Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
PRACTICAL CONSIDERATIONS IN THE EXECUTION
The duration of a study will vary with local circumstances. For example, the reactions of the community and the level of enthusiasm for co-operating in the project can never be precisely known beforehand and are likely to change as a result of many different circumstances. Local facilities and problems of transport can also affect the organization of the project. Therefore, estimation of the duration of the different types of studies can be only approximate.
Certain assumptions also have to be made about the initial planning of all of these projects, which means that, prior to the definite selection of a specific community, there would be contingency plans for communities in different parts of the world that would approximately suit the requirements for short-term, longer-term, and seasonal projects. If this information is not effectively organized beforehand, the duration of the projects of each type could be greatly extended. Prior decisions are needed on selecting a suitable general area and a specific community to fit the several essential conditions, such as the high probability of true existence of chronic energy deficiency and a population of the appropriate size to allow selection of an adequate sample. It is also desirable that a scientific institute or laboratory should be located nearby and be staffed by individuals who have local knowledge and experience. All require considerable time-consuming effort.
The time schedule for a short-term intervention study would include the following components.
1. Preparation-selection of the precise population to be studied; recruitment and training of local helpers; final decisions about which of the standardized methods is to be used; organization of data collection and handling; setting up of a small laboratory; etc.: 2-3 months.
2. Baseline data-collection of the relevant data to provide baseline information: 2-3 months.
3. Intervention and post-intervention intervals. It will obviously take some time before an intervention has any beneficial effect (if indeed it does have any), probably not less than five to six months. It will thus be necessary to wait that long after food supplements to individuals have been begun before making repeat measurements. If the intervention can be introduced to some individuals in the sample as soon as their baseline data have been collected, the investigators can begin collecting the repeat measurements five to six months later, but it will still take two to three months to complete them. In practice it may not be practical or desirable to stagger the introduction of the food supplements; they may need to be started for everyone at the same time. The total duration of this phase, therefore, is likely to be 7-9 months.
4. Analysis and report. Analysis of the data, interpretation of the results, and preparation of the report should be completed at least in a preliminary fashion in 3 months.
Total time: up to 18 months.
Longer- Term Study
For a longer-term intervention study, the time schedule would differ from that for a short-term study in the following respects.
1. Baseline data. The period for collection of the
baseline data will have to be longer because of the greater
number, diversity, and complexity of the measurements.
2. Intervention period. The duration of the intervention period before the final measurements are made should be extended to 12-18 months.
3. Analysis and report. Data analysis and reports will also take longer, perhaps 4-6 months.
Total time: 2' -3 years.
Study of Seasonal Variation
Studying the effects of seasonal variations in food availability or energy demands would require times similar to those for the other projects, the main difference being that the "intervention" is a natural one. Many areas with seasonal influences are subjected to them twice a year, so that a six month period would include measurements both before and after the fluctuation. A full year of measurements would be required, however, to follow these seasonal effects fully, which, together with time for planning and for analysis of the results, means that the study would entail about 18 months.
Nature of the Intervention
The purpose of providing a food supplement is to increase the energy intake of the chronically deficient population in a way that will be acceptable for several months as a true supplement and not simply a substitute. The supplement could be either more of the habitual diet of the community or a special high-energy preparation. On the whole, it is probably easier to ensure that a high-energy supplement reaches the appropriate individuals and to monitor that it is a supplement to the normal energy intake.
Organizing the distribution in a form that will ensure that the supplement is regularly consumed but does not otherwise affect the normal eating patterns may be difficult; the best approach will almost certainly vary considerably with the population. The frequency of consuming the supplement may vary with circumstances.
The amount of energy contained in the supplement must be sufficient to produce a measureable effect in a reasonably short period of time, and yet not so much that it will be consumed in very differing amounts by the individuals in the study.
