Growth velocity assessment in paediatric AIDS: smoothing, penalized quantile regression and the definition of growth failure

Growth trajectories for executive and social cognitive abilities in an Indian population sample: Impact of demographic and psychosocial determinants

Cognitive abilities are markers of brain development and psychopathology. Abilities, across executive, and social domains need better characterization over development, including factors that influence developmental change. This study is based on the cVEDA [Consortium on Vulnerability to Externalizing Disorders and Addictions] study, an Indian population based developmental cohort. Verbal working memory, visuo-spatial working memory, response inhibition, set-shifting, and social cognition (faux pas recognition and emotion recognition) were cross-sectionally assessed in > 8000 individuals over the ages 6-23 years. There was adequate representation across sex, urban-rural background, psychosocial risk (psychopathology, childhood adversity and wealth index, i.e. socio-economic status). Quantile regression was used to model developmental change. Age-based trajectories were generated, along with examination of the impact of determinants (sex, childhood adversity, and wealth index). Development in both executive and social cognitive abilities continued into adulthood. Maturation and stabilization occurred in increasing order of complexity, from working memory to inhibitory control to cognitive flexibility. Age related change was more pronounced for low quantiles in response inhibition (β∼4 versus  -1 versus -0.25 for lower quantiles). Wealth index had the largest influence on developmental change across cognitive abilities. Sex differences were prominent in response inhibition, set-shifting and emotion recognition. Childhood adversity had a negative influence on cognitive development. These findings add to the limited literature on patterns and determinants of cognitive development. They have implications for understanding developmental vulnerabilities in young persons, and the need for providing conducive socio-economic environments.

Dynamic child growth prediction: A comparative methods approach

Abstract:

We introduce a class of dynamic regression models designed to predict the future of growth curves based on their historical dynamics. This class of models incorporates both baseline and time-dependent covariates, start with simple regression models and build up to dynamic function-on-function regressions. We compare the performance of the dynamic prediction models in a variety of signal-to-noise scenarios and provide practical solutions for model selection. We conclude that (a) prediction performance increases substantially when using the entire growth history relative to using only the last and first observation; (b) smoothing incorporated using functional regression approaches increases prediction performance; and (c) the interpretation of model parameters is substantially improved using functional regression approaches. Because many growth curve datasets exhibit missing and noisy data, we propose a bootstrap of subjects approach to account for the variability associated with the missing data imputation and smoothing. Methods are motivated by and applied to the CONTENT dataset, a study that collected monthly child growth data on 197 children from birth until month 15. R code describing the fitting approaches is provided in a supplementary file.

Modelling the distribution of health-related quality of life of advanced melanoma patients in a longitudinal multi-centre clinical trial using M-quantile random effects regression

Health-related quality of life assessment is important in the clinical evaluation of patients with metastatic disease that may offer useful information in understanding the clinical effectiveness of a treatment. To assess if a set of explicative variables impacts on the health-related quality of life, regression models are routinely adopted. However, the interest of researchers may be focussed on modelling other parts (e.g. quantiles) of this conditional distribution. In this paper, we present an approach based on quantile and M-quantile regression to achieve this goal. We applied the methodologies to a prospective, randomized, multi-centre clinical trial. In order to take into account the hierarchical nature of the data we extended the M-quantile regression model to a three-level random effects specification and estimated it by maximum likelihood.

CD4+ gain percentile curves for monitoring response to antiretroviral therapy in HIV-infected adults

OBJECTIVES

We constructed CD4 cell count gain percentile distributions standardized by baseline CD4 cell count and assessed the association between poor CD4 cell count gain and subsequent death and virological failure on antiretroviral therapy (ART).

DESIGN

Secondary analysis of 10 years of clinical data from a cohort of adults initiated on ART at the Themba Lethu clinic in Johannesburg, South Africa.

METHODS

The generalized additive model for location, scale and shape was used to construct percentile curves for CD4 cell count gain standardized by baseline CD4 cell count in the first 28 months of ART. Cox proportional models were used to assess the association between lower percentiles (<50th) of CD4 cell count gain, and subsequent death and virological failure.

