点击显示 收起
1 From the Faculty of Physical Education and Sport Sciences, Universidade Pedagógica, Maputo, Mozambique (AP); the Faculty of Sports Sciences, Universidade do Porto, Porto, Portugal (JARM); the Faculty of Medicine, Universidade Eduardo Mondlane, Maputo, Mozambique (AP and AD); and Katholieke Universiteit Leuven, Leuven, Belgium (GB).
2 Supported by Fundação Calouste Gulbenkian (Portugal). 3 Address reprint requests to A Prista, Faculty of Physical Education and Sport Sciences, CP 2107, Maputo, Mozambique. E-mail: aprista{at}virconn.net.
ABSTRACT
Background: Studies are needed to test the relevance of the anthropometric criteria for health and well-being, particularly in developing countries.
Objective: The objective of the study was to identify the relevance of anthropometric indexes as indicators of nutritional status.
Design: The sample consisted of 2316 subjects (n = 1094 males, 1222 females) aged 6–18 y from Mozambique. Anthropometric variables, maturity stage, physical fitness, physical activity, and metabolic fitness were measured. Samples of blood, urine, and feces were obtained. Subjects were classified in 5 nutritional groups labeled normal, low height-for-age (stunted), low weight-for-height (wasted), low height-for-age and low weight-for-height (stunted and wasted), and overweight, according to cutoffs set by a World Health Organization expert committee. Socioeconomic status was classified according to region of residence.
Results: Prevalence rates for males and females, respectively, in the nutritional groups were 3.0% and 2.3% (stunted group), 21.9% and 10.0% (wasted group), 3.0% and 0.8% (stunted and wasted group), and 4.8% and 7.7% (overweight group). With control for age, socioeconomic status, and maturity stage, the overweight group performed significantly worse than did the other groups on most of the fitness tests. Compared with the normal group, the 3 undernourished groups performed significantly worse in absolute strength tasks, better in endurance tasks, and equally in flexibility and agility. Very few differences were found in physical activity scores. The 3 undernourished groups had scores for the biochemical indicators that were similar to those of the normal group and had more favorable profiles for blood pressure and cholesterol.
Conclusions: In this population, the cutoffs used to classify overweight status appear to appropriately identify potential health problems. No relevance to health was found for the lower cutoffs identifying undernourished children.
Key Words: Nutritional status • anthropometric indicators • Africa • children • adolescents • physical fitness • stunting • wasting • overweight • Mozambique
INTRODUCTION
Although the unfavorable influence of malnutrition on the growth of infants and children is well established, the classification of the nutritional status of a child by means of the child’s anthropometric percentile position can be misleading (1). Thus, the definition of normality in populations of developing countries on the basis of reference values from studies of children and adolescents in developed countries is controversial (2–5).
Because the prevalence of malnutrition in adolescents is considerably less than that in children, relatively few studies have been carried out on this topic with school-age populations. In addition, limitations in establishing relations between anthropometric indicators and nutritional status in adolescents appear to have hindered research in this area (6). Despite the lack of a robust biological meaning and the difficulties in interpreting the cutoffs in adolescents, anthropometric indexes are still the most common tools used in public health, particularly in developing countries. Thus, determining the biological meaning of those indicators is of interest.
In developing countries, health and well-being are also measured by a person’s capacity to perform work and resist disease (1). However, the effect of nutritional status as an indicator of health and disease is unknown. Thus, the present study was conducted to evaluate the accuracy of the anthropometric indicators currently used to classify nutritional status in the school-age population of Maputo, Mozambique. We hypothesized that children and adolescents classified as malnourished would exhibit lower levels of physical activity, less work capacity, and poorer health indicators.