The energy intake of a population that suffers from chronic energy deficiency and yet continues to be reasonably active is likely to be as much as 20 per cent below the desirable requirement. The energy content of the supplement should be slightly more than this, around 25 per cent of the requirement. For children the amount should be higher, about 30 to 35 per cent of the requirement, to allow the possibility of some catch-up growth.
Characteristics of the Supplement
To minimize the risk of the supplement's becoming a substitute, it should be energy-rich and therefore of relatively small volume. It should also be eaten at a time of day when it is least likely to have an effect on appetite. The supplement should of course be palatable and acceptable to the local community. The foods most likely to satisfy these stipulations would be carbohydrate-rich beverages and high-fat and sweet biscuits (or cookies). If, for compelling reasons, it is decided to provide extra food instead of a supplement, foods as energy-rich as are acceptable should be selected.
If it appears probable that the study population may be suffering from protein as well as energy deficiency, the supplement or the extra food should also be capable of providing additional protein. This is especially necessary in children to avoid any adverse effects of exaggerating the energy-protein imbalance.
For most types of studies, micronutrients need not be considered, although long-term projects in communities with possible micronutrient deficiencies should provide additions to the supplement.
Distribution and Consumption
The supplement can be given to the subjects on a daily or weekly basis with instructions to eat it at home, or it can be distributed and eaten daily under direct supervision of the experimenter. The latter is more complex and costly in a free-living population, but is sometimes possible.
The former approach must take into account the following considerations: (a) Enough supplement must be provided for the whole family, or for all children in the household, to avoid undesired intra-family distribution of the target individual's share. (b) It should be consumed at the end of a meal to minimize any effect it might have in reducing appetite. (c) A system should be devised to assess how much of the supplement has been eaten and whether or not it has had an influence on normal dietary intake. Id) The shelf-life of the supplement should be appropriate for the distribution intervals. (e) If possible, its re-sale value should be low, although if this is impractical it may not have much importance if the people in the study are sufficiently motivated and monitored.
Ethical problems may arise with respect to the control group. To use a group of persons clearly suffering from serious energy deficiency simply as guinea pigs for comparison with another group being given nutritional supplements would not be ethically acceptable. Such difficulty should not occur when an investigation is to be made on a community whose members apparently have adapted reasonably effectively to chronic energy deficiency and are still capable of working and partaking in some social activities. As some compensation for the inconvenience to the control group, however, and perhaps also as a stimulant for their continued co-operation in the project, some suitable means of rewarding them must be devised; simple presents and clothing are often quite acceptable.
If some individuals in the experimental group are found to have replaced some of their normal diet with the supplement, the supplement can be reduced gradually toward the end of the intervention. Otherwise at the end of the study the energy deficit could be greater than it was at the beginning.
If the intervention is shown to have a marked beneficial effect, considerable effort is necessary to induce the appropriate authorities and policy-makers to implement measures that would increase the food energy available to the population and to other similar populations in the country.
Nutritional, physiological, and socio-cultural measurements are required, first to define the baseline state and second to assess any changes that have occurred during the study. Some details of the methodologies used for these various measurements are described below, with suggestions of minimal and desirable ones to include. Table 1 lists the measurements needed for the different kinds of projects, distinguishing in the case of short-term studies those that are essential and those that are desirable.
Accurate data on the energy intake of the population are of obvious importance. At the theoretical level, it is apparent that we need to know the intake of any individual being studied before the intervention be gins and then how much extra food is being consumed by reason of the intervention, and we must be sure that the extra intake as a result of the supplement continues for the relevant period. Techniques are available to measure energy intake with adequate accuracy. Unfortunately, a major problem i. the considerable variability of responses both within and between individuals.
TABLE 1. Minimum and desirable measurements for short-term, longer-term, and seasonal studies
In well-fed populations, individuals can have a mean energy intake of 2,000 kcal per day and easily have a standard deviation of 300 to 400 kcal in day-today variability. Equally, similar persons leading more or less comparable life-styles can have as great inter individual differences. In populations suffering from chronic energy deficiency, this variability may not necessarily be as great, but it is still a possible complicating factor. Therefore, energy intake must be measured as carefully as possible on a suitable sample size.
