Predicting the drop out from the maternal, newborn and child healthcare continuum in three East African Community countries: application of machine learning models.

BACKGROUND: For optimal health, the maternal, newborn, and child healthcare (MNCH) continuum necessitates that the mother/child receive the full package of antenatal, intrapartum, and postnatal care. In sub-Saharan Africa, dropping out from the MNCH continuum remains a challenge. Using machine learning, the study sought to forecast the MNCH continuum drop out and determine important predictors in three East African Community (EAC) countries. METHODS: The study utilised Demographic Health Surveys data from the Democratic Republic of Congo (DRC) (2013/14), Kenya (2014) and Tanzania (2015/16). STATA 17 was used to perform the multivariate logistic regression. Python 3.0 was used to build five machine learning classification models namely the Logistic Regression, Random Forest, Decision Tree, Support Vector Machine and Artificial Neural Network. Performance of the models was assessed using Accuracy, Precision, Recall, Specificity, F1 score and area under the Receiver Operating Characteristics (AUROC). RESULTS: The prevalence of the drop out from the MNCH continuum was 91.0% in the DRC, 72.4% in Kenya and 93.6% in Tanzania. Living in the rural areas significantly increased the odds of dropping out from the MNCH continuum in the DRC (AOR:1.76;95%CI:1.30-2.38), Kenya (AOR:1.23;95%CI:1.03-1.47) and Tanzania (AOR:1.41;95%CI:1.01-1.97). Lower maternal education also conferred a significant increase in the DRC (AOR:2.16;95%CI:1.67-2.79), Kenya (AOR:1.56;95%CI:1.30-1.84) and Tanzania (AOR:1.70;95%CI:1.24-2.34). Non exposure to mass media also conferred a significant positive influence in the DRC (AOR:1.49;95%CI:1.15-1.95), Kenya (AOR:1.46;95%CI:1.19-1.80) and Tanzania (AOR:1.65;95%CI:1.13-2.40). The Random Forest exhibited superior predictive accuracy (Accuracy = 75.7%, Precision = 79.1%, Recall = 92.1%, Specificity = 51.6%, F1 score = 85.1%, AUROC = 70%). The top four predictors with the greatest influence were household wealth, place of residence, maternal education and exposure to mass media. CONCLUSIONS: The MNCH continuum dropout rate is very high in the EAC countries. Maternal education, place of residence, and mass media exposure were common contributing factors to the drop out from MNCH continuum. The Random Forest had the highest predictive accuracy. Household wealth, place of residence, maternal education and exposure to mass media were ranked among the top four features with significant influence. The findings of this study can be used to support evidence-based decisions in MNCH interventions and to develop web-based services to improve continuity of care retention.

Malaria attributable fractions with changing transmission intensity: Bayesian latent class vs logistic models.

BACKGROUND: Asymptomatic carriage of malaria parasites is common in high transmission intensity areas and confounds clinical case definitions for research studies. This is important for investigations that aim to identify immune correlates of protection from clinical malaria. The proportion of fevers attributable to malaria parasites is widely used to define different thresholds of parasite density associated with febrile episodes. The varying intensity of malaria transmission was investigated to check whether it had a significant impact on the parasite density thresholds. The same dataset was used to explore an alternative statistical approach, using the probability of developing fevers as a choice over threshold cut-offs. The former has been reported to increase predictive power. METHODS: Data from children monitored longitudinally between 2005 and 2017 from Junju and Chonyi in Kilifi, Kenya were used. Performance comparison of Bayesian-latent class and logistic power models in estimating malaria attributable fractions and probabilities of having fever given a parasite density with changing malaria transmission intensity was done using Junju cohort. Zero-inflated beta regressions were used to assess the impact of using probabilities to evaluate anti-merozoite antibodies as correlates of protection, compared with multilevel binary regression using data from Chonyi and Junju. RESULTS: Malaria transmission intensity declined from over 49% to 5% between 2006 and 2017, respectively. During this period, malaria attributable fraction varied between 27-59% using logistic regression compared to 10-36% with the Bayesian latent class approach. Both models estimated similar patterns of fevers attributable to malaria with changing transmission intensities. The Bayesian latent class model performed well in estimating the probabilities of having fever, while the latter was efficient in determining the parasite density threshold. However, compared to the logistic power model, the Bayesian algorithm yielded lower estimates for both attributable fractions and probabilities of fever. In modelling the association of merozoite antibodies and clinical malaria, both approaches resulted in comparable estimates, but the utilization of probabilities had a better statistical fit. CONCLUSIONS: Malaria attributable fractions, varied with an overall decline in the malaria transmission intensity in this setting but did not significantly impact the outcomes of analyses aimed at identifying immune correlates of protection. These data confirm the statistical advantage of using probabilities over binary data.

