The content of African diets is adequate to achieve optimal efficacy with fixed-dose artemether-lumefantrine: a review of the evidence

Lumefantrine, like several other lipophilic antimalarial agents, is absorbed slowly after oral administration, and bioavailability varies between individuals [2, 11‐13]. Concomitant food increases lumefantrine bioavailability [14]. In a double-blind trial of 260 patients with uncomplicated malaria in Thailand, Ezzet et al showed that both the extent and variability of lumefantrine absorption improved with clinical recovery as normal food intake was resumed [14]. Oral bioavailability increased by 108% when a normal meal was eaten close to AL dosing, compared to the fasting condition. Adequate lumefantrine absorption is important in order to avoid low plasma concentrations with risk of treatment failure [2].

The key question for the healthcare team is how much dietary fat is necessary to achieve complete parasite clearance with AL. This question has been investigated by Ashley et al in a cross-over pharmacokinetic study in which lumefantrine exposure (area under the curve, AUC) was measured in healthy volunteers after a single dose of AL was administered with different volumes of soya milk or with no milk [15]. Based on these data, the authors constructed a population model which indicated that in this population of healthy volunteers, drinking 36 mL of soya milk (containing only 1.2 g of fat) resulted in 90% of the lumefantrine exposure obtained with 500 mL milk (16 g fat); 74 ml of soya milk (2.4 g of fat) achieved 99% of the exposure seen with 500 mL (Figure 1). Ashley and colleagues pointed out that 1.2 g of fat can be obtained in a remarkably small amount of food: for example, a single teaspoon of vegetable oil contains 5 g of fat.

Diets in Africa vary widely according to regional, economic, environmental and cultural factors. Nevertheless, there are certain typical patterns of food consumption (Table 1). African diets are traditionally based on a carbohydrate staple that is supplemented by soups, relishes and sauces derived from a wide range of other foodstuffs. The carbohydrate component may be primarily cereals (e.g. maize, sorghum, millet, rice), starchy roots (e.g. potatoes, sweet potatoes, yams, cassava) or fruit (plantain), depending on the country and local area. In recent years there has been a trend away from the consumption of roots towards maize, rice and wheat, particularly in cities [16]. Accompaniments to the carbohydrate staple are usually composed of vegetables (e.g. eggplant, cabbage, carrots, spinach, kales, French beans, onions), pulses, nuts (particularly peanuts), meat (chicken, beef or pork) or fish, depending on local availability, household income, season and community habits, with the greatest variety seen in coastal regions and the fertile highlands [16].

The Food and Agriculture Organization (FAO) of the United Nations compiles data on various aspects of food production and availability [17]. These statistics offer a valuable source of information, although it should be borne in mind that estimates are based on the amount of food available for human consumption, rather than actual amounts eaten, and do not account for wastage. Table 2 summarizes the consumption of selected food groups in sub-Saharan countries based on FAO data, and highlight the predominance of cereals and starchy root vegetables [18].

The most important sources of dietary fat in African countries are oil crops (e.g. peanuts, soy beans) and oils obtained from vegetables or plants such as red palm oil, groundnut oil, coconut oil and sesame oil. These are widely used in food preparation, for frying or as additions to sauces and stews. Additional contributions to fat intake are made by other commonly-used foodstuffs such as nuts and pulses. Table 3 describes the typical fat content of some staple foodstuffs [19]. While food composition inevitably varies according to crop variety, growing conditions and freshness, these illustrate that relatively small amounts of certain widely-consumed items such as soya beans and ground nuts contain high levels of fat. A further contribution comes from whole-grain cereals if the cereal germ is not separated out prior to milling: for example, maize contains 4.8 g/100 g [19] and is a staple food in many countries.

Total energy intake per person across sub-Saharan Africa is currently estimated to be 2220 kcal/day, compared to 3340 kcal/day in developed countries [10]. Regional differences exist, most notably a higher average energy intake in West Africa compared to other areas (Table 4). Estimates of total fat intake from the FAO for the period 2001–2003 suggest that the average total amount of fat consumed per person per day vary considerably between subSaharan African countries (Table 5) [10]. The majority of countries, however, fall into the range 30–60 g/person/day. The daily intake of fat across subSaharan Africa is estimated to be 43 g/person/day, with the highest fat intake in West Africa (Table 1); for comparison, people in developed countries consume 123 g/person/day, on average [10]. The per capita daily intake of fat has generally remained stable or shown an upward trend in the majority of African countries over the last three decades [10], with exceptions being due to major conflicts or famines.

