This is the old United Nations University website. Visit the new site at http://unu.edu
Abstract
Introduction
Methods
Principal findings of interest
Comments
Conclusions
References
Noel W. Solomons, Ivan Mendoza, Luis Gutierrez, and Carlos Monteiro
Noel Solomons and Ivan Mendoza are affiliated with the Center for Studies of Sensory Impairment, Aging and Metabolism (CeSSIAM) in Guatemala City, Guatemala. Luis Gutierrez is with the Department of Physiology of the National Institute of Nutrition in Mexico City, Mexico. Carlos Monteiro is affiliated with the Department of Nutrition in the School of Public Health in the University of Sao Paulo in São Paulo, Brazil.
Mention of the names of firms and commercial products does not imply endorsement by the United Nations University.
The essential units for testing hypotheses about the influences of ageing and urbanization in the CRONOS (Cross-Cultural Research on the Nutrition of Older Subjects) process are the responses to the questions on the questionnaire and the physical measurements of individuals, aggregated into distributions by age, sex, geography, and their combinations. In the Reconnaissance experience, the number of subjects per country (about 24) was insufficient to provide meaningful group or subgroup averages. However, the process allowed the investigators to get a sense of which measurements were consistent across countries and which were aberrant in some sites. In general, despite the lack of inter-site standardization, most of the items in the protocol seem to have been measured with acceptable consistency and accuracy across geographic areas. Problem areas identified were in the measurement of arm span, the determination of arterial blood pressure, and the accuracy of 24-hour recall methods for total energy and nutrient intakes.
The project Nutritional and Food Security Situation of Rural and Urban Elderly from Selected East Asian and Latin American Developing Countries was intended to be conducted in its entirety as originally proposed. That is, we planned to gather data in each country from the entire 600-person sample at one time, and not in a multiphase, staged, or piecemeal fashion. If things had gone as planned, an intensive analysis of the merits and validity of the questions and procedures would, and could, only have come after all 4,800 people in the eight countries had been enrolled and studied. The exigencies of the funding process, however, imposed a change. Some of the professionals involved consider it a mixed blessing that the entire funding was not obtained, as it provided the luxury for a more profound analysis of the questions and procedures in the protocols during a pilot testing phase. This, in turn, allowed for revision and the production of a superior protocol instrument [1].
This paper is concerned with the 104 items in the original Reconnaissance protocol that dealt with the individual as the unit of analysis.
A number of agendas that were hoped to be mutually compatible are embodied in the original motivation of the conceivers of the Reconnaissance project. These included:
· providing descriptive information on rural and urban elderly people that could be useful for policy makers concerned with preventive and curative services for the older populations of the respective countries;Although all eight countries in the Reconnaissance study are developing countries, the third objective placed a high priority on obtaining information that was not only analogous to, but also homologous with, that in the EURONET-SENECA study [1]. We expected that sufficient elements of modernization would be present in the urban populations, which made up two-thirds of the sample, that homologous information could be obtained. Cultural incompatibilities with the rural third of the sample and with the populations of the IUNS-FHLL and EURONUT-SENECA studies might be expected. With China and the Philippines as part of the catchment area of the IONS Committee on Nutrition and Aging for their study Food Habits in Later Life, we expected that some attention had already been given to cultural interpretations of the questions. Our pilot phase turned out to be, in part, an examination of how successfully and universally we could apply questions extracted from the interview protocols of the two aforementioned multicentre studies in our eight populations of interest.· combining the perspective of ecological and social experiences in diverse countries in Asia and Latin America;
· adding data comparable in origin with that from the multicentre EURONUT-SENECA study in Europe and the International Union of Nutritional Sciences (IUNS) multicentre study Food Habits in Later Life (FHLL) to a conjoint database;
· providing experience in urbanization, ageing, and multicentre collaborative research for investigators from selected developing countries.
The original Reconnaissance protocol had a series of 104 individual, numbered response items in the obligatory portion, and some form of appraisal of dietary and nutrient intake was an option. In each site, data were supposed to be obtained from 24 individuals, distributed among two groups according to sex, two according to age, and three according to social conditions and geography, for a total of about 192 individuals in the entire study. Hence, there was a universal admonition not to succumb to the temptation to try to draw any comparative conclusions across subgroups. However, broad generalities that cut across countries and any exceptional (outlier) findings were considered to be worthy of identification. This paper scrutinizes the individual values to draw attention to the broad trends and specific unusual measurements or responses.
