Raw data that are generated in surveys are of little use unless they are packaged and presented as information that can be understood and used by planners and programme-managers. Choosing the appropriate indicators and summary statistics to capture the information contained in the data is vital in this process. It is these issues that are discussed in Chapter 5.
Based on the Caribbean experiences in the conduct of other surveys/projects, data analysis has received little attention at the planning stages of the project. This has resulted in inappropriate analysis and restricted use of the information for policy development and programme-planning. It is vital, therefore, that all the steps of data analyses are thought out, planned and budgeted for during the planning stages of the survey
Integral to the plans are the individuals who perform the tasks - statistician, software specialist, food consumption specialist, and data entry clerks. Appropriate food composition tables and computer analysis software are also critical elements in transforming food consumption and anthropometric data into information on nutrient intake that can be used in programme and policy development.
food composition tables (FCT) are databases containing a list of foods with their respective amounts of energy, nutrients and other substances (usually expressed per 100 grams of edible portion). Most methods for evaluating food consumption use FCT to estimate the energy and nutrient content of each food reported by subjects. The dietary variables contained in the FCT should provide information about foods and nutrients that are relevant to the objectives of the particular
information survey. Survey personnel must also know if each food item in the FCT is raw or cooked, and if cooked, the method of preparation. This is crucial as incorrect classification can lead to under- and over-estimations of nutrients.[4]
COMPUTER APPLICATIONS FOR ANALYSIS OF FOOD CONSUMPTION DATA
Computer software for nutrient calculation should be selected according to the objectives of the survey and the availability of local resources such as trained personnel and appropriate hardware. These latter resources are very important in the Caribbean context where there is a shortage of trained personnel and equipment. There are several types of computer software available for data entry and analysis of food consumption data. Choosing the appropriate software requires careful consideration. This is because food consumption studies often contain different types of data for analyses and as a result require different types of software.
On the one hand, non-nutrient data from the questionnaires - such as demographic details, household-purchasing practices, perceptions of consumption of various food items, food-preparation practices, risk factors associated with health and nutrition, food practices, etc. - are usually entered using one of the general statistical applications widely available, for example, SPSS, SAS or STATA.
On the other hand, another type of software is needed for analysis of nutrient or dietary variables. For example, energy, protein, Vitamin C, iron, foods and their amounts consumed by the individuals must be entered in software that is capable of processing this type of data. Extensive information about computer applications for nutrient computation and food consumption data analysis can be found at The Nutrient Databank Directory1-9 and at the Food and Nutrition Information Center (FNIC) of the United State's Department of Agriculture's National Agricultural Library. Box 5.1 describes the software CERES, developed jointly by FAO and Institute of Nutrition and Food Hygiene of Cuba.
DATA, INDICATORS AND INFORMATION FOR PLANNING
Trends in dietary habits are influenced profoundly by the current expansion of global food trade. Whether this will bring positive or negative nutritional outcomes will depend on whether the relevant sectors within countries will be able to utilize opportunities to exploit the liberalization process. Governments and planners must therefore be aware of groups at risk of deprivation, such as vulnerable groups in the population, so as to implement and appropriately manage the country's social security and nutrition programmes. (Boxes 5.2 and 5.3 provide examples of how data collected from food consumption and anthropometric surveys were used in this regard for Barbados and Guyana.)
Information garnered from food consumption surveys are essential for generating information about a country's food and nutrition situation. By using the data collected in a survey, various indicators may be derived. Selection of indicators is usually based on the prevailing nutrition problems in the country. In addition, data can be used in order to present a picture of the population at risk of a particular threat related to nutrition.