It is standard to breast-feed infants up to 4–6 months in all regions of Africa, although supplementary cow's milk may be given. The fat content of breast milk in rural African areas varies with the season, stage of lactation and between individual mothers, but an investigation in a farming community in The Gambia recorded an average fat content of 40 g/L breast milk [20], similar to that reported in other parts of the world [21‐23]. Breast-fed infants in The Gambia have been shown to receive 15–30 g fat/day from breast milk up to the age of 18 months, with breast milk continuing to make a major contribution to fat intake even after partial weaning [24]. Comparing infants in The Gambia and the United Kingdom, Prentice et al found that fat intake from breast milk was at least as high in the Gambian children while being breast-fed exclusively [24].

Infants usually continue to receive breast milk until two years of age, with some food introduced from three months or later. The first weaning food is usually porridge made from a cereal such as rice, corn or millet mixed with water, or possibly cow's milk, with salt and sugar. Such porridges have a relatively low fat content (e.g. 0.1–0.3 g/100 g); although addition of soya milk or peanuts in some regions increases the fat content (1.2–1.9 g/100 g) [25]. In West Africa there tends to be a greater diversity of early foodstuffs, including chocolate, mashed potatoes or beans (Table 1).

The transition from breast milk to weaning foods and then to adult foods is associated with a marked reduction in the proportion of energy that is obtained from fat. A study in Kenya was undertaken in infants during the second year of life (mean 14–20 months) to assess fat intake over the period of weaning [26]. In these young children, the mean quantity of breast milk ingested was calculated to be 502 ± 239 g per 24 hours, although this varied widely between individuals [26]. The authors calculated the average fat intake from milk to be approximately 17 g/day, which was supplemented by 10.5 g/day from weaning foods. In the period prior to weaning, fat intake from breast milk decreased to approximately 11 g/day but after complete weaning the increased food consumption provided only an additional 3.5 g/day. A further detailed study in The Gambia has shown that total fat intake remained relatively stable over the first 17 months of life, but as the percentage of other foods (notably cereals and peanuts) increased in the diet, the percentage of energy from fat decreased from 50% during breast-feeding to 30% during mixed feeding and to 15% after total weaning [24]. Nevertheless, total fat intake, which was ~25 g/day until complete weaning, was >10 g/day subsequently.

Cow's milk is only given to infants infrequently in many African countries because of the low milk yield per animal and scarcity of preservation equipment, although it is more common in some countries (e.g. Kenya, Somalia and Sudan). Where cow's milk is used, however, it is a good source of fat with a fat content of 30 g/L [19], only slightly lower than breast milk.

In most trials of AL in African populations, AL has been administered under supervision with food or drink given in the hospital or clinic [4, 7, 27‐29]. Two studies, however, have assessed 'real-life effectiveness' of AL, in which the drug was taken unsupervised at home [6, 30]. These are helpful in providing an indication of whether normal, unsupervised food consumption during AL administration ensures good efficacy. Piola et al undertook a randomized trial in 957 patients in Uganda to compare the effectiveness of AL when administered under supervision in hospital with fat intake ensured by a meal comprising ~300 mL milk (10 g fat) with 30 g peanuts (13 g fat) versus AL taken unsupervised at home [6]. In the unsupervised group, only the first dose was taken in the clinic, and patients were then discharged and advised to take subsequent doses with a meal that contained fat (with no fat content specified), or breast milk in breastfed infants. The 28-day PCR-corrected cure rate in the evaluable population was 100% in both cohorts (95% CI 98–100 in the supervised group and 99–100 in the unsupervised group). The type of meal eaten at home was not recorded, so it cannot be determined if patients ate especially fatty meals or standard food. Lumefantrine plasma concentration was measured on days 3 and 7 in a random selection of 433 patients taking part in the study [31], and multivariate modelling of these data confirmed that supervised treatment was significantly associated with higher lumefantrine concentrations (p < 0.0001) [31]. In another trial, undertaken in an area of Tanzania with a high level of resistance to sulfadoxine-pyrimethamine and amodiaquine, Mutabingwa and colleagues randomized 1292 children aged 4–59 months with uncomplicated malaria to AL or to conventional drugs [30]. All agents were taken at home, without supervision, and no specific advice was given regarding concomitant food consumption. At 28 days, 97.3% of patients receiving AL showed parasitological cure after adjustment for recrudescence.

Together, these findings suggest that food consumption during unsupervised AL dosing in African patients, including children, provides sufficient fat to ensure effective parasitological clearance.

As discussed above, the fat consumption of infants shortly after weaning can be relatively low. It is highly relevant, therefore, that randomized trials in African children aged from five months [32, 33] or 12 months [34] up to 59 months have consistently shown 28-day PCR-corrected parasitological cure rates of greater than 95% i.e. similar to cure rates observed in older populations.

It should be noted that data are scarce concerning lumefantrine absorption, and the effect of food consumption, in patients with HIV/AIDs or other coinfections such as tuberculosis. Pharmacokinetic and pharmacodynamic studies are required in the HIV/AIDS population, particularly in the light of increased antimalarial failure rates in such patients due to increased parasite burden and reduced host immunity associated with HIV infection (1).