, Woon FC.
1. Falk LW, Sobal J, Bisogni CA, Connors M, Devine CM. Managing healthy Evaluation of diet quality and its associated factors among adolescents in
eating: definitions, classifications, and strategies. Health Educ Behav. 2001; Kuala Lumpur, Malaysia. Nutr Res Pract. 2015;9(5):511–6.
28(4):425–39. 27. Hughes LA, Simons CC, van den Brandt PA, van Engeland M, Weijenberg
2. Maillot M, Drewnowski A, Vieux F, Darmon N. Quantifying the contribution MP. Lifestyle, diet, and colorectal cancer risk according to (epi) genetic
of foods with unfavourable nutrient profiles to nutritionally adequate diets. instability: current evidence and future directions of molecular pathological
Br J Nutr. 2011;105(8):1133–7. epidemiology. Current colorectal cancer reports. 2017;13(6):455–69.
3. DeSalvo KB, Olson R, Casavale KO. Dietary guidelines for Americans. JAMA. 28. Darmon N, Ferguson EL, Briend A. Impact of a cost constraint on
2016;315(5):457–8. nutritionally adequate food choices for French women: an analysis by linear
4. Ahmad N, Jaafar MS, Bakhash M, Rahim M. An overview on measurements programming. J Nutr Educ Behav. 2006;38(2):82–90.
of natural radioactivity in Malaysia. J Radiat Res Appl Sci. 2015;8(1):136–41. 29. Rajikan R, Zaidi NIA, Elias SM, Shahar S, Manaf ZA, Yusoff NAM. Construction
5. Azizah A, Nor Saleha I, Noor Hashimah A, Asmah Z, Mastulu W. Malaysian of healthy and palatable diet for low socioeconomic female adults using
National Cancer Registry Report 2007–2011: Malaysia Cancer statistic, data linear programming. Int J Adv Sci Eng Inf Technol. 2017;7(1):125–31.
and figure. Malaysia: National Cancer Institute; 2016. 30. Franceschi S, Bidoli E, Negri E, Zambon P, Talamini R, Ruol A, Parpinel M,
6. Ghazi HF, Hasan TN, Isa ZM, AbdalQader MA, Abdul-Majeed S. Nutrition and Levi F, Simonato L, La Vecchia C. Role of macronutrients, vitamins and
breast cancer risk: review of recent studies. Malaysian J Pub Health Med. minerals in the aetiology of squamous-cell carcinoma of the oesophagus.
2016;16(1):75–80. Int J Cancer. 2000;86(5):626–31.
7. Cuco G, Arija V, Marti-Henneberg C, Fernandez-Ballart J. Food and 31. Andriolo A. Diet and cancer. J Bras Patol Med Lab 2016; 52(6):366–366.
nutritional profile of high energy density consumers in an adult 32. D’Elia L, Rossi G, Ippolito R, Cappuccio FP, Strazzullo P. Habitual salt intake
Mediterranean population. Eur J Clin Nutr. 2001;55(3):192. and risk of gastric cancer: a meta-analysis of prospective studies. Clin Nutr.
8. Dachner N, Ricciuto L, Kirkpatrick SI, Tarasuk V. Food purchasing and food 2012;31(4):489–98.
insecurity: among low-income families in Toronto. Can J Diet Pract Res. 33. Asmaa A, Zzaman W, Tajul A. Effect of superheated steam cooking on fat
2010;71(3):e50–6. and fatty acid composition of chicken sausage. Int Food Res J. 2015;22(2):
9. Darmon N, Briend A, Drewnowski A. Energy-dense diets are associated with 598-605.
lower diet costs: a community study of French adults. Public Health Nutr.
2004;7(1):21–7.