The method most recommended is a combination of diet interview and weighing of a food inventory over a period of time appropriate to the purchase patterns and menu cycle of the population. It requires carefully trained and experienced field workers to obtain accurate and valid dietary-intake information.
Duration of Observation
The weighed-inventory method is usually carried out for an individual subject over several consecutive days, the duration ideally depending on the nature of the day-to-day diet and on the degree of co-operation of the person. Even if the local diet appears to vary only minimally (as might often be the case in this instance), a duration of fewer than three consecutive days is not recommended. For many populations, seven days represents a usual cycle of food purchases and meals.
Frequency of Assessment
Dietary data should be obtained during the baseline period and twice more during short-term studies or quarterly during long-term studies.
Even though seasonal variations may not appear to be large enough to exert any important influence in relation to energy deficiency or adaptation, they may be sufficient to cause significant dietary changes, and this should be taken into account in planning the project.
Energy intake is usually calculated from the weights of the foods eaten, using energy values obtained from food-composition tables. Sometimes the relevant tables do not have precise data on the energy values of certain foods in the local diet. In this case, it is best to try to estimate, from other comparable foods, an appropriate value that can be used with relative accuracy. It is possible to obtain a gross energy value by bomb calorimetry or by proximate analysis, and then to apply a factor to allow for energy losses in faeces and urine. Unless the food item is of very considerable importance, however, the time and expense involved are not justifiable.
Anthropometry and Body Composition
Height ( or length in infants) should be measured under standardized conditions and at a fairly constant time of day. Adults should have the baseline measurement repeated at some time during the project to check for consistency. Children should be measured monthly.
Weight should also be taken in reasonably standardized conditions, although it is pointless carrying this to inconvenient extremes because of the effect on weight of normal fluctuations in body water. In adults this can easily be up to 1 kg from one day to the next. Weight should be taken for all individuals twice during the baseline period; after the intervention begins, it should be taken monthly for adults and every two weeks for children and pregnant women.
Skinfold thickness provides the best field information about the level of fatness of an individual. Because it would be anticipated that persons suffering from chronic energy deficiency would have minimal quantities of fat in the body, and therefore very small amounts of subcutaneous fat, these measurements are among the most important.
In general, the more skinfold measurements that are taken, the lower the error in assessing the fat content of the body. Frequently four such measurements -biceps, triceps, subscapular, and supra-iliac -are taken. If it is practical in terms of time and convenience to take all four measurements, it is worth doing so. However, very little extra error will ordinarily arise if only two measurements are taken - preferably the triceps and the subscapular, or the triceps and the supra-iliac. On the other hand, in populations with unusual physiques, it may even be useful to take more than four measurements.
Circumferences may act as useful indicators of muscularity, especially as any population that is genuinely suffering from chronic energy deficiency will have minimal amounts of subcutaneous fat. At minimum, the sites are the mid-upper arm and the largest area of the calf. The upper thigh and greatest buttock circumferences are also potentially of use and should be obtainable in most populations.
Skeletal diameters are also indicators of the different types of development that may characterize certain populations. They are slightly more difficult to measure accurately than other anthropometric variables, but with experience they pose few problems and take only a few seconds to do. The most useful are the bi-acromial and bi-iliac diameters. They do not need to be repeated, as they simply help to identify a population and should not alter during the project.
Other measurements: No other measurements are recommended. Those that might be performed are, in general, too complex or too difficult for routine field use. Nevertheless, body density, total body water, perhaps electrical impedance, and ultra-sound may on occasion be possible and desirable.