This report was developed from a detailed review of the disaggregated data for the specific individuals, as reported in the tabular appendix materials in the specific country reports. The report for the urban population of the Philippines cited these tables, but they were not available. Only the summary of the data in the narrative text was available for this group.
The quantitative dietary intake study was optional. Individual nutrient intake data, based on a 24-hour intake, were available in the archival record for only two countries, China and Thailand.
Interview questions of general interest
In those areas for which disaggregated or aggregated data existed, people in rural areas generally had more poorly constructed houses and fewer water and sanitation facilities than corresponding lower income urban people. Electrical appliances were common in rural households in both Asia and Latin America in areas where electricity was available.
With respect to self-rated health (a variable often used in ageing research that is very predictive of actual health outcome and survival [2]), we generally did not see a stark polarization between the older and younger subjects with respect to lower and higher ratings. In these third world countries, people between 30 and 40 years old often rated their health as “below excellent” and even as “fair.” Data on self-rated health may be available from elderly populations but may be generally lacking from adults in midlife. With the realization of the full protocol, there is a truly rich possibility of exploring the self-rated health paradigm across generations in low-income populations.
Anthropometry and physical data
As commented by de Groot et al. [3], a meticulous process of standardization was felt to be necessary so that all of the data collected in the European EURONUT-SENECA study could be compared or pooled. The financing of the Reconnaissance phase did not allow the teams to be brought together before taking measurements in the field. Only the instructions in the Reconnaissance manual and the investigators’ background in field research provided guidelines for standardizing the independent data collection ventures of the eight countries. In cross national studies, new knowledge about the normative aspects of long-bone measurements, such as knee height and arm span, might be an important outcome of the full protocol.
The body mass index (BMI) is constructed from measurements of weight in kilograms and height in metres and is expressed as kilograms per square metre. It is relatively robust without cross-centre standardization, as weight and height have relatively small errors with acceptable techniques and measuring devices.
The BMI can be interpreted in terms of energy nutriture [4]. A BMI less than 18.5 kg/m2 is a proposed cut-off point for chronic energy deficiency. Overall, 12 persons had a BMI below this level: 2 in China, 5 in Indonesia, and 1 each in Malaysia, the Philippines, Thailand, Guatemala, and Mexico. In Brazil, no subject fell into this range. A BMI less than 17 kg/m2 has been defined as indicating advanced chronic energy deficiency in adults. Of the aforementioned, 3 subjects in the series had a BMI in this range: 1 in the Philippines and 2 in Indonesia. At the other end of the energy-status scale, a BMI greater than 30 kg/m2 is considered to represent overweight. This level was found in 1 subject each in China, Indonesia, Malaysia, Thailand, and Guatemala, 4 subjects in Brazil, and 6 in Mexico. This suggests, in the broadest terms, a tendency towards more underweight people in Asia and more overweight people in Latin America.
The knee-height [5] estimate deserves mention. For most of the research groups, this was the first time that they had made the measurement. Knee-height calipers with instructions were donated to each group (courtesy of Ross Laboratories, Columbus, OH, USA). Analysis of the data shows it to be in the appropriate range, in relation to standing height, indicating a successful application.
Similarly, most research groups measured arm span [6] for the first time. Arm-span measurements are used to correct measurements of stature in the elderly, who lose height as part of the process of senescence, combined with osteoporotic vertebral compression in some individuals [7]. In theory, the height/arm span index is close to 1 in young adults and less than 1 in the elderly [6]. The absolute difference between measured height and measured arm span in the elderly is on the order of 0 to 7 cm. In the Malaysian experience, an obvious error was detected, as arm spans were reported that were 12 to 17 cm greater than the corresponding height. In the rural Philippines, the investigators measured arm length rather than arm span as defined by the manual; this was a departure from the intent of the protocol.