In deciding how to transform the data into information, characteristics and current trends in food acquisition, availability, and nutritional status may be considered. In the Caribbean over the past few decades, there has been a gradual change in the prevailing nutritional problems - from undernutrition affecting mainly young children to overweight which is prevalent among both the young and the old. Associated with the increased prevalence of overweight there has been an increase in the chronic nutrition-related diseases such as: hypertension, diabetes, heart disease and some cancers. Concurrently, there has been a large increase in the amount of food available from animals, fats, oils and sugars and a smaller increase in roots and tubers, cereals, fruits, vegetables, legumes and nuts.
|
Box 5.1 CERES SOFTWARE The computer application, CERES, developed by the Institute of Nutrition and Food Hygiene of Cuba and the FAO has a fundamental objective of evaluating and processing data from food consumption surveys. The development of the application is based on the experiences gained while using other applications and trying to overcome the limitations and constraints identified. Up to this point, training courses in the use of CERES have been conducted in Cuba, Mexico, Colombia, Venezuela, Peru, Panama, Barbados, Jamaica, Guyana and Holland. General description CERES performs calculations for various kinds of food consumption surveys: individual, family, and individual-family level surveys by 24-hour recall, daily record, and direct weighing of foods, semi-quantitative food frequency at the individual level and food access and family consumption surveys. There is also an option for the estimation of nutrient contents and price of a food basket. The software offers maximum flexibility because it uses several kinds of reference values (i.e. food composition tables and recommended daily allowances) and the structure, components and dietary variables included as reference values are freely defined by the user. Data entry is performed using all the facilities offered by a Windows environment, including searching for food items and household measuring units. |
There are direct links between these trends in poor health and food intake. The consumption of too many fats, especially saturated fats, has been linked to an increased risk of obesity and associated disorders such as cardiovascular disease, hypertension, diabetes and some cancers. Recent evidence has also emerged of the contribution of trans fats to disease.10, 11 Dietary cholesterol contributes to elevation of serum cholesterol, a risk factor for many major chronic diseases,11 while high sodium intake is associated with elevated blood pressure and increased risk of stroke in some populations.11 On the other hand, fruits and vegetables contribute a mix of nutrients and bioactive substances such as fibre and phytochemicals to the diet and help to maintain good health and reduce the risk of cancer and heart disease.11
Indicators relating to population nutrient intake goals, dietary guidelines (where they exist) and recommended dietary allowances (RDA)[5] can all be used to identify risks related to diseases resulting from dietary deficiencies or excesses. These indicators may be represented as ranges, or may have only either an upper or a lower limit.
Based on the nutritional profile of the Caribbean, the following is a list of suggested indicators for consideration when planning and analysing data for food consumption surveys in the region.
FOOD CONSUMPTION INDICATORS
1. Percentage of persons consuming >30 percent of total energy as fats.
2. Percentage of persons consuming >15 percent of total energy as proteins.
3. Percentage of persons consuming >75 percent of total energy as carbohydrates.
4. Percentage of persons consuming >300 mg/day of cholesterol.
5. Percentage of persons consuming >5 g/day of sodium.
6. Percentage of persons consuming <400 g of fruits and vegetables.
7. Percentage adequacy of energy, protein, fats, carbohydrates.
8. Percentage intake of dietary iron among women of child-bearing age.
9. Percentage contribution of food items to selected dietary variables (i.e. what are the main food sources of energy and other nutrients).
The purpose of food consumption surveys is to compare energy and nutrient intakes with the estimated requirements of the population. Energy in the diet is derived from carbohydrates, fats and proteins. There is no single "best value" for population nutrient intake goals[6] but based on the relationship between diet and good health, CFNI recommends that contributions of these macronutrients to energy should range from 15 to 30 percent from fats, 55 to 75 percent from carbohydrates and 10 to 15 percent of energy from proteins.
In addition to protein, fats and carbohydrates, there are many other dietary components of importance to health. Over-consumption of some elements can contribute to poor health. For instance, there is a great deal of evidence associating cholesterol with increased incidence of cardiovascular disease. Since cholesterol is synthesized in the body, there is no requirement for dietary cholesterol. Hence, it is advisable to keep dietary cholesterol to low levels, less than 300 mg/day. Sodium intake is directly associated with increased blood pressure. Dietary iron is of importance because of its association with anaemia, which is especially common in pregnant women.
Food consumption surveys can also yield information on consumption of particular foods by the population, such as fruits and vegetables, which have many dietary components important to good health. Information on how particular foods and food groups contribute to the diet is also important for planning. The number of different foods or food groups consumed by a household is a measure of the dietary diversity.