Basal Metabolic Rate
The basal metabolic rate is one of the commonly affected biological variables in some types of chronic energy deficiency and should always be measured when low energy intake is discovered unless circumstances are very difficult. Even in the baseline state, it can indicate possible energy deficiency together with some degree of adaptation. Because the BMR can be an important adaptation to reduced energy requirements, and because this adaptation may be altered rapidly by the consumption of extra food, it is highly desirable to measure it both during the baseline period and after intervention. After the intervention begins, measurements should be made at two and three-month intervals, each time on two different days within a day or two of each other to allow for day-to-day variability.
A simple exercise step-test can provide useful functional information. It is important to remember that static measurements do not necessarily indicate a functional change. For example, increases in body weight and body fat might not by themselves necessarily coincide with increased capacity. An exercise test can demonstrate whether or not there has been any change in the ability of the individual to do some standardized physical activity with regard to the stress to the cardiovascular and respiratory systems. If increased capacity for exercise is a consequence of the nutritional supplements, it can be very significant and practical benefit to the person.
There seems to be a strong impression in the minds of many nutritionists that an exercise test in the field, especially on a population possibly suffering from energy deficiency, is either too artificial or too complicated. It is neither. A very simple step-test can be performed on quite unsophisticated populations, and, with relatively simple equipment, heart rate and oxygen consumption can be determined at three different levels of exercise and the maximum exercise capacity of individuals be predicted with reasonable accuracy. This information can be used to categorize individuals into different levels of fitness.
The measurement is primarily of cardio-vascular and respiratory fitness, but as these are the aspects of fitness most closely correlated with the general ability of the individual to move around, work, engage in active leisure, and so on, they are also the most helpful to measure in the general assessment of physical health. Thus a simple exercise test is highly desirable and should be seriously considered for inclusion in anything other than a very restricted project.
The victims of chronic energy deficiency that is not flagrant or severe may adapt in several ways. One obvious way is to reduce the level of physical activity. If such persons' food is supplemented, it can reasonably be assumed that either more active leisure might become apparent or work output (not necessarily only in economic terms) might increase, or both might occur.
Measurement of the type and duration of all physical activity is therefore of basic importance for two reasons: (1) to set the baseline from which deductions can be made as to whether or not activity is likely to be reduced, and (2) to determine whether the increased energy intake from dietary supplements results in higher levels of activity. Information must be obtained on the duration and strenuousness of activity both at work and in leisure.
Physical activity can be measured by several methods, one of the simplest being a time-motion study carried out by the observation of selected individuals. It may be most informative to obtain some general picture of, say, the average adult male, non pregnant and pregnant women, and children. Such a picture could be built up by observing key individuals during selected periods to cover typical work and leisure situations. A more comprehensive account at the community level can be obtained by the time sampling technique, during perhaps a two-week period in the baseline phase and at two- and three month intervals after the start of the intervention.
A different approach to assessing changes in physical activity is to measure work output, if the work is capable of being assessed in this way. An increase in work output is, of course, not necessarily the only indicator of increased physical activity, but it is certainly one possible outcome of an improvement in nutritional status as a result of dietary supplementation. If the amount of work performed by individuals in the study is such that it can be quantified, this should be done in the baseline period and at two and three-month intervals after the start of the intervention.
A form of quantitative assessment that can be integrated with time-motion data is obtainable by heart-rate recording. Numerous instruments can monitor heart rate over variable periods of time from a few minutes up to several hours. If such instruments are available and can be used ( sometimes humid tropical environments make their use almost impossible), they may detect more energy-intensive work or leisure activities. Examples include an increased effort expended in work over the same duration as previously, and an enhanced total level of activity during certain periods of the day. Heart-rate recording integrated with time-motion studies may usefully aid in the interpretation of social information on activity patterns.
It may be useful, though probably only in a restricted way, to make some direct measurements of energy expenditure. It may be interesting to compare the total energy expended at work in the baseline period and after intervention. Such a comparison can be made in an approximate fashion using the time motion record and the energy values for various work activities obtained from published tables; that is, no actual measurements of energy expenditure need be made.