RESULTS

Among 9640 nonpregnant adults 7406, with available CD4 cell count results for CD4 gain calculation at 4 months of ART, 843 (8.7%) died subsequently and 1101 (11.4%) experienced virologic failure, respectively. For CD4 gains below the third percentile, the adjusted hazard ratios at different time points ranged between 2.72 and 5.73 for death, and between 1.48 and 6.93 for virologic failure. The CD4 percentile curves revealed a gradient of increasing risk of subsequent death and virological failure, with lower CD4 gain percentiles and increasing time on ART, and were more informative than the WHO criteria for immunological failure or current CD4 cell count.

CONCLUSION

Percentile curves of CD4 cell count gain provide a simple tool for healthcare workers in low-resource settings to monitor response to ART with improved information regarding risk of death and virological failure compared to current WHO criteria for immunological failure.

Age-specific and sex-specific weight gain norms to monitor antiretroviral therapy in children in low-income and middle-income countries

BACKGROUND

Viral load and CD4% are often not available in resource-limited settings for monitoring children's responses to antiretroviral therapy (ART). We aimed to construct normative curves for weight gain at 6, 12, 18, and 24 months following initiation of ART in children, and to assess the association between poor weight gain and subsequent responses to ART.

DESIGN

Analysis of data from HIV-infected children younger than 10 years old from African and Asian clinics participating in the International epidemiologic Databases to Evaluate AIDS.

METHODS

The generalized additive model for location, scale, and shape was used to construct normative percentile curves for weight gain at 6, 12, 18, and 24 months following ART initiation. Cox proportional models were used to assess the association between lower percentiles (< 50th) of weight gain distribution at the different time points and subsequent death, virological suppression, and virological failure.

RESULTS

Among 7173 children from five regions of the world, 45% were underweight at baseline. Weight gain below the 50th percentile at 6, 12, 18, and 24 months of ART was associated with increased risk of death, independent of baseline characteristics. Poor weight gain was not associated with increased hazards of virological suppression or virological failure.

CONCLUSION

Monitoring weight gain on ART using age-specific and sex-specific normative curves specifically developed for HIV-infected children on ART is a simple, rapid, sustainable tool that can aid in the identification of children who are at increased risk of death in the first year of ART.

Effect of probiotic bacteria on microbial host defense, growth, and immune function in human immunodeficiency virus type-1 infection

The hypothesis that probiotic administration protects the gut surface and could delay progression of Human Immunodeficiency Virus type1 (HIV-1) infection to the Acquired Immunodeficiency Syndrome (AIDS) was proposed in 1995. Over the last five years, new studies have clarified the significance of HIV-1 infection of the gut associated lymphoid tissue (GALT) for subsequent alterations in the microflora and breakdown of the gut mucosal barrier leading to pathogenesis and development of AIDS. Current studies show that loss of gut CD4+ Th17 cells, which differentiate in response to normal microflora, occurs early in HIV-1 disease. Microbial translocation and suppression of the T regulatory (Treg) cell response is associated with chronic immune activation and inflammation. Combinations of probiotic bacteria which upregulate Treg activation have shown promise in suppressing pro inflammatory immune response in models of autoimmunity including inflammatory bowel disease and provide a rationale for use of probiotics in HIV-1/AIDS. Disturbance of the microbiota early in HIV-1 infection leads to greater dominance of potential pathogens, reducing levels of bifidobacteria and lactobacillus species and increasing mucosal inflammation. The interaction of chronic or recurrent infections, and immune activation contributes to nutritional deficiencies that have lasting consequences especially in the HIV-1 infected child. While effective anti-retroviral therapy (ART) has enhanced survival, wasting is still an independent predictor of survival and a major presenting symptom. Congenital exposure to HIV-1 is a risk factor for growth delay in both infected and non-infected infants. Nutritional intervention after 6 months of age appears to be largely ineffective. A meta analysis of randomized, controlled clinical trials of infant formulae supplemented with Bifidobacterium lactis showed that weight gain was significantly greater in infants who received B. lactis compared to formula alone. Pilot studies have shown that probiotic bacteria given as a supplement have improved growth and protected against loss of CD4+ T cells. The recognition that normal bacterial flora prime neonatal immune response and that abnormal flora have a profound impact on metabolism has generated insight into potential mechanisms of gut dysfunction in many settings including HIV-1 infection. As discussed here, current and emerging studies support the concept that probiotic bacteria can provide specific benefit in HIV-1 infection. Probiotic bacteria have proven active against bacterial vaginosis in HIV-1 positive women and have enhanced growth in infants with congenital HIV-1 infection. Probiotic bacteria may stabilize CD4+ T cell numbers in HIV-1 infected children and are likely to have protective effects against inflammation and chronic immune activation of the gastrointestinal immune system.