SUBJECTS AND METHODS
Subjects and design
The subject population consisted of 2316 students (n = 1094 males, 1222 females) aged 6–18 y who were selected from 5 schools from different regions of Maputo and who represented a large range of socioeconomic status (SES) and living conditions. Anthropometric variables, physical fitness, physical activity, and metabolic fitness were measured, and samples of blood, urine, and feces were collected between June 1999 and May 2000. All information was collected by a team of 20 researchers and technicians, including physicians, exercise physiologists, nurses, and physical educators. To reduce interobserver variability, each variable was assessed by a maximum of 2 observers. The parents or educators of the subjects were asked to give their written informed consent after receiving a letter explaining the study. The study was approved by the National Institute of Health of Mozambique and the scientific and ethics committees of the Faculty of Medicine of Eduardo Mondlane University.
Anthropometric measures and maturity stage
Height was measured (± 0.1 mm) with a Harpender stadiometer (Holtain, Crymych, United Kingdom). Subjects were weighed naked on a weight scale (M 01-22-07-245; Secca, Germany). Values were recorded to the nearest 100 g. Maturity stage was assessed according to the criteria of Tanner (7) for secondary sexual characteristics, namely, the growth of pubic hair.
Physical fitness
Physical fitness was assessed according to a battery of standardized tests (8–10) that included those for sit and reach, trunk lift, curl-up, flexed arm hang, standing long jump, handgrip, 10 x 5-m run, and 1-mile run. These tests are considered measures of flexibility, dynamic muscular strength and endurance, muscular power, isometric strength, agility, and cardiovascular endurance.
Physical activity
Physical activity was assessed by a questionnaire developed and validated for this specific population (11). Briefly, subjects were asked how many times per week they performed each activity in a list of activities characteristic of the population. For practical reasons, activities were grouped by type, such as household tasks, sports, playing, and walking. Because the living situation in Maputo had changed since the validation study was done, questions about TV viewing and video-game playing were added to the questionnaire. The energy cost of each activity was determined according to tables of estimated energy cost in metabolic equivalents. For activities not included in the tables, metabolic equivalents were estimated from observations in a pilot study based on heart-rate recordings during the specific activities. Each subject was assigned an activity coefficient, calculated as the sum of the products of the energy cost of each activity (in metabolic equivalents) multiplied by the number of times per week the subject performed each activity.
Clinical evaluation
In a subsample of 555 males and 616 females, samples of fasting blood, urine, and feces were collected. Samples of 3 and 5 mL venous blood were drawn into evacuated tubes containing EDTA and into plain evacuated tubes, respectively, for hematologic and biochemical analyses, respectively. Hematologic values were measured with a Coulter counter (Coulter Electronics, Hialeah, FL), and biochemical values were measured with a Beckman analyzer (Beckman Instruments, Fullerton, CA).
Blood pressure was determined as the mean of 2 evaluations obtained while the subject was in a seated position after 5 min of rest; we waited 3 min between readings and used phase 1 and phase 5 Korotkoff sounds. All subjects underwent medical examinations and were considered clinically healthy.
Socioeconomic status
The subjects’ schools were used as the criterion for SES. Students from private schools were classified as having high SES (HSES), those from public schools in the center of the town were classified as having middle SES (MSES), and those from the public schools on the periphery of the town were classified as having low SES (LSES). This classification was previously shown to provide valid determinations of socioeconomic groups in Maputo (12).
Nutritional status
Classification of nutritional status was made according to public health criteria recommended by a World Health Organization expert committee (6). For analytic purposes, the subjects were divided into 5 groups: 1) normal, 2) low height-for-age (stunted), 3) low weight-for-height (wasted), 4) low height-for-age and low weight-for-height (stunted and wasted), and 5) overweight. The criteria used for the classification of nutritional status are described in Table 1.
View this table:
TABLE 1 . Criteria for classification of nutritional status1
Statistical analysis
Analysis of covariance was used to test for differences among groups with the use of age, SES, and maturity stage as covariates and of nutritional group as the main factor. When one or more covariates were not significant, the analyses were repeated without those variables. Post hoc examinations were made by using Bonferroni correction. All the analyses were sex-specific. When analyses of proportions were needed, a chi-square test was performed. The statistical analysis was done with the use of SPSS software, version 10.0 (SPSS Inc, Chicago). The level of significance was set at P < 0.05 for all analyses.