The actual energy expended in certain activities, either at work or in leisure, can be measured directly and compared in an individual. It might be shown, for example, that after intervention more energy was expended in specific work tasks. It can also be calculated from heart-rate data if calibration of the heart rate with oxygen consumption is also obtained.
It is difficult to measure school performance accurately as an independent variable because it is open to many indirect influences. In studies involving schoolchildren, however, scholastic performance, social interactions, and activity patterns during school hours should be assessed in the baseline period and near the end of the intervention. Absenteeism from school and the reasons for it-illness, the need to work or to look after other children in the family- should also be recorded. School performance might easily reflect chronic nutritional deficiencies, but it is improbable that it varies due to seasonal influences, and so it is usually not worth assessing in seasonal projects.
OTHER ASPECTS OF DESIGN
Statistical advice is required to determine the size of the population sample. As most of the data analysis depends on fairly straightforward tests of significant differences between mean results obtained on two or three occasions (e.g., the mean of two baseline values compared to one result two to three months and to another five to six months after the start of the intervention), the number of individuals to be studied need not be very large. It is necessary to remember, however, that individual variability is so great that there may well be very little correlation between energy intake and several different aspects of activity; an adequate sample size is required to compensate for this variability.
It is also important for these individuals to be representative of a reasonably large group; so it will be helpful if the population to be studied is as homogeneous as possible. Ideally, the population should inhabit a village, or a district in a town, where differences between households appear to be comparatively minor. It is critically important that great care should be taken and as much precise information as possible be acquired before a community is selected for the project.
The control group must be chosen with the same care as the experimental group. In general, it is undesirable to use a control group from the same village or district.
To obtain data for continuing analysis and to allow efficient collection of information, the use of small local computers should be encouraged.
For proper comparison of data from different study centres, the methodology should be standardized as far as possible. To ensure this standardization, consideration might be given to having the senior field workers spend a few weeks together in the same laboratory before any project starts.
EXAMPLES OF CONSIDERATIONS FOR SPECIFIC DESIGNS
The information given above should provide basic assistance for planning and carrying out an intervention study on a population suffering from chronic energy deficiency. The brief sections that follow give examples of factors to be considered in designing specific projects that might be appropriate for a village in Africa and a district in a large city (e.g., Lima, Peru).
An African Village
It has been mentioned several times that it is of extreme importance to select an environment suitable for the purpose. It is assumed that colleagues working in the particular country and preferably with local knowledge of the region will be collaborating. Therefore, the village chosen should be, from prior direct information, one where chronic energy deficiency is likely to affect most of the population. The population should also be large enough for the purpose. The senior investigator should always confirm this personally by visiting the village and spending sufficient time there to verify its suitability. Prior to this visit, adequate official approval of the project down to the local level should have been obtained to gain initial co-operation from the community.
During the visit, time must be taken to meet the local notables on friendly terms, to explain enough of the purpose of the project to make it acceptable without necessarily giving information that may influence the outcome, and to arrange living accommodations and a suitable laboratory for the investigators.
Depending on the time and money available, the extent and number of measurements to be made should be decided in a preliminary way. Whether or not the project will be short-term, longer-term, or seasonal will presumably already have been determined.
When the field work starts after the arrival of the investigators, the first subjects should be chosen very carefully after interviewing a range of volunteers, as successful initiation of the study will greatly facilitate later work. Serious problems in this early phase might jeopardize the success of the whole venture.
Local constraints also have to be taken into account. For example, many regions of Africa have populations that are Muslim, and attention must be paid to avoiding actions that might be unacceptable on religious grounds. Marked seasonal weather effects, even if they have little nutritional or socio-cultural impact, might influence the time of year when the study should be started; a very wet rainy season with flooding and impassable roads, for example, would probably not be the best time to begin the project.
A suitable water supply and electricity (perhaps requiring a generator) will be needed.
Almost all of these requirements apply to village populations in other parts of the world, with the exception that the religious problems would be different.