Longitudinal anthropometric assessment of infants born to HIV-1-infected mothers, Belo Horizonte, Southeastern Brazil

OBJECTIVE: To evaluate the growth parameters in infants who were born to HIV-1-infected mothers. METHODS: The study was a longitudinal evaluation of the z-scores for the weight-for-age (WAZ), weight-for-length (WLZ) and length-for-age (LAZ) data collected from a cohort. A total of 97 non-infected and 33 HIV-infected infants born to HIV-1-infected mothers in Belo Horizonte, Southeastern Brazil, between 1995 and 2003 was studied. The average follow-up period for the infected and non-infected children was 15.8 months (variation: 6.8 to 18.0 months) and 14.3 months (variation: 6.3 to 18.6 months), respectively. A mixed-effects linear regression model was used and was fitted using a restricted maximum likelihood. RESULTS: There was an observed decrease over time in the WAZ, LAZ and WLZ among the infected infants. At six months of age, the mean differences in the WAZ, LAZ and WLZ between the HIV-infected and non-infected infants were 1.02, 0.59, and 0.63 standard deviations, respectively. At 12 months, the mean differences in the WAZ, LAZ and WLZ between the HIV-infected and non-infected infants were 1.15, 1.01, and 0.87 standard deviations, respectively. CONCLUSIONS: The precocious and increasing deterioration of the HIV-infected infants' anthropometric indicators demonstrates the importance of the early identification of HIV-infected infants who are at nutritional risk and the importance of the continuous assessment of nutritional interventions for these infants.

Six-month gain in weight, height, and CD4 predict subsequent antiretroviral treatment responses in HIV-infected South African children

OBJECTIVES

Construct percentile curves for 6-month gain in weight, height, CD4 cell count, and CD4 percentage (CD4%) in children initiating ART, and to assess the association between lower percentiles and subsequent ART responses.

DESIGN

Cohort of 1394 HIV-infected children initiating ART between April 2004 and March 2008, Johannesburg, South Africa

METHODS

The generalized additive model for location, scale, and shape was used to construct percentile curves for 6-month gain in weight, height, CD4 cell count, and CD4%. Cox proportional models were used to assess the association between lower percentiles of each distribution and death, virological suppression, and treatment failure between 6 to 36 months post-ART initiation.

RESULTS

Lower percentiles for gain in weight, CD4, and CD4% count after 6 months of ART, but not height, were associated with poor subsequent treatment outcomes independent of baseline characteristics, with increasing strength of association as percentiles decreased. Age-specific 6-month post-ART weight gain in our cohort was substantially higher compared with 6-month weight gain in non-HIV-infected American children of the Fels Institute cohort and the attained weight-for-age at 6 months post-ART plotted on WHO weight-for-age growth charts were not associated with subsequent treatment outcomes.

CONCLUSION

Gain in CD4% in the first 6 months of ART was the best predictor of poor subsequent ART outcomes. In areas with limited access to CD4%, weight gain post-ART using our newly developed reference distributions for HIV-infected children on ART is a good alternative to CD4%, and clearly superior to the commonly used 'Road-to-Health' weight-for-age charts.

Modeling and prediction of children's growth data via functional principal component analysis

We use the functional principal component analysis (FPCA) to model and predict the weight growth in children. In particular, we examine how the approach can help discern growth patterns of underweight children relative to their normal counterparts, and whether a commonly used transformation to normality plays any constructive roles in a predictive model based on the FPCA. Our work supplements the conditional growth charts developed by Wei and He (2006) by constructing a predictive growth model based on a small number of principal components scores on individual's past.

Longitudinal growth of infants born to HIV-1-infected mothers in Belo Horizonte, Brazil

Objective

To prospectively evaluate growth parameters assessed by weight and length in infected and uninfected infants born to HIV-1-infected mothers and followed from birth to 18 months.

