Lack of association between nutritional status and change in clinical category among HIV-infected children in Brazil

ABSTRACT CONTEXT AND OBJECTIVE: Malnutrition is common among HIV-infected children. Our objective was to study the occurrence of malnutrition and its relationship with changes in clinical category among HIV-infected children. DESIGN AND SETTING: Longitudinal study, at the Pediatrics Department and Pediatrics Investigation Center (CIPED), Faculdade de Ciências Médicas da Universidade Estadual de Campinas (Unicamp). METHODS: We reviewed the hospital records of 127 vertically HIV-infected children. Anthropo-metric measurements were obtained at the beginning of follow-up, at clinical category change and five months later. These were converted to z-scores of weight/age, height/age and weight/height. Data were presented as means, standard deviations, frequency counts and percentages. The Wilcoxon and Kruskal-Wallis tests and odds ratios were used in the analysis. RESULTS: We found that 51 (40.2%) were undernourished and 40 (31.5%) were stunted, with higher risk of being included in clinical category C. There was an association between nutritional condition and the clinical categories of the Centers for Disease Control classification (1994), and with age at symptom onset (except for height z-score). During follow-up, 36 patients (28.4%) changed their clinical category, which occurred early among the undernourished patients. The group that changed its clinical category maintained the same z-score distribution for weight, height and weight/height throughout follow-up. CONCLUSION: Aids manifestation severity was associated with nutritional status and with age at symptom onset, but change in clinical category was not followed by worsening of nutritional status.

of Campinas, State of São Paulo, who were referred to the Aids children division of Universidade Estadual de Campinas university hospital between January 1989 and July 1999, we selected a total of 127 vertically infected children.These infants received medical care in the outpatient service at least once a month on a routine basis.Of these 127 selected children, 71 were males (55.9%) and 56 were females (44.1%), and all fulfi lled the diagnostic criteria for Aids. 3,4ain measurements: From each patient's hospital record we obtained information about gender, birth weight, breastfeeding, age at onset of symptoms, antiretroviral therapy, weight, height and CDC clinical category at diagnosis (measurement M1), at the time of category change (M2) and fi ve months after that change (M3).The clinical categories N and A were analyzed together.
According to their ages at disease onset, the children were grouped as rapid progressors (signs before 12 months old) or slow progressors (signs after 12 months old).
Weight was measured using a Filizola electronic balance.Up to the age of two years, the length was measured with a horizontal wooden anthropometer and, after this age, with a vertical one.
Waterlow et al. 8 proposed a functional classification for child malnutrition that separated children who had acute malnutrition from those with chronic malnutrition.The acutely malnourished children were those with adequate height for age but inadequate weight for height (wasted).The chronically malnourished children were those that had inadequate height for age (stunted).Chronically malnourished children could also be acutely malnourished, in which case they would be both stunted and wasted.

METHODS:
We reviewed the hospital records of 127 vertically HIV-infected children.Anthropometric measurements were obtained at the beginning of follow-up, at clinical category change and fi ve months later.These were converted to z-scores of weight/age, height/age and weight/ height.Data were presented as means, standard deviations, frequency counts and percentages.The Wilcoxon and Kruskal-Wallis tests and odds ratios were used in the analysis.

RESULTS:
We found that 51 (40.2%) were undernourished and 40 (31.5%) were stunted, with higher risk of being included in clinical category C.There was an association between nutritional condition and the clinical categories of the Centers for Disease Control classifi cation (1994), and with age at symptom onset (except for height z-score).During follow-up, 36 patients (28.4%) changed their clinical category, which occurred early among the undernourished patients.The group that changed its clinical category maintained the same z-score distribution for weight, height and weight/height throughout follow-up.
CONCLUSION: Aids manifestation severity was associated with nutritional status and with age at symptom onset, but change in clinical category was not followed by worsening of nutritional status.

Statistical methods:
The weight and height values were transformed into z-scores, utilizing National Center for Health Statistics (NCHS) 9 as the reference curve.The Waterlow classification was used to determine the nutritional status, taking z scores of less than −2 standard deviations (SD) as the cutoff points. 8Z-scores outside of these parameters were not included.
Descriptive data were reported as means with standard deviations, frequency counts and percentages.For statistical analysis the Wilcoxon and Kruskal-Wallis tests and odds ratios were used.