Markov modelling of viral load adjusting for CD4 orthogonal variable and multivariate conditional autoregressive mapping of the HIV immunological outcomes among ART patients in Zimbabwe.

BACKGROUND: This study aimed to jointly model HIV disease progression patterns based on viral load (VL) among adult ART patients adjusting for the time-varying "incremental transients states" variable, and the CD4 cell counts orthogonal variable in a single 5-stage time-homogenous multistate Markov model. We further jointly mapped the relative risks of HIV disease progression outcomes (detectable VL (VL ≥ 50copies/uL) and immune deterioration (CD4 < 350cells/uL) at the last observed visit) conditional not to have died or become loss to follow-up (LTFU). METHODS: Secondary data analysis of individual-level patients on ART was performed. Adjusted transition intensities, hazard ratios (HR) and regression coefficients were estimated from the joint multistate model of VL and CD4 cell counts. The mortality and LTFU transition rates defined the extent of patients' retention in care. Joint mapping of HIV disease progression outcomes after ART initiation was done using the Bayesian intrinsic Multivariate Conditional Autoregressive prior model. RESULTS: The viral rebound from the undetectable state was 1.78times more likely compared to viral suppression among patients with VL ranging from 50-1000copies/uL. Patients with CD4 cell counts lower than expected had a higher risk of viral increase above 1000copies/uL and death if their VL was above 1000copies/uL (state 2 to 3 (λ23): HR = 1.83 and (λ34): HR = 1.42 respectively). Regarding the time-varying effects of CD4 cell counts on the VL transition rates, as the VL increased, (λ12 and λ23) the transition rates increased with a decrease in the CD4 cell counts over time. Regardless of the individual's VL, the transition rates to become LTFU decreased with a decrease in CD4 cell counts. We observed a strong shared geographical pattern of 66% spatial correlation between the relative risks of detectable VL and immune deterioration after ART initiation, mainly in Matabeleland North. CONCLUSION: With high rates of viral rebound, interventions which encourage ART adherence and continual educational support on the barriers to ART uptake are crucial to achieve and sustain viral suppression to undetectable levels. Area-specific interventions which focus on early ART screening through self-testing, behavioural change campaigns and social support strategies should be strengthened in heavily burdened regions to sustain the undetectable VL. Sustaining undetectable VL lowers HIV transmission in the general population and this is a step towards achieving zero HIV incidences by 2030.

The relationship between childhood trauma and schizophrenia in the Genomics of Schizophrenia in the Xhosa people (SAX) study in South Africa.

BACKGROUND: Evidence from high-income countries suggests that childhood trauma is associated with schizophrenia. Studies of childhood trauma and schizophrenia in low and middle income (LMIC) countries are limited. This study examined the prevalence of childhood traumatic experiences among cases and controls and the relationship between specific and cumulative childhood traumatic experiences and schizophrenia in a sample in South Africa. METHODS: Data were from the Genomics of Schizophrenia in the South African Xhosa people study. Cases with schizophrenia and matched controls were recruited from provincial hospitals and clinics in the Western and Eastern Cape regions in South Africa. Childhood traumatic experiences were measured using the Childhood Trauma Questionnaire (CTQ). Adjusted logistic regression models estimated associations between individual and cumulative childhood traumatic experiences and schizophrenia. RESULTS: Traumatic experiences were more prevalent among cases than controls. The odds of schizophrenia were 2.44 times higher among those who experienced any trauma than those who reported no traumatic experiences (95% CI 1.77-3.37). The odds of schizophrenia were elevated among those who experienced physical/emotional abuse (OR 1.59, CI 1.28-1.97), neglect (OR 1.39, CI 1.16-1.68), and sexual abuse (OR 1.22, CI 1.03-1.45) compared to those who did not. Cumulative physical/emotional abuse and neglect experiences increased the odds of schizophrenia as a dose-response relationship. CONCLUSION: Childhood trauma is common in this population. Among many other benefits, interventions to prevent childhood trauma may contribute to a decreasing occurrence of schizophrenia.