Methods

A cohort consisting of ninety-seven uninfected and forty-two infected infants born to HIV-infected mothers enrolled from 1995 to 2004, and admitted during their first 3 months of life at a referral Pediatric AIDS Clinic in Belo Horizonte, Brazil. Infants were followed until 18 months of age. Data were analysed using mixed-effects linear regression models for weight and length fitted by restricted maximum likelihood.

Results

Infected infants contributed to 466 weight and 411 recumbent length measurements. Uninfected infants provided 924 weight and 907 length measurements. Mean birth weight and length were similar in both groups, 3·1 (sd 0·4) and 3·0 (sd 0·5) kg, and 48·7 (sd 1·4) and 48·8 (sd 2·9) cm for uninfected and infected infants, respectively. However, HIV-1 infection had an early impact in growth impairment: at 6 months of age, HIV-infected children were 1 kg lighter and 2 cm shorter than the uninfected.

Conclusions

Growth faltering in weight, but not length, in HIV-infected children in Brazil is more marked than that reported in a European cohort, probably reflecting background nutritional deficiencies and concomitant infections. In these settings, early and aggressive nutritional management in HIV-1-infected infants should be a priority intervention associated with the antiretroviral therapy.

Failure to thrive: still a problem of definition

The term 'failure to thrive' (FTT) is widely used to describe inadequate growth in early childhood. However, no consensus exists concerning the specific anthropometrical criteria to define this description. The aim of this study was to make an updated assessment concerning the use of FTT definitions and describe possible trends regarding the use of specific criteria. A cross-sectional review was done covering English-language articles published from January 2003 until June 2004, and recent textbooks of general pediatrics. Most of the reviewed literature broadly defined FTT as inadequate growth and total agreement existed to define FTT based solely on anthropometrical parameters. Large differences, however, were seen regarding which growth parameters to use and whether to use attained values or velocities. Weight was the most predominant choice, but many included more than one anthropometrical parameter. Failure to thrive in children is currently described solely based on anthropometrical indicators, with weight gain as the predominant choice of indicator and cut off around the 5th percentile. Discussion is needed as to whether the term 'failure to thrive' is still a useful common term for pediatric undernutrition of different types.

Quantile regression methods for reference growth charts

Estimation of reference growth curves for children's height and weight has traditionally relied on normal theory to construct families of quantile curves based on samples from the reference population. Age-specific parametric transformation has been used to significantly broaden the applicability of these normal theory methods. Non-parametric quantile regression methods offer a complementary strategy for estimating conditional quantile functions. We compare estimated reference curves for height using the penalized likelihood approach of Cole and Green with quantile regression curves based on data used for modern Finnish reference charts. An advantage of the quantile regression approach is that it is relatively easy to incorporate prior growth and other covariates into the analysis of longitudinal growth data. Quantile specific autoregressive models for unequally spaced measurements are introduced and their application to diagnostic screening is illustrated.

Do growth monitoring and promotion programs answer the performance criteria of a screening program? A critical analysis based on a systematic review

OBJECTIVE

Growth Monitoring and Promotion programs (GMP) have been intensively promoted to improve children's health in developing countries. It has been hoped that regularly weighing children would result in the early detection of growth falterers, and that the growth chart would serve as an educational tool to make that state apparent to both health workers and caretakers in order to trigger improved caring practices. Our objective was to review whether GMP answers the theoretical grounds of a screening and intervention program.

METHOD

A systematic literature review was performed. The WHO framework developed by Wilson and Jungner for planning and evaluating screening programs guided the analysis.

RESULTS

Sixty-nine studies were retrieved. Overall, evidence is weak on the performance of GMP as a screening program for malnutrition through early detection of growth falterers. The main results are: (1) malnutrition remains a public health problem, but its importance is context specific; (2) the value of a low weight velocity to predict malnutrition is unknown and likely to vary in different contexts; (3) the performance of GMP for improving nutrition status of children and in reducing mortality and morbidity is unknown; (4) the performance of the screening is affected by the unreliability of weight measurements; (5) the promotional and educational effectiveness of GMP is low, in particular the growth chart is poorly understood by mothers; (6) the acceptability seems low in regards of low attendance rates; (7) evidence is lacking regarding cost-effectiveness.

CONCLUSIONS

We conclude that there is too little scientific evidence to indiscriminately support international promotion of GMP. However GMP could constitute a valid strategy of public nutrition in specific situations. We indicate paths for further research and how prevention programs could be developed.