The measurement of arterial blood pressure without initial cross-standardization of field teams presents some interesting findings in the preliminary data. “Digit preference,” or rounding off the final digit of a diastolic or systolic measurement to five or zero, is a well-documented phenomenon in reports of blood-pressure measurements by human examiners. Careful training or the use of “zero-muddling” devices with the sphygmomanometer can eliminate this tendency. In the present series, the occurrence of digit preference varied between countries. All diastolic and systolic measurements from the Philippines ended in zero. Of 48 measurements from Mexico, 47 ended in zero and 1 ended in five. In measurements from Guatemala and China, on the other hand, the whole range of digits was represented.
The upper limit for normal systolic and diastolic arterial pressure was 140/90 mm Hg. By this criterion, 7 of 23 Brazilian subjects measured were classified as hypertensive; 6 of them were between 60 and 70 years old. In the rural site in the Philippines, 50% of the subjects had a diastolic reading greater than 90 mm Hg. In China, 4 individuals with diastolic readings greater than 90 mm Hg were found. By contrast, in Guatemala, only 1 subject had a diastolic reading greater than 90 mm Hg.
The haematological status of young and old alike, with respect to the single indicator (haemoglobin concentration), was generally acceptable across countries. Most subjects had haemoglobin concentrations greater than 11 g/dl, and only one value, 9.6 g/dl from a subject in Thailand, was clearly in the anaemic range. Of course, not all populations were living at sea level. In Guatemala and Mexico highland populations were represented, so that cross-national comparisons of haemoglobin would have to be adjusted for altitude [8].
Dietary intake
Dietary intake and dietary behaviour were assessed by standard questions in the interview format, as well as by 24-hour recall in some settings. It was a constant finding across all sites that the staple cereal of the region was consumed with the greatest frequency and that meat, poultry, and fish were consumed sparingly in the low-income groups but with greater frequency in the urban middle class. This finding confirms the basic tenets of the DULL multicentre study that people can reach advanced ages in good health while consuming a variety of different diets [9].
Twenty-four-hour recall data were optional for each participating team. All groups except the one in Guatemala performed 24-hour recalls, a single day’s inquiry for each of the participants. However, in the final reporting, only China and Thailand reported the disaggregated data for individuals. Although these qualify as individual data, intake of nutrients cannot be estimated with stability for an individual with only one 24-hour recall estimate [10-12]. However, the group average of various nutrients would be valid if the number of individuals were adequate.
If we assume that differences exist within countries across sex, age, and location, then combining data within countries for inter-site comparisons presents certain obvious risks. With the symmetrical representation of the samples, however, such an exercise can be undertaken, albeit only with the two countries, China and Thailand, that provided dietary data. On the basis of data from 24 subjects in each country, the average total daily caloric intake in China (2,100 kcal; range, 1,066-4,125 kcal) was greater than that in Thailand (1,981 kcal; range, 654-4,033 kcal). In terms of internal validity, assuming that the basal metabolic requirement for elderly people is at least 1,000 kcal/day, and that 1.2 x energy intake/energy expenditure is compatible with only the most sedentary existence, an intake of 1,200 kcal would represent the minimal daily energy needs. In China 3 subjects reported an energy intake of less than 1,200 kcal on the day of the interview. Two of these were between 60 and 70 years old. In Thailand 8 of 24 subjects reported an energy intake of less than 1,200 kcal for the day; 4 of these were elderly. However, individual values for a single day are notoriously unreliable [11,12].
Some other findings of individual nutrient intake estimates derive from the two sets of 24-hour recalls. In China there was a wide spread in vitamin C intake on the day of the interview, from 0 to 235 ma. In Thailand some subjects reported no intake of either preformed vitamin A or carotenoids. Also interesting was a trend in Thailand towards uniformly low riboflavin intakes and relatively high iron intakes. The range of reported 24-hour calcium intake values was 74 to 893 mg in China and 142 to 1,198 mg in Thailand. The median value for daily calcium intake was greater in Thailand than in China.
Multicentre and multinational cross-cultural studies represent a challenge for researchers who are used to operating within the confines of their own countries or regions.
The conduct of the interviews and measurements, without prior standardization, in this study was similar to that of the FHLL study by the IONS Committee on Nutrition and Aging [1] but quite distinct from that of the EURONUT-SENECA study [3]. Judging from the overall impression of the present pilot data, however, an acceptable degree of consistency in questioning and measuring across sites was obtained.