RESULTS
The general characteristics of the children are shown in Table 1 Concerning the children's nutritional status, 68 (53.5%) were classified as eutrophic, 40 (31.5%) as stunted, 8 (6.3%) as chronically undernourished and 3 (2.4%) as wasted.Eight children could not be classified because of incomplete data.The distribution of the z-scores for weight-for-age (p < 0.01), height-for-age (p = 0.002) and weight-for-height (p = 0.01), according to the clinical category at M1, showed statistically significant differences (Table 2).
The mean age at onset of symptoms was recorded for only 108 of the HIV-infected children.It was significantly different between the eutrophic and undernourished groups (p = 0.002).In the eutrophic group, the mean age at onset of the symptoms was 24.19 months (minimum = 1, maximum = 127 and median = 16 months) (Table 3).In this group, height-forage was −1.95 (0.84), weight-for-age was −0.53 (0.97), and weight-for-height was 0.03 (1.05).
In the undernourished group, the mean age at onset of symptoms was 13.16 months (minimum = 2, maximum = 86 and median = 7 months) (Table 3).The height-for-age was −2.87 (0.97), weight-for-age was −2.76 (1.01) and weight-for-height was −1.11 (1.21).In this group, the chance of being included in clinical category C was higher than for the eutrophic group (odds ratio = 3.23, confidence interval = 1.40 -7.53) (Table 4).We found significant associations for weight-for-age (p < 0.05) and weight-for-height (p < 0.05) with age at onset of symptoms, with the lowest values in the rapid progressor group.There was no association for height-for-age (p > 0.05).
During follow-up, 36 patients (28.4%), of whom 22 were eutrophic and 14 undernourished, changed their clinical category over a mean period of 20 and 8 months, respectively (p = 0.008).The group that changed clinical category maintained the same weight-for-age, height-for-age and weight-for-height at M1, M2 and M3 (p > 0.05) (Table 5).

DISCUSSION
In the group of children studied, the occurrence of malnutrition was associated with the age at onset of symptoms and the severity of CDC clinical category at diagnosis.The clinical category change was earlier among those patients who were already undernourished when diagnosed, although there was no significant change in weight-for-age, height-for-age and weight-forheight for any of the children evaluated.
Our data showed that undernutrition led to a 3.23-times increased risk that the children would be classified in clinical category C (Table 4).][14] The correlation between clinical category and undernutrition reinforces the concept of natural evolution that underlies the CDC classi-Sao Paulo Med J. 2005;123(2):62-5.fication.This classification system represents the present moment of HIV infection as well as the natural evolution of the disease.The symptoms described in each category follow those that progressively occur within the immunological system among non-treated Aids cases.The age at which symptoms began was found to be strongly related to nutritional status.The rapid progressors had lower mean weight-for-age and weight-for-height, while the mean age at onset of symptoms was lower among the undernourished children than among the eutrophic children.
Early development of symptoms has been described as a mark of disease severity and decreased survival time. 15,16Recently, the use of highly active antiretroviral therapy has improved life expectancy and the quality of life for these children. 17,18rogress in the treatment of Aids has changed it into a chronic disease.In fact, the nutritional damage pattern found in the present study, of low weight-for-age and height-for-age with normal weight-for-height, is the same pattern as observed for other chronic diseases.
Malnutrition among HIV-infected patients is multifactorial.Increased energy requirements, decreased calorie intake, intestinal malabsorption and economic problems contribute towards malnutrition among this population. 19,20rom the outset of follow-up, such patients undergo antiretroviral therapy, opportunistic infection prevention and early treat-ment for secondary bacterial infections.These actions may result in viral load reduction, improvement of clinical status and less wasting syndrome.Furthermore, social workers play an important role in the overall initiative for Aids therapy by aiding the family with their social, economic and affective problems.
Patient adherence to the overall treatment ensures enhancement of their quality of life.Through this, patients who changed their clinical category during the follow-up may be able to maintain the same initial weight-forage, height-for-age and weight-for-height, in the way that was observed in our study.

CONCLUSION
We observed that the severity of Aids manifestations was associated with nutritional status and with the age at onset of symptoms, but changes in clinical category were not followed by worsening of the nutritional status.

Table 1 .
General characteristics of HIV-1 vertically infected children at diagnosis

Table 2 .
Comparison of the mean weight, height and weight/height z-scores of 127 vertically HIV-1 infected children, in clinical categories N, A, B and C at diagnosis

Table 3 .
Comparison of the general characteristics between eutrophic and malnourished children at diagnosis of HIV infection

Table 4 .
3isk for undernourished and eutrophic HIV-infected children to be classified in clinical category C of the Centers for Disease Control and Prevention revised classification system for HIV infection in children, 19943

Table 5 .
Comparison of the mean weight, height and weight/height z-scores of HIV- *A/N: mild or no symptoms; B: moderate symptoms; C: severe symptoms.