The effect of a home exercise intervention on persons with lower limb amputations: a randomized controlled trial.

OBJECTIVE: To establish if a home-based exercise and education programme is more effective than usual treatment in improving function, mobility and quality of life in people with lower limb amputation due to peripheral vascular disease. METHOD: A blocked randomized single-blinded controlled trial (RCT) with 154 participants (54 female; mean age 58) compared a home-based exercise and education programme (n = 77) with usual care (control) (n = 77). Participants were measured at baseline, immediately post intervention at three months, and after a further three months without any intervention. The outcome measures were the Barthel Index, Participation Scale, EuroQuol 5D, Modified Locomotor Capability Index and Timed Up and Go Test. Changes over time were established using generalized estimating equations and analysis of covariance, (P < 0.05). RESULTS: The Participation Scale (18.73 ± 14.91 against 26.67 ± 19.14; P = 0.011), the EuroQuol5D visual analogue scale (69.10 ± 20.31 against 55.37 ± 27.67; P = 0.003), EuroQuol5D utility index (0.672 ± 0.300 against 0.532 ± 0.358; P = 0.25) and the Modified Capability Index (21.03 ± 15.79 against 15.91 ± 13.67; P = 0.034) improved in the intervention group compared with the control group at three months. At six months, there was no difference between the groups in any of the measures except for the EuroQuol5D visual analogue scale (74.52 ± 16.14 against 66.34 ± 22.91; P = 0.033). Although there were no differences between the Timed Up and Go test at 3 (35.39 ± 32.48 against 45.08 ± 41.52; P = 0.192) and six months (28.22 ± 20.96 against 36.08 ± 36.19; P = 0.189) between the groups the intervention group was more mobile. CONCLUSION: This intervention improved function, mobility and quality of life in persons following lower limb amputation in the first three months post amputation.

Four decades of pulmonary tuberculosis in deceased South African miners: trends and determinants.

OBJECTIVE: We investigated trends and determinants of pulmonary tuberculosis (PTB) in deceased South African miners. METHODS: Statutory autopsies are performed on miners for occupational lung disease compensation, irrespective of cause of death. Data were extracted from the PATHAUT (Pathology Automation System) autopsy database. PTB trends were analysed and explanatory variables (year of autopsy, age at death, gold employment duration, silicosis and HIV) were evaluated using binary logistic regression modelling. Analyses were stratified by population group because of racial differences in socioeconomic status, employment patterns and access to facilities for autopsies. The analyses were segmented to represent the pre-HIV (1975-1989), rapid HIV spread (1990-2004) and antiretroviral therapy (2005-2014) periods. RESULTS: The proportions of men with PTB at autopsy increased from 4.62% in 1975 to 27.18% in 2014 in black miners, and from 2.07% to 5.19% in white miners, with peaks in 2007 (43.12% and 9.51%, respectively). The magnitude and significance of adjusted ORs of determinants differed by population group and calendar period. PTB was largely associated with silicosis, increasing gold employment duration and year of autopsy (a surrogate for unmeasured confounders, such as unknown HIV status and tuberculosis transmission). CONCLUSIONS: Changes in PTB time trends and determinants reflect the complex social and political environment in which mining occurs. Silica dust reduction remains a key intervention for tuberculosis reduction, together with tuberculosis and HIV treatment and management. The autopsy data provide reliable information to monitor progress towards the achievement of industry and national targets to reduce tuberculosis.

The effects of progressive-resisted exercises on muscle strength and health-related quality of life in persons with HIV-related poly-neuropathy in Zimbabwe.

