Project topics and materials
Get Complete Project Material File(s) Now! »
CHAPTER 6 – ANTHROPOMETRIC MEASUREMENTS AMONG HIV-INFECTED WOMEN OVER A 24 MONTH PERIOD
OBJECTIVES
To establish the longitudinal changes in body composition, as measured by select anthropometric measurements, amongst a cohort of HIV- infected women from six weeks until 24 months after 2 To determine the factors that impact on maternal anthropometric measurements over a 24-month period of postnatal follow-up.
SUBJECTS AND METHODS
HIV-infected women were consecutively recruited from four clinics offering antenatal care (ANC) and PMTCT services in Tshwane between 2003 and 2005 and were followed-up for a period of 24 months after delivery. The four clinics from which the women were recruited are in the peri-urban Mamelodi and Atteridgeville townships. Details on the methodology are described in Chapter 4.
A sample of 53 HIV-negative women was recruited as a comparison group at six weeks postpartum and they were assessed on nutritional status, biomarkers and infant feeding practices at this time only.
Anthropometric measurements
Anthropometric measurements were taken of mothers during the six-week visit as proxy indicators of body composition. These included mid-upper arm circumference measurements (MUAC) and determination of body mass index.
Body mass index (BMI) was calculated as weight in kg divided by height in metres squared. MUAC is the circumference of the left upper arm, measured at the mid-point between the tip of the shoulder and the tip of the elbow (olecranon process and the acromium). Mid-upper arm circumference was measured using a non-stretchable tape.
Height was taken without shoes and measured to the nearest 0.1 cm using a stadiometer (Scales 2000, Durban, SA) and weight was measured in light clothing to the nearest 100g using an electronic digital scale (Scales 2000, Durban, SA). The control mothers were only measured at six weeks postnatally whilst the HIV-infected mothers were measured at intervals between six weeks postnatally and 104 weeks (equivalent to 24 months).
STATISTICAL ANALYSES
Distributions of anthropometric measurements and indices by HIV status were determined. Anthropometric measurements among HIV-negative women were only taken at six weeks after delivery and not continued thereafter. All the anthropometric measurements that were not normally distributed were logarithmically transformed. Adjusted differences by HIV status were obtained from ANOVA models for repeated measures, in which CD4 count and ferritin were covariates. The same analysis was repeated by infant feeding practice of the mothers. Differences were considered to be statistically significant between the groups at p≤0.05. All analyses were carried out using the STATA statistical software package version 9.
To compare differences in anthropometric measurements and body composition between the HIV-infected and non-infected women, the t-test was used. The next phase was to compare groups controlling for ferritin and CD4 cell count, both of which are measures of the inflammatory response.
RESULTS
Comparison between the baseline anthropometric measurements of HIV-infected and un-infected women at six weeks post-delivery
Table 6.1 indicates that at six weeks postpartum, the HIV-negative women weighed on average less and had lower BMI than HIV-infected women. The differences between the groups was significant only for MUAC (p<0.05). The differences between the two groups of women remained significant only for BMI (p=0.037), with HIV-infected women having a greater BMI than their HIV- negative counterparts. The differences were not significant for MUAC and weight after controlling for baseline CD4 count. Using ferritin concentration as a marker of the inflammatory response and controlling for it, there was no significant difference between the groups for weight (p=0.6549) and for BMI (p=0.148), however for MUAC the difference remained significant (p=0.0466). Both groups of mothers had a mean BMI falling into the overweight category of BMI³25 and none of the mothers at six weeks postpartum had a MUAC < 23 cm, which is the cut-off for underweight. Both comparison and study subjects were well nourished.
Anthropometric measurements and infant feeding practices
In order to determine if there was any effect of infant feeding practice on anthropometric measurements, we assessed anthropometric data at six weeks and also at six months between HIV-infected formula- and breastfeeding women (see Table 6.2). The sample sizes differ between the two tables as measurements because of missing data.
As depicted in Table 6.2, at six weeks and at six months after delivery the most significant differences in anthropometric measurements between formula-feeding mothers and their breastfeeding counterparts were for BMI and MUAC (p<0.05). At both visits, the BMI and MUAC of the breastfeeding mothers were lower. At this same time there was no significant difference in CD4 cell count by feeding group.
DISCUSSION
Our study findings are important in that, unlike other studies in Africa that have documented anthropometric measurement changes among HIV-infected women in pregnancy80,83,192, or among rural HIV-infected lactating women8, we investigated anthropometric measurement changes as proxy measures of body composition among peri-urban women over a 24-month period after delivery.