Most of the centres felt extremely comfortable with the 24-hour recall method. Although it was an optional measure, the data were collected in the pilot sample by most of the centres, but the data were only enumerated and reported by two participating sites. A single determination of this measure cannot provide a stable measurement of individual intake for any nutrient [10-12]. One can conclude that some individuals existed on an unreasonably low chronic energy intake on the basis of the assumptions in the WHO/FAO/UNU report [13]. On any given day, one or another individual may not consume enough energy to balance energy expenditure. However, over time, the average energy intake for both an individual and a population must be in balance with the energy requirement based on body composition, physiological state, and level of activity. This is obviously not the case with famine, acute infectious disease, or excessive weight gain or obesity.
For the authors of this review, with their predominant personal and collective scientific experience in Latin America, it was interesting and enriching to share experiences with Asian colleagues and to learn from their measurements and observations. Because it was possible to enrol only 24 subjects from each country, rather than the 600 originally planned (i.e., 100 from each age, sex, and sociogeographic group), we do not know what consistency would have remained, or what systematic inter-site biases might have emerged, if this multicentre study had been conducted according to the full conditions of the design without face-to-face standardization. The indication from these small data sets is that the diverse centres were in close agreement in the physical measurements, but many culturally related issues arose with regard to the interview questions. It was the consensus of the participants that rewording, consolidation, and even elimination of some of the questionnaire items would make for a better and more universal study. Their suggestions were incorporated into the final protocol [14].
1. Hsu-Hage BH-H, Wahlqvist ML. The IONS study of Food Habits in Later Life: a critique. Food Nutr Bull 1997;18:239-47.
2. Lawton MP, Moss M, Fulcomer M, Kleban MH. A research and service oriented multilevel assessment instrument. J Gerontol 1982;37:91-9.
3. de Groot LCPGM, van Staveren WA, Hautvast JGAJ. Experience of the EURONUT-SENECA study in design, implementation, and data analysis. Food Nutr Bull 1997;18:235-8.
4. Shetty P. James WP. Body mass index: a measure of chronic energy deficiency in adults. FAO Food and Nutrition Paper No. 56. Rome: Food and Agriculture Organization, 1994.
5. Chumlea W. Guo S. Equations for predicting stature in white and black elderly individuals. J Gerontol 1992;47:M197-203.
6. Dequeker JV, Baeyens JP, Claessens J. The significance of stature as a clinical measurement of ageing. J Am Gerontol Soc 1969;17:169-79.
7. Rabe B, Thamrin H, Gross R, Solomons NW. Body mass index of the elderly derived from height and from arm-span. Asia Pacific J Clin Nutr 1996;5:79-83.
8. Dirren H, Logman MHGM, Barclay DV, Friere WB. Attitude correction for hemoglobin. Eur J Clin Nutr 1 994;48:625-32.
9. Wahlqvist ML, Hsu-Hage BH-H, Kouris-Blazos A, Lukito W, IUNS Study Investigators. The IUNS cross cultural study of “food habits in later life”-an overview of key findings. Asia Pacific J Clin Nutr 1995; 4:233-43.
10. Beaton OH, Milner J. McGuire V, Feather TE, Little JA. Sources of variance in 24-hour dietary recall data: implications for nutrition study design and interpretation. Carbohydrate sources, vitamins, and minerals. Am J Clin Nutr 1983;37:98695.
11. Sempos CT, Johnson NE, Smith EL, Gilligan C. Effects of intraindividual and interindividual variation in repeated dietary records. Am J Epidemiol 1985;121:12030.
12. Nelson M, Black AK, Morris JA, Cole TJ. Between and within-subject variation in nutrient intake from infancy to old age: estimating the number of days required to rank dietary intakes with desired precision. Am J Clin Nutr 1989;50:155-67.
13. WHO/FAO/UNU. Energy and protein requirements. Report of a joint WHO/FAO/UNU expert committee. Geneva: World Health Organization, 1985.
14. Gross R. ed. CRONOS (Cross-Cultural Research on the Nutrition of Older Subjects). 3rd ed. Food Nutr Bull 1997;18:267-303.