Distal symmetrical poly-neuropathy (DSP) is a neurological complication associated with HIV/AIDS and stavudine (d4T) containing antiretroviral therapy. People with DSP experience pain, numbness and muscle weakness, which affect their quality of life (QOL). The purpose of this study was to establish the effect of a progressive-resisted exercise (PRE) intervention on health-related quality of life (HR-QOL) in people living with HIV/AIDS-related DSP. An assessor-blinded randomised controlled trial was conducted, with participants sourced from 10 clinics with HIV services, the family care clinic at Wilkins Hospital and 2 large hospitals in Harare, Zimbabwe. A 12-week PRE intervention was conducted twice weekly for 80 participants, while the control group with 80 participants continued with usual daily activities. The main outcome variable was HR-QOL for which we controlled for demographic and clinical measures in generalised estimating equation population-averaged models. Data were summarised and analysed using an intention to treat analysis approach using the Stata v10 program. Mean age of participants was 42.2 years (SD = 8.5). While d4T was used by 59% (n = 94), an equal proportion of the participants also had moderate to severe neuropathy. PRE was found to significantly improve HR-QOL in the intervention group based on the mean difference between the intervention group mean change and the mean change in the control group (F ratio 4.24; p = .04). This study established that PREs have positive effects on HR-QOL for people living with HIV/AIDS-related DSP.

Prevalence of latent tuberculosis infection and predictive factors in an urban informal settlement in Johannesburg, South Africa: a cross-sectional study.

BACKGROUND: South Africa has one of the highest burdens of latent tuberculosis infection (LTBI) in high-risk populations such as young children, adolescents, household contacts of TB cases, people living with HIV, gold miners and health care workers, but little is known about the burden of LTBI in its general population. METHODS: Using a community-based survey with random sampling, we examined the burden of LTBI in an urban township of Johannesburg and investigated factors associated with LTBI. The outcome of LTBI was based on TST positivity, with a TST considered positive if the induration was ≥5 mm in people living with HIV or ≥10 mm in those with unknown or HIV negative status. We used bivariate and multivariable logistic regression to identify factors associated with LTBI RESULTS: The overall prevalence of LTBI was 34.3 (95 % CI 30.0, 38.8 %), the annual risk of infection among children age 0-14 years was 3.1 % (95 % CI 2.1, 5.2). LTBI was not associated with HIV status. In multivariable logistic regression analysis, LTBI was associated with age (OR = 1.03 for every year increase in age, 95 % CI = 1.01-1.05), male gender (OR = 2.70, 95 % CI = 1.55-4.70), marital status (OR = 2.00, 95 % CI = 1.31-3.54), and higher socio-economic status (OR = 2.11, 95 % CI = 1.04-4.31). CONCLUSIONS: The prevalence of LTBI and the annual risk of infection with M. tuberculosis is high in urban populations, especially in men, but independent of HIV infection status. This study suggests that LTBI may be associated with higher SES, in contrast to the well-established association between TB disease and poverty.

Direct and indirect determinants of childhood malaria morbidity in Malawi: a survey cross-sectional analysis based on malaria indicator survey data for 2012.

BACKGROUND: Children under the age of five are most vulnerable to malaria (malaria is a major health challenge in sub-Saharan Africa) with a child dying every 30 s from malaria. Hampered socio-economic development, poverty, diseconomies of scale, marginalization, and exploitation are associated with malaria. Therefore establishing determinants of malaria in affected sub-Saharan populations is important in order to come up with informed interventions that will be effective in malaria control. METHODS: The study was a cross-sectional survey design based on data from the Malawi 2012 Malaria indicator Survey obtained from Demographic and Health Survey (DHS) programme website. The outcome variable was positive laboratory-based blood smear result for malaria in children less than 5 years, after an initial positive rapid malaria diagnostic test done at the homestead. Statistical modelling was done using survey logistic regression as well as generalized structural equation modelling (G-SEM) to analyse direct and indirect effects of malaria. RESULTS: The propensity score matched data had 1 325 children with 367 (27.7%) having blood smear positive malaria. Female children made up approximately 53% of the total study participants. Child related variables (age, haemoglobin and position in household) and household wealth index were significant directly and indirectly. Further on G-SEM based multivariable analysis showed socio-economic status (SES) [Odds ratio (OR) = 0.96, 95% Confidence interval (CI) = 0.92, 0.99] and primary level of education [OR = 0.50, 95% CI = 0.32, 0.77] were important direct and indirect determinants of malaria morbidity. CONCLUSION: Socio-economic status and education are important factors that influence malaria control. These factors need to be taken into consideration when planning malaria control programmes in order to have effective programmes. Direct and indirect effect modelling can also provide an alternative modelling technique that incorporates surrogate confounders that may not be significant when modelled directly. This holistic approach is useful and will help in improving malaria control.