RESULTS
Nutritional classification
Descriptive statistics for height, weight, and body mass index (BMI; in kg/m2) by age and sex are given in Table 2. Results for nutritional status by age and sex are shown in Table 3. On average, Maputo children had negative z scores for height-for-age. However, boys aged 11 y and girls aged 12 y had positive scores, which suggested differences in the prevalence of low height-for-age across age cohorts. The proportion of students classified as stunted was also greater in the older groups (Table 3).
View this table:
TABLE 2 . Descriptive statistics for height, weight, and BMI by age and sex1
View this table:
TABLE 3 . z Scores for height-for-age by age and sex and prevalence of normal weight, stunting, wasting, stunting and wasting, and overweight by age and sex
The prevalence of wasting in the age groups ranged from 9.1% to 27.2% in the males and from 5.1% to 20.7% in the females (Table 3). The males had a significantly higher incidence of wasting than did the females (chi-square; P < 0.001). The prevalence of those who were simultaneously stunted and wasted was low (males: 3.1%; females: 0.8%). The prevalence of overweight was significantly (P = 0.002) higher in the females (7.7%) than in the males (4.8%). No specific tendency in overweight prevalence across age groups was observed (Table 3).
Nutritional classification by SES is shown in Table 4. The observation of the adjusted residual scores from the chi-square test indicates that, among the males, the HSES group had a significantly (P < 0.001) lower prevalence of stunting and wasting than did the LSES and MSES groups, but a significantly (P < 0.001) higher prevalence of overweight. Among the females, the LSES group had the lowest prevalence of overweight, the MSES group had the lowest prevalence of wasting, and the HSES group had the lowest prevalence of stunting (Table 4).
View this table:
TABLE 4 . Prevalence of normal weight, stunting, wasting, stunting and wasting, and overweight by socioeconomic status1
Physical fitness
The differences in physical performance between the nutritional groups are shown in Table 5. With the exception of the sit-and-reach data for the males and the trunk-lift data for the females, the results indicated significant differences between 2 groups. Thus, after we controlled for age, SES, and maturity stage, the nutritional groups performed significantly differently on several tasks (Table 5). The overweight males and females performed significantly worse on most of the tasks than did the other groups. However, they performed significantly (P < 0.001) better than did the other groups on handgrip exercise. Post hoc comparisons between the normal group and the wasted group showed that the normal group performed significantly (P < 0.001) better in the standing long jump (males only) and handgrip (both sexes). Wasted females performed significantly (P < 0.001) better in the 1-mile run than did the females in the other groups. Compared with the normal group, the stunted group performed significantly worse in the curl up and standing long jump (males) and handgrip (both sexes), but significantly better in the 1-mile run (females). The stunted and wasted group had the same profile as the wasted group, except for the standing long jump (males) and handgrip (females), in which their performances were significantly worse.
View this table:
TABLE 5. Comparison between nutritional groups of fitness indicators with age, socioeconomic status (SES), and maturity stage (MS) as covariates1
Physical activity
Physical activity was rated in 5 subcategories, and the results among the nutritional groups are shown in Table 6. No significant differences were found for most of the comparisons. The males in the stunted and stunted and wasted groups had significantly higher scores on household tasks than did the males in the normal group, whereas the males in the normal group had significantly higher scores in TV watching than did the males in the stunted and stunted and wasted groups. Among the females, the stunted group had the lowest scores on household tasks. The difference among the females in sports participation was due to a higher score in the overweight group than in the other groups.
View this table:
TABLE 6 . Comparison between nutritional groups of physical activity scores with age, socioeconomic status (SES), and maturity stage (MS) as covariates1
Clinical indicators
Metabolic fitness was also compared among the nutritional groups (Table 7). Systolic and diastolic blood pressures were significantly (P < 0.001) higher in the overweight males and females than in the other groups. Conversely, the wasted males had significantly lower systolic and diastolic blood pressures than did the normal males, whereas the wasted females had significantly lower systolic blood pressure than did the normal females (both P < 0.001). Mean cholesterol concentrations were significantly higher in the males and females in the overweight group, but, in the males, the statistical significance disappeared after control for covariates. No significance was found for the other indicators.