A District in a Large City
There are advantages and disadvantages in a city compared to a village. The study is easier for the investigators with respect to living conditions and some of the technology. They can live in some more convenient area of the city, and usually a laboratory (a rented house, perhaps) with running water and electricity can easily be obtained. It is unlikely that the population will be as homogeneous as in a village, however, and it may be much more difficult to control all the variables. Monitoring physical activity and energy intake may be especially complicated. In this type of environment it is even more important to select the participating individuals carefully. Even if money and facilities are available to undertake a longer-term study, it is preferable to do a short-term pilot study first so that all the practical problems may be uncovered and tackled.
The urban poor, many of whom suffer from chronic energy deficiency, may well represent the greater part of the world-wide problem of nutritional deprivation. They must be studied in the present context, but it will undoubtedly require greater and more complex effort to obtain good scientific data on them than on most village populations.
These studies could yield several results. The simplest would be a measurable improvement in several variables in the experimental population, such as an increased rate of growth in children, increased body weight and to a limited extent fatness in adults, more active leisure pursuits, and greater work output -whereas there would be no significant alteration in any of these variables in the control group. Longer term consequences might be increased energy put into economic activities, home improvement, household food production, community development, and political organization.
Values of the BMR might or might not change, but the importance of the measurement is mainly to indicate adaptation to low energy intakes. There might also be an improvement in results of the exercise tests.
Such clear-cut results are unlikely, especially in short-term studies. A change in some variables but not in others may cause much difficulty in interpretation and in recommending subsequent action. For example, if the only positive results are a very slight increase in the rate of growth of children and an addition to the body weight of adults that proves to be mostly fat, with no changes in physical activity, work output, morbidity, and so on, it becomes somewhat of a dilemma to decide whether they are of practical significance.
One of the biggest problems in effecting a clear-cut outcome in an intervention study is in ensuring that the distribution of food supplements really results in a significant increase in food intake that continues for a sufficiently long time. It is probably true that, of all the thousands of times that food supplements have been distributed to populations supposedly in need of them (but not so severely as to be facing nearstarvation), very few indeed have resulted in an important increase in food intake. Supervising and monitoring the food-supplementation programme is essential, and, unless this is successfully accomplished, implementing the rest of the project is more or less useless.
If the supplementation is adequate, there should certainly be some detectable change -for instance, in body weight or growth. If, however, there is not also some functional improvement, it will be difficult to generate much enthusiasm for a programme of further intervention. If the programme is wide-ranging, its expense may well be almost prohibitive and justifiable only if it can be shown that it will have results of real nutritional, social, or economic value.
A change in functional values-such as work output, social interactions, and leisure pursuit-together with some physical improvement for the population would be a sufficiently persuasive indication that the intervention would be of significant relevance to the well-being of the population to induce further action.
If the measurement of all the variables shows only partial or negative results, the outcome is likely to be disastrous, assuming that the population investigated is genuinely affected by chronic energy deficiency.
An energy deficiency, almost by definition, implies that anything that lessens the effect of that deficiency can be shown to have beneficial effects. If, therefore, we have an intervention in a population suffering from true chronic energy deficiency and we can not clearly show a beneficial result, the only reasonable explanation is inaccurate and inadequate methodology.
This topic is of enormous and world-wide importance, and there are no methodological shortcuts. Whether a short-term or a longer-term project is· decided on will depend on finances, technical assistance, local constraints, and so on, but there is little room for compromise on the measurements made. Small economies are possible: two skinfold measurements instead of four or more, fewer skeletal dimensions, and perhaps no measurements of energy expenditure. Basically, however, all the measurements mentioned are necessary in order to demonstrate clearly (a) that chronic energy deficiency exists, (b) that food supplementation has indeed taken place and has continued, (c) that some anthropometric changes have or have not occurred, and (d) that an improvement in some or all areas connected with activity, fitness, work output, and social interactions has been demonstrated.
Contents - Previous - Next