Circulating anti-citrullinated peptide antibodies, cytokines and genotype as biomarkers of response to disease-modifying antirheumatic drug therapy in early rheumatoid arthritis.

BACKGROUND: To measure circulating anti-citrullinated peptide antibodies (ACPA) and cytokines pre- and 6 months post-therapy as a strategy to predict and optimize responses to traditional disease-modifying antirheumatic drugs (DMARDs) in early RA, which is an unmet need in developing countries. PATIENTS AND METHODS: A cohort of 140 predominantly (88.5 %) black female South African patients with early RA was treated with synthetic DMARDs, mostly methotrexate (MTX) alone, or in combination with low-dose oral corticosteroids (CS). Circulating ACPA and a panel of circulating cytokines/chemokines/growth factors were measured at baseline and after 6 months of therapy in relation to disease activity and Shared Epitope (SE). RESULTS: Following 6 months of therapy, the median simplified disease activity index (SDAI) declined from a baseline of 41.4 to 16.0 (p = 0.0001) for the entire cohort, which was paralleled by significant falls in median serum ACPA levels (516.6 vs. 255.7 units/ml, p = <0.0001) and several of the circulating cytokines (IL-4, IL-7, IL-8, G-CSF, VEGF; p < 0.0010 - p < 0.0001) which were most evident in the subgroup of patients treated with a combination of MTX and CS. Although biomarker concentrations decreased most notably in the low-disease activity group post-therapy, no significant correlations between these biomarkers and disease activity were observed, Baseline ACPA levels, but not SDAI or cytokines, were significantly higher in the subgroup of risk allele-positive patients (561.1 vs. 331.9 units/ml, p < 0.05), while no associations with ACPA and a smoking history were evident. CONCLUSIONS: The use of DMARDs in RA is associated with significant decreases in ACPA and cytokines which did not correlate with changes in SDAI, precluding the utility of serial measurement of these biomarkers to monitor early responses to therapy, but may have prognostic value.

Spatial analysis of factors associated with HIV infection among young people in Uganda, 2011.

BACKGROUND: The HIV epidemic in East Africa is of public health importance with an increasing number of young people getting infected. This study sought to identify spatial clusters and examine the geographical variation of HIV infection at a regional level while accounting for risk factors associated with HIV/AIDS among young people in Uganda. METHODS: A secondary data analysis was conducted on a survey cross-sectional design whose data were obtained from the 2011 Uganda Demographic and Health Survey (DHS) and AIDS Indicator Survey (AIS) for 7 518 young people aged 15-24 years. The analysis was performed in three stages while incorporating population survey sampling weights. Maximum likelihood-based logistic regression models were used to explore the non-spatially adjusted factors associated with HIV infection. Spatial scan statistic was used to identify geographical clusters of elevated HIV infections which justified modelling using a spatial random effects model by Bayesian-based logistic regression models. RESULTS: In this study, 309/533 HIV sero-positive female participants were selected with majority residing in the rural areas [386(72%)]. Compared to singles, those currently [Adjusted Odds Ratio (AOR) =3.64; (95% CI; 1.25-10.27)] and previously married [AOR = 5.62; (95% CI: 1.52-20.75)] participants had significantly higher likelihood of HIV infections. Sexually Transmitted Infections [AOR = 2.21; (95% CI: 1.35-3.60)] were more than twice likely associated with HIV infection. One significant (p < 0.05) primary cluster of HIV prevalence around central Uganda emerged from the SaTScan cluster analysis. Spatial analysis disclosed behavioural factors associated with greater odds of HIV infection such as; alcohol use before sexual intercourse [Posterior Odds Ratio (POR) =1.32; 95% (BCI: 1.11-1.63)]. Condom use [POR = 0.54; (95% BCI: 0.41-0.69)] and circumcision [POR = 0.66; (95% BCI: 0.45-0.99)] provided a protective effect against HIV. CONCLUSIONS: The study revealed associations between high-risk sexual behaviour and HIV infection. Behavioural change interventions should therefore be pertinent to the prevention of HIV. Spatial analysis further revealed a significant HIV cluster towards the Central and Eastern areas of Uganda. We propose that interventions targeting young people should initially focus on these regions and subsequently spread out across Uganda.