View this table:
TABLE 7. Comparison between nutritional groups of blood pressure and serum concentrations of cholesterol, calcium, and hemoglobin with age, socioeconomic status (SES), and maturity stage (MS) as covariates1
Subjects were also classified as having or not having parasites in urine, feces, or both. The prevalence was 31.1% in the males and 34.2% in the females. With the use of adjusted residuals from a chi-square test to compare prevalence between nutritional groups, results showed that, in the males, the overweight group had a significantly lower prevalence of parasites (data not shown). In the females, the 3 stunted and wasted groups had higher prevalences than expected. The comparison among SES groups showed that the percentage of students with parasites increased in both sexes as SES declined (HSES: 11.4%; MSES: 29.4%; and LSES: 41.2%).
DISCUSSION
The primary observation of this study was that the prevalence of stunting, defined according to the criteria used in this study, has decreased dramatically in Maputo schoolchildren, compared with the findings in earlier studies conducted in Mozambique (5). In contrast, the prevalence of overweight has increased. Over the period of this study (September 1992–June 2000), the proportion of subjects with stunted growth dropped from 34.5% to 3% in males and from 24.6% to 2.3% in females, whereas the proportion of overweight subjects increased from 0% to 4.8% in males and from 4.9% to 7.7% in females. These changes may be a result of the changes observed in socioeconomic conditions of Mozambique, particularly in Maputo (13), and they confirm the trend in developing countries toward greater prevalence of obesity (14, 15). Since 1994, a dramatic change in the political and economic situation in Mozambique has taken place, particularly in Maputo. Three decades of war in Mozambique, especially the period from 1980 to 1992, were characterized by food restrictions and a lack of consumables. This situation changed after the peace agreement of 1992, which led to increases in urbanization, sedentary occupations, and the availability of private transport and to the advent of "fast food." The mean height-for-age z scores in this study were positive until 11 y in boys and until 12 y in girls, which may suggest that the higher prevalence of stunting in older adolescents is associated with the fact that their infancy and childhood occurred during the difficult times of war, which was not the case for the younger children. However, a delay in maturation among the study subjects, relative to the reference population, could also be a cause of the observed negative z scores.
According to the criteria used, the primary nutritional deficiency in the study population was wasting. The males had a greater prevalence of wasting than did the females. This difference could be due to cultural factors, socioeconomic factors, or both. Among poor families, females are more likely than are males to drop out of school. As a consequence, there will be a higher percentage of males than of females from poor families attending school, which could explain the higher prevalence of wasting in males in school.
Nutritional deficiences also differed by the SES of the groups in the study. In the HSES group, there was a higher prevalence of overweight, whereas combined stunting and wasting was more prevalent in the LSES males. Although these findings were expected, the higher prevalence of wasting among the HSES females was unexpected. No obvious explanation for this finding was established.
The negative effects of overweight on fitness performance are widely evident (16–18). The exception of static strength (as tested by handgrip) in the overweight group may be due to greater amounts of muscle mass in overweight children and adolescents.
The undernourished groups performed better than the overweight group on most fitness tests. This observation cannot be compared with other results because previous studies that compared malnourished groups did not include overweight subjects (19–23). In general, studies that compared marginally malnourished children and adolescents with their normal peers found that absolute levels of maximal oxygen uptake, muscle strength, and anaerobic power are higher in the normal groups, but if the difference in body mass is taken into account, motor performance tends to be equal or even better in the malnourished groups (12, 24–27).