Selection of both MUAC and BMI as measures of body composition among the study participants was primarily based on the simplicity of their collection at the clinics and the fact that they are less invasive techniques and affordable within the public health system. However, some researchers19,193 state that whilst both BMI and MUAC are useful in predicting fat mass they are not useful for determining fat free mass among HIV-positive women and that bio-impedence spectroscopy (BIS) may be a better alternative for this purpose. Others have used skinfold measurements to assess body composition, however this measurement is considered problematic and unreliable as there is a need for “fastidious attention to standardisation and significant training and practice in order to obtain accurate results.”63
In accordance with recommendations made by others8,79 , we undertook baseline measurements of all the HIV-infected and non-infected women at six weeks postpartum, by which time we had estimated that the anthropometric measurements indices would have returned to pre-pregnancy values.
In comparison to the cut-off points for BMI, very few of the study mothers and the control group were underweight (BMI < 18.5). Most of the study mothers had BMI ranges falling into the normal range (18.5 to 24.9 kg/m²) and even the overweight range (25.0 to 29.9 kg/m²).62 These findings may indicate that the pre-pregnancy body weight measurements of the women in this study were either high or within the expected range for their height. This is similar to findings from Rwanda where none of the HIV-infected and non-infected women suffered from chronic energy deficiency in the pre-pregnancy period.83
The finding that BMI levels were high but not significantly different (p = 0.1236) among the HIV-infected women (26.3kg/m²) as compared to the HIV-negative controls (25.0kg/m²) in this study is consistent with findings from the North West Province, where the mean BMI among HIV-infected women was 26.1kg/m² and for the uninfected women 27kg/m². These North West study researchers attributed the high BMI to the fact that most of the infected women were asymptomatic and in the early stages of the disease.139 This same explanation could be the case for our study participants who were also at the asymptomatic stage of HIV disease for the most part for the first two years.
Our findings of a relatively small (1kg only) weight loss among the breastfeeding HIV-infected mothers between six weeks and six months is highly similar to data in KwaZulu-Natal where weight loss (1.4kg between eight weeks and 24 weeks) amongst lactating women occurred, even though 95% of the mothers in their study had CD4 cell counts above 200cells/mm3. The KwaZulu-Natal study established that whilst the breastfeeding women lost weight between the two visits, their BMI levels remained high, BMI > 20kg/m².8 Similarly, we found that the mean CD4 cell count among the breastfeeding mothers at six weeks and six months was greater than 200 cells/mm³, implying that there was no evidence of severe immuno-suppression at this time. Even at six months the breastfeeding mothers still had mean CD4 cell counts (399 cells/mm³) that were slightly higher than amongst the formula-feeders at this time. The difference in weight was due to breastfeeding even though there was no effect on the immune status. From a socio-economic perspective, the formula-feeding HIV-infected mothers had an insignificantly higher socio-economic score than their breastfeeding counterparts and this may have had a positive or protective effect on their weight as they could have also had a greater food-purchasing power, though this aspect was not fully investigated in this study.
All lactating women have physiologically increased energy needs post-partum regardless of their HIV status. If these physiological needs for energy are not met, it is possible that the energy cost of lactation that may result in weight loss due to increased energy requirements.5 It has also been reported elsewhere that the weight and fat loss that is observed among women during lactation is independent of the length of breastfeeding, but rather that it results from a negative energy balance and dietary restriction that is self-imposed by mothers wanting to lose body fat accumulated during pregnancy, or it may be attributable to metabolic or hormonal influences.78 We did not investigate these factors. Regardless of our findings, wherein we had a minority of mothers chosing to breastfeed and where post-partum weight loss was minimal, it has been recommended by other authors that in particular breastfeeding HIV-infected mothers should be provided with nutritional support to avoid any risks to maternal health such as weight loss due to fat mass loss or fat free mass reduction.185 Whilst our findings did not necessarily corroborate these recommendations, on a public health level it may be more appropriate to support a targeted nutritional supplementation approach, prioritising HIV infected women with low anthropometric indices and multiple micronutrient deficiencies.
There is a need for caution in the interpretation of the results in this study, especially when comparing anthropometric measurements between HIV-infected and non-infected breastfeeding women, as there were very few HIV-infected breastfeeding women in the study and this may have had an effect on the results.
Furthermore, this study did not assess trends in anthropometry among the HIV- negative mothers beyond the six weeks after delivery. Perhaps if this had been done it would have provided a better indication as to whether the trends in weight loss among breastfeeding HIV-infected mothers fall within a physiological norm or whether this change is only attributable to the HIV infection itself. Given the fact that our research was conducted in a peri-urban setting, it is possible that there was access to greater variety of foods and possibly more energy- dense sources, which could have resulted in greater weight gain in our study mothers. Our study findings reflect that there was a true difference between HIV-infected breastfeeding women and non-breastfeeding HIV-infected women, with the former losing 1Kg of body weight, whilst the latter remained at the same weight between six weeks and six months. However caution is warranted in the interpretation of this result as this difference may be reflective of a normal physiological occurrence and is to be expected regardless of the HIV status of individual women.