Prevalence of peripheral neuropathy and its associated demographic and health status characteristics, among people on antiretroviral therapy in Rwanda.

BACKGROUND: The introduction of antiretroviral therapy (ART) has dramatically reduced the mortality rate of people living with HIV (PLHIV). However, complications of both HIV and ART, such as peripheral neuropathy currently affect PLHIV. The purpose of this study was to establish the prevalence of peripheral neuropathy of the lower extremity and, its association with demographic and health status, characteristics among people on ART in Rwanda. METHODS: A cross sectional study was conducted among 507 women and men aged between 18 and 60 years, on ART, randomly selected from eight selected ART clinics in Rwanda. Brief Peripheral Neuropathy Screen was used to assess peripheral neuropathy. RESULTS: Peripheral neuropathy prevalence was 59% overall, mean age of the participants was 39.7 (±9.2) and a slightly older age was associated with peripheral neuropathy; [42(±9.2) vs 37 (±8.8) (p < 0.001)]. 78% of participants living in urban settings compared to 40% in rural settings reported peripheral neuropathy, 69% of participants with higher levels of education (secondary level and above) reported lower extremity neuropathy.The three factors were significantly associated with peripheral neuropathy in multivariable model analysis: older age [aOR = 1.1, 95% CI (1.0, 1.2), p < 0.001], primary education level [aOR = 0.6 95% Cl (0.3, 1.0), p = 0.04] and urban setting [aOR = 0.1, 95% CI (0.06, 0.3), p < 0.001], after adjusting for other factors. None of the health status characteristics namely; the level of CD4 cell count, duration of HIV infection and duration on ART, was independently associated with peripheral neuropathy. CONCLUSIONS: The prevalence of peripheral neuropathy among PLHIV on ART in Rwanda is high. It is unclear why urban setting has an effect on PN levels in this cross sectional study, but does suggest that unidentified social and lifestyles factors may have a role in subjective symptoms and objective signs, of PN.

Comorbidities in Privately Insured South Africans With Systemic Lupus Erythematosus.

Introduction Comorbidities in systemic lupus erythematosus (SLE) impact negatively on health-related quality of life (HRQoL) and life expectancy. We investigated the frequency and spectrum of comorbidities in privately insured South Africans with SLE. Methods The data of SLE patients based on International Classification of Diseases, Tenth Revision (ICD-10) codes and insured with Discovery Health Medical Scheme (DHMS), South Africa, aged ≥16 years at diagnosis and with ≥6 months of follow-up were reviewed. Demographics, comorbidities listed in the Charlson comorbidity index (CCI), other common comorbidities, intercurrent illnesses, and drug therapy were documented. Results Of the 520 patients coded as SLE, 207 met the inclusion criteria. Most were females (90.8%), with a median (interquartile range {IQR}) age and follow-up duration of 39 (30.3-53.0) and 6.1 (3.7-8.1) years, respectively. All patients had at least one comorbidity; the most frequent CCI comorbidities were pulmonary disease (30.9%), congestive heart failure (CHF, 15%), and renal disease (14.0%). Other common comorbidities were hypertension (53.1%) and mood and anxiety disorders (46.9%). Urinary tract infections (UTIs, 37.7%) and pneumonia (33.8%) were common intercurrent illnesses. The independent predictors of CHF were renal disease (odds ratio {OR}=855), dyslipidemia (OR=15.3), and male gender (OR=43.0); the independent predictors of hypertension were age at diagnosis (OR=1.03), type 2 diabetes (OR=4.45), and renal disease (OR=4.34); and the independent predictors of mood and anxiety disorders were female gender (OR=3.98), stroke (OR=3.18), UTI (OR=2.39), and chloroquine use (OR=1.94). Conclusion In this study of privately insured South Africans with SLE, comorbidities were common, and all patients had at least one comorbidity. Hypertension, infections, and mood and anxiety disorders were the leading comorbidities overall, and pulmonary disease was the most common CCI comorbidity. There is an obvious need to formally study the burden of mental health disorders in South African SLE patients.