ANTHROPOMETRIC INDICATORS OF UNDERNUTRITION[7]
1. Percentage of young children (0-5 years) who are stunted (height-for-age <-2SD of reference population).
2. Percentage of young children (0-5 years) who are wasted (weight-for-height <-2SD of reference population).
3. Percentage of young children (0-5 years) who are underweight (weight-for-age <-2SD of reference population).
4. Percentage of older children (10-18 years) who are underweight (BMI for age <5th percentile).
5. Percentage of adults who are underweight (BMI <18.5).
Body Mass Index (BMI) is calculated by dividing the weight of a person (kg) by the square of his or her height(m) and is used mainly for adolescents and adults. Foradolescents, BMI for age is plotted on gender-specific growth charts.13, 14 BMI for age less than 5th percentile is indicative of underweight in adolescents.13 In adults, the established normal range of BMI is not gender specific and ranges from 18.5 to 24.9 kg/m2.14 BMI has a drawback in that it does not differentiate between muscle mass and fat mass, and body shape and composition are not taken into consideration. However, generally, a BMI lower than 18.5 is considered to be an indicator of energy deficiency.
|
Table 5.1 World Health Organization
(WHO) |
||
|
Classification |
BMI |
Risk of co-morbidities[8] |
|
Underweight |
< 18.5 |
Low |
|
Normal range |
18.5-24.9 |
Average |
|
Overweight |
> 25.0 |
|
|
Pre-obese |
25.0-29.9 |
Increased |
|
Obese Class I |
30.0-34.9 |
Moderate |
|
Obese Class II |
35.0-39.9 |
Severe |
|
Obese Class III |
> 40.0 |
Very severe |
ANTHROPOMETRIC INDICATORS OF OVERWEIGHT AND OBESITY
1. Percentage of children (0-5 years) who are overweight (weight-for-height >2SD of reference population).
2. Percentage of older children (10-18 years) who are at risk for overweight (BMI for age 85th-95th percentile).
3. Percentage of older children (10-18 years) who are overweight (BMI for age > 95th percentile).
4. Percentage of adults who are overweight (BMI 25.0-29.9).
5. Percentage of adults who are obese (BMI ³ 30).
6. Percentage of adults with high waist to hip ratio (men: >1 and women: > 0.85).
7. Percentage of adults with high abdominal fat (waist circumference for men ³ 102 cm and for women ³ 88 cm).
Overweight and obesity, which refer to an excess of body weight compared to set standards, are associated with an increased occurrence of many health problems, including cardiovascular disease, hypertension and stroke, gallbladder disease, diabetes and certain forms of cancer.11, 14 There is an increased risk of premature death due to these and other serious chronic conditions that also reduce the overall quality of life. Of especial concern is the increasing incidence of child obesity.
In young children, weight-for-height more than 2 SD above the reference population is used to define obesity, while in older children BMI charts are used.13, 14 Adolescents are seen as at risk of overweight if BMI for age is between the 85th and 95th percentile and overweight if BMI for age is greater than 95th percentile.15
For adults, the classification of obesity using BMI is based on the association between BMI and mortality. A BMI of 25 or greater is categorized as overweight and a BMI of 30 or greater as obesity, with further subcategories (see Table 5.1).11, 14 Care must, however, be taken in the interpretation of BMI measurements, due to ethnic differences between populations, differences between the body composition of the sexes, and differences in the pattern of physical activity which results in variances in the adiposity of the body.
Waist circumference and waist-hip ratio are other valuable indicators used to identify people at increased risk of being affected by the co-morbidities of obesity due to accumulation of fat in the abdominal region.