Regardless of infant feeding practice, overall, formula-feeding mothers had no significant change in weight, MUAC and BMI. Similarly, anthropometric trends among the breastfeeding mothers also did not change significantly between six weeks and six months. As such, it is not possible to attribute the observed 1kg weight loss among breastfeeding mothers to feeding mode only. It is possible that the weight loss was in line with the expected levels postpartum or that the mothers were returning to their pre-pregnancy weight levels.
Our research did not use more sophisticated and accurate measurements of body composition which have been used in other studies of this nature and thus could not determine if the weight loss was attributable to greater lean or fat mass loss. Others have documented, using bioelectrical impedence analysis, that HIV-infected underweight women in the USA tended to preferentially lose fat mass whilst conserving their body cell mass.85
Interestingly there was a significant difference between the first and the last mean BMI measurements in the study group, reflecting an overall increase of 0.57kg/m². This could, in part, be attributable to a better disease profile, increased access to a varied diet or fewer reported opportunistic infections. However, considering that several of the mothers in this study were also accessing micronutrients and other dietary supplements (see Chapter 7), this may have also resulted in the changes observed. The study findings are similar to those from the Free State Province, which did not find significant reductions in anthropometric measurements among HIV-infected patients and HIV-negative patients, primarily because the former were asymptomatic and in the early stages of disease progression.139
The South African Demographic and Health Survey (SADHS) of 200357 indicated that in the age group 15–24 years, 11.2% of South African women were classified as obese; this age group being the one closest to the ages of our study participants. Furthermore, the SADHS indicated that 23% of all women were obese with a BMI > 30kg/m² and 29% of these women were classified overweight with a BMI between 25kg/m² and 29.9kg/m². It would appear that being overweight is particularly prevalent among black women, of whom 28.4% were obese and 27.8% were overweight. It is important to note, however, that the HIV sero-status of these women in the SADHS was unknown and it was assumed that most were healthy persons. In our study, 48.5% of women were considered obese by six weeks postpartum. It is possible that HIV infected women in our study were over-compensating for their HIV status by consuming a higher energy dense diet or that based on local health messaging they too had come to believe that HIV infected persons required increased intake of energy sources. Given that the prevalence of obesity increased to 65% at the end of the follow-up period, there is a need to promote consumption of a prudent diet for all persons in the South African society regardless of HIV status. The notable increase in mean BMI levels among our study group was surprising considering that the claimed median per capita monthly income in the households in which the study mothers resided was R320.00 and the Inter-quartile Range (IQR) was R345.97. There were 185 (63%) participants whose per capita income was below R431.00, the national poverty line in 2006. Given the poverty data we found, it is probable that mothers were able to consume foods or lead sedentary lifestyles which could have resulted in the higher BMI levels we observed. It is of concern that some researchers have found very few overweight or obese African women in South Africa who view themselves as being overweight and instead associate thinness with HIV and AIDS. 194
The assessment of body composition among HIV-infected persons needs to take into consideration any other co-infections that may be present.65 Whilst we did not systematically verify the illnesses that the mothers in our study had, at every visit they were asked to state any illnesses they had experienced since the last visit. Unlike other findings among HIV-infected men and women, it would appear that there was minimal co-infection in our study mothers.
An additional factor that may have influenced the trends we observed in anthropometric measurements among the study mothers could have been that some of the study participants had initiated HAART. In South Africa, HAART was introduced within the public health institutions from 1 April 2004, halfway through this study’s follow-up period. By the end of the 24 month follow-up period there were 31 women who were on HAART. It has been documented that amongst persons on HAART disturbed fat compartmentalisation and elevated CRP levels may occur.195 Others have not found fat mass changes amongst persons on HAART, but rather increased bone mass loss.90 Considering that none of our patients had been on HAART for longer than two years, it is highly unlikely that during the 24-month period of observation these metabolic changes would have been observed. Our study was also not designed to determine the levels of adherence to ARV therapy among the clients and the impact on anthropometry, so we relied on hospital records and the participants’ own recall of taking ARV therapy.
The importance of continued monitoring and assessment of nutritional parameters among HIV-infected persons has been emphasised to enable early intervention as required and to avoid more detrimental consequences of HIV- related immuno-suppression and malnutrition.196 Others have recommended that in South Africa the prevention and treatment of obesity should focus on, amongst other interventions, high level political support and community mobilisation, and behaviour change communication. Further, there should be emphasis on healthy weight goals, increasing levels of physical activity, and identification of persons at risk of obesity at the primary health care level through routine monitoring.194 Whilst our follow-up period was limited to 24 months, we observed minimal weight loss in our study cohort and, instead, we observed that the majority of the mothers enrolled fell into the overweight and obese BMI categories, which in itself raises concern and requires further monitoring to prevent the onset of non-communicable diseases of lifestyle.