The topic of the advantages and disadvantages of being small has been controversial. Some authors suggested that a reduced growth rate is an adaptive response to adverse conditions and that people from developing countries are mechanically more efficient (28). This notion was rejected by studies in which muscle efficiency was compared between nutritional groups and no differences were found (26, 29). The advantage of being small is also contested by the argument that, in countries where manual labor plays an important role in productivity, the absolute capacity to perform work would be more important than capacity relative to size (30, 31). In addition, it has been shown that the relation between body dimensions and motor performance changes according to the task and population tested, because of an important cultural influence on motor performance that must be taken into account in the analysis (32, 33). Indeed, habits of physical activity, related to socioeconomic conditions and culture, have been suggested as having a strong influence on the performance of some tests, which may interfere with the validity of the results.
Very few differences were found in physical activity when nutritional groups were compared. Instead, data from the present study suggest a large influence of SES on the habits of physical activity, because SES was a significant covariate in the analyses of almost all variables (Table 6). Prista et al (12) found that LSES children and adolescents in Maputo had higher levels of physical activity because of the higher demands of survival activities and playing, but that they spent less time in formal sports than did their more privileged peers. Despite this socioeconomic effect, the present study does not show a consistent relation between nutritional status and levels of physical activity. This inconsistency was also observed in several other studies comparing nutritional groups within school-age populations (5, 20, 22, 30, 34). Peer pressure and survival needs are cited as the main reasons for this finding (22). The absence of reduced levels of physical activity in nutritionally deficient groups becomes controversial because it would be reasonable to expect a significant reduction in activity if a nutritional deficit is present, as has been observed in infants (35–37). Limitations of the assessment of physical activity may impair more robust conclusions about this association.
Data on risk factors for cardiovascular disease provide some evidence for the sensitivity of the BMI cutoff in classifying children and adolescents as overweight because risk factors for cardiovascular disease are associated with increased proportions of fat. Thus, the cutoffs used and proposed by a World Health Organization expert committee (6) have some biological meaning. However, no relevance was found for the lower cutoff, which is supposed to identify children classified as wasted. The 3 wasted and stunted groups also did not have anemia or reduced calcium concentrations, and nor were any other biochemical indicators reduced (data not shown).
A low prevalence of parasites in the overweight group may be related to SES rather than to body size per se. Although the presence of parasites has been shown to negatively influence performance on fitness tests (38, 39), this was not observed in the present study, maybe because of the strong relation with SES.
Anthropometric indexes in preschool children may be linked to increased mortality and morbidity (40–43), but data on school-age populations are scarce and controversial (1, 6, 29). Thus, despite the worldwide use of anthropometric criteria to classify nutritional status in children and adolescents, the biological meaning of these criteria is not yet well established. With the exception of the tasks that are markedly associated with body size, this study was unable to elucidate any biological meaning for the cutoffs proposed to identify undernourished children aged 8–17 y (6). Previous studies of schoolchildren from Maputo had similar results (5). This weakness of anthropometric data in the classification of nutritional status has been recognized by the World Health Organization expert committee. That group recommended, that, for uniform reporting purposes and in the absence of other data specifying optimum cutoffs for BMI in adolescence, that the BMI-for-age data for US children be used on a provisional basis until better reference data for adolescent growth are available (6).
Although efforts have been made to construct international references as alternatives (44), a valid cutoff for low BMI or another weight-for-height indicator with an associated health significance in school-age children and adolescents has not yet been identified. The findings of the present study indicate that the upper limit (overweight) seems to be sensitive enough to identify nutritional problems in school-age children and adolescents. Conversely, the indicator of underweight was not shown to be useful, and further research is needed to identify the appropriate indicators in developing countries for classifying a person as underweight. However, it must be recognized that the construction of local norms, thought to be one of the best solutions, has operations and cost implications that are not easy to resolve in developing countries.
ACKNOWLEDGMENTS
We express our appreciation to the Exercise Science Department at Syracuse University and to Bo Fernhall, chair of the department. AP was a Fulbright Scholar in the Exercise Science Department.
AP participated in the study design, data collection, data analysis, and writing of the manuscript. JM and GB participated in the study design, data analysis, and writing of the manuscript. AD participated in the study design, data collection, and data analysis. None of the authors had any financial or personal interest in the Fundação Calouste Gulbenkian.
REFERENCES