OTHER RELEVANT INDICATORS ASSOCIATED WITH FOOD CONSUMPTION AND ANTHROPOMETRIC DATA
1. Percentage of pregnant women with anaemia (Hb < 11 g/dL).
2. Percentage of mothers with babies 0-6 months who are breastfeeding exclusively.
3. Percentage of adults with hypertension.
4. Percentage of adults with diabetes.
Based on the purpose of the survey, other data may be collected. It is therefore critical that data on demographic and socio-economic characteristics are also collected and integrated into the analyses of the data so as to provide useful guidance to policy makers and programme managers. For example, iron-deficiency anaemia is the most common micronutrient deficiency worldwide. Data from surveys and health centres in the Caribbean show that the problem is also widespread in this region, and affects many pregnant women as well as young children. Pregnant women with a haemoglobin (Hb) level less than 11 g/dL are regarded as anemic by WHO standards.16, 18 Because of differences in ethnicity, a cut-off of 10.0 g/dL is sometimes used in populations of African origin.18
|
Box 5.2 USE OF INDICATORS TO IDENTIFY GROUPS AT RISK IN BARBADOS The main purpose of the Food consumption and Anthropometric Survey in Barbados was to determine the food intake and dietary patterns of the adult population (18 years and over), the prevalence of obesity and diagnosed nutrition-related non-communicable chronic diseases, and their relationship to dietary patterns as a basis for policy and programme planning. The survey results included:
From the data, a vulnerability profile was derived, identifying lower income households, single sex households (male or female), male headed households, younger men and older persons living alone as most likely to have diets which are less diverse and/or less adequate. Specific recommendations for action could then be made. This information can be used by policy-makers and programme-planners in targeting intervention strategies and follow-up studies. For example, based on the findings, CFNI/PAHO commissioned a qualitative study on the factors affecting diet and exercise patterns in Barbados. |
Exclusive breastfeeding is recommended for babies in the first six months of life: except for medicine, the infant is given no other foods or liquids except for breastmilk. Breastmilk (especially when offered exclusively) has many benefits including decreased rates of diarrhoea, respiratory tract infections, otitis media and other infections. Intellectual and motor development is improved and the risk of chronic disease later in life is decreased. The benefits of breastfeeding can have implications for the economic welfare of the family and lessen the burden on the health care system.
|
Box 5.3 GENERATION OF INFORMATION TO TARGET SOCIAL WELFARE PROGRAMMES IN GUYANA In order to effectively implement programmes for the relief of vulnerable groups, identification of these groups should not be limited only to a country-wide basis but according to their specific locations (regions, parishes, towns etc.). Additional available information on other threats to food security and nutrition, such as food production or disasters in the specific area, may be used to target programs. If risk-mapping technology e.g. GIS is available, this can greatly enhance depiction of the data to represent the nutrition problems. The food consumption survey in Guyana was appropriate for this purpose. In the 2002 survey, in addition to food consumption and anthropometry, data collection included:
These data were collected from the country's major ethnic groups (Afro-Guyanese, Indo-Guyanese and Amerindians) and dwelling areas (urban, rural and interior). Data on age, education, employment and other socioeconomic characteristics are also available from the survey. Indicators were developed and the resulting information presented to key stakeholders. Use of this information will improve the targeting of social programmes for vulnerable groups. |
Hypertension and diabetes are chronic noncommunicable diseases. Information may be collected on either reported prevalence or by actual measurements. However, more than one measurement is necessary to confirm the existence of either of these conditions.
As mentioned earlier in this chapter, other data on the survey population is needed to make the information useful for policy and program development.
It is therefore critical that data on demographic and socio-economic characteristics be collected and integrated into the analyses of the data so as to provide useful guidance to policy makers and program managers
|
[4] CFNI has compiled food
composition tables for use in the English-Speaking Caribbean. In addition a food
composition data base of commonly eaten Caribbean foods is being developed on an
ongoing basis as part of process of planning and implementing food consumption
surveys in the region. [5] Recommended dietary allowances (RDA) - authoritative, quantitative estimates of human requirements for essential nutrients, usually set out with different amounts (in weight/day) considered to be adequate to meet the nutrient needs of practically all healthy persons. They are also referred to as recommended nutrient intakes (RNI), recommended dietary intakes (RDI) or dietary reference values (DRV). [6] Population nutrient goals - the average intake judged to be consistent with maintenance of health in a population. [7] Indicators i-iii are discussed in Chapter 4. [8] Risk of other clinical problems increased |