DECLARATION
DEDICATION
ABSTRACT
OPSOMMING
ACKNOWLEDGMENTS
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES AND ILLUSTRATIONS
APPENDICES
TERMINOLOGY AND ACRONYMS
CHAPTER 1 – INTRODUCTION
CHAPTER 2 – BACKGROUND AND LITERATURE REVIEW
2.1 HIV AND INFANT FEEDING CHOICES
2.1.1 Mother-to-child transmission of HIV
2.1.2 Infant feeding choices and prevention of HIV transmission
2.1.3 HIV and infant feeding policy guidelines
2.1.4 Recall bias in estimating exclusivity or duration of infant feeding practices
2.1.5 Unsafe formula-feeding
2.1.6 Problems and risks in the PMTCT programme
2.1.7 Practical considerations in the application of HIV transmission and infant feeding guidelines
2.1.8 Practical issues associated with early cessation of breastfeeding
2.1.9 Maternal viral load and mastitis as risk factors for HIV transmission
2.1.10 Breastfeeding and maternal outcome
2.1.11 Infant Feeding practices in South Africa
2.1.12 Summary
2.2 NUTRITIONAL STATUS AND HIV INFECTION AMONG WOMEN
2.2.1 Interactions between Nutrition and HIV infection
2.2.2 Body composition and HIV infection
2.2.2.1 Prenatal and postnatal body composition trends among HIV-infected women
2.2.2.2 Body composition trends, survival and initiation of HAART
2.2.2.3 Summary
2.2.3 Maternal micronutrient status and HIV infection
2.2.3.1 Mechanisms by which HIV infection impacts on blood micronutrient levels
2.2.3.2.The role of vitamins and minerals on pregnancy outcome and MTCT of HIV
2.2.3.3 Micronutrient deficiencies and HIV disease progression
2.2.3.4 Dietary micronutrient intake, blood micronutrient levels and HIV disease
2.2.3.5 Micronutrient levels and initiation of HAART
2.2.3.6 Summary
2.3 MATERNAL HEALTH, HIV AND GROWTH OF HIV-EXPOSED CHILDREN
2.3.1 Overview of maternal HIV infection and nutritional status on child outcomes
2.3.2 Trends in child mortality and HIV prevalence
2.3.3 Child growth, morbidity, mortality and HIV infection
2.3.4 Maternal caring capacity, psychosocial wellbeing and child growth
2.3.5 Summary
CHAPTER 3 – SCOPE OF RESEARCH AND HYPOTHESIS/PROBLEM STATEMENT
3.1 INTRODUCTION
3.2 SCOPE OF RESEARCH
3.3 RESEARCH QUESTIONS
CHAPTER 4 – PARTICIPANTS AND METHODS
4.1 INTRODUCTION
4.2 METHODS
4.2.1 Socio-demographic information
4.2.2 Anthropometric measurements
4.2.3 Infant feeding assessment
4.2.4 Clinical assessment
4.2.5 Nutritional biomarkers and immunological assessment
4.2.6 HIV transmission assessment
4.2.7 Measures of Psychosocial well-being
4.2.7.1 Disclosure
4.2.7.2 Stigma
4.2.7.3 Depression
4.2.7.4Coping
4.3 STATISTICAL ANALYSES
4.4 ETHICAL CONSIDERATIONS
5.1 OBJECTIVES.
5.2 SUBJECTS AND METHODS
5.3 STATISTICAL ANALYSIS
5.4 RESULTS
5.5 DISCUSSION
5.6 SUMMARY
CHAPTER 6 – ANTHROPOMETRIC MEASUREMENTS AMONG HIV- INFECTED WOMEN OVER A 24 MONTH PERIOD
6.1 OBJECTIVES
6.2 SUBJECTS AND METHODS
6.3 STATISTICAL ANALYSES
6.4 RESULTS
6.5 DISCUSSION
6.6 SUMMARY
CHAPTER 7 – MICRONUTRIENT STATUS AMONG HIV-INFECTED MOTHERS IN TSHWANE, 2003-2005
7.1 OBJECTIVES
7.2 SUBJECTS AND METHODS
7.4 RESULTS
7.5 DISCUSSION
7.6 SUMMARY
CHAPTER 8 – CHILD OUTCOMES IN RELATION TO MATERNAL HEALTH
8.1 OBJECTIVES
8.2 SUBJECTS AND METHODS.
8.3 STATISTICAL ANALYSES
8.4 RESULTS
8.5 DISCUSSION
8.7 SUMMARY
CHAPTER 9 – CONCLUSIONS AND RECOMMENDATIONS
9.1 CONCLUSIONS
9.2 RECOMMENDATIONS
REFERENCES