Investigation into longitudinal dietary behaviours and household socio-economic indicators and their association with BMI Z-score and fat mass in South African adolescents: the Birth to Twenty (Bt20) cohort.

OBJECTIVE: The present study aimed to assess the relationship between dietary habits, change in socio-economic status and BMI Z-score and fat mass in a cohort of South African adolescents. DESIGN: In the longitudinal study, data were collected at ages 13, 15 and 17 years on a birth cohort who have been followed since 1990. Black participants with complete dietary habits data (breakfast consumption during the week and at weekends, snacking while watching television, eating main meal with family, lunchbox use, number of tuck shop purchases, fast-food consumption, confectionery consumption and sweetened beverage consumption) at all three ages and body composition data at age 17 years were included in the analyses. Generalized estimating equations were used to test the associations between individual longitudinal dietary habits and obesity (denoted by BMI Z-score and fat mass) with adjustments for change in socio-economic status between birth and age 12 years. SETTING: Birth to Twenty (Bt20) study, Soweto-Johannesburg, South Africa. SUBJECTS: Adolescents (n 1298; 49·7 % male). RESULTS: In males, the multivariable analyses showed that soft drink consumption was positively associated with both BMI Z-score and fat mass (P < 0·05). Furthermore, these relationships remained the same after adjustment for socio-economic indicators (P < 0·05). No associations were found in females. CONCLUSIONS: Longitudinal soft drink consumption was associated with increased BMI Z-score and fat mass in males only. Fridge ownership at birth (a proxy for greater household disposable income in this cohort) was shown to be associated with both BMI Z-score and fat mass.

Serum matrix metalloproteinase-3 in comparison with acute phase proteins as a marker of disease activity and radiographic damage in early rheumatoid arthritis.

Matrix metalloproteinase-3 (MMP-3) is involved in the immunopathogenesis of rheumatoid arthritis (RA), but little is known about its relationship to genetic susceptibility and biomarkers of disease activity, especially acute phase reactants in early RA. MMP-3 was measured by ELISA in serum samples of 128 disease-modifying, drug-naïve patients and analysed in relation to shared epitope genotype, a range of circulating chemokines/cytokines, acute phase reactants, autoantibodies, cartilage oligomeric protein (COMP), and the simplified disease activity index (SDAI). MMP-3 was elevated >1.86 ng/ml in 56.25% of patients (P < 0.0001), correlated with several biomarkers, notably IL-8, IL-6, IFN γ , VEGF and COMP (r values = 0.22-0.33, P < 0.014-0.0001) and with CRP and SAA levels (r = 0.40 and 0.41, resp., P < 0.0000) and SDAI (r = 0.29, P < 0.0001), but not with erosions or nodulosis. However, the correlations of CRP and SAA with SDAI were stronger (respective values of 0.63 and 0.54, P < 0.001 for both). COMP correlated with smoking, RF, and MMP-3. MMP-3 is significantly associated with disease activity, inflammatory mediators and cartilage breakdown, making it a potential biomarker of disease severity, but seemingly less useful than CRP and SAA as a biomarker of disease activity in early RA.

Bayesian analysis of zero inflated spatiotemporal HIV/TB child mortality data through the INLA and SPDE approaches: Applied to data observed between 1992 and 2010 in rural North East South Africa.

Longitudinal mortality data with few deaths usually have problems of zero-inflation. This paper presents and applies two Bayesian models which cater for zero-inflation, spatial and temporal random effects. To reduce the computational burden experienced when a large number of geo-locations are treated as a Gaussian field (GF) we transformed the field to a Gaussian Markov Random Fields (GMRF) by triangulation. We then modelled the spatial random effects using the Stochastic Partial Differential Equations (SPDEs). Inference was done using a computationally efficient alternative to Markov chain Monte Carlo (MCMC) called Integrated Nested Laplace Approximation (INLA) suited for GMRF. The models were applied to data from 71,057 children aged 0 to under 10 years from rural north-east South Africa living in 15,703 households over the years 1992-2010. We found protective effects on HIV/TB mortality due to greater birth weight, older age and more antenatal clinic visits during pregnancy (adjusted RR (95% CI)): 0.73(0.53;0.99), 0.18(0.14;0.22) and 0.96(0.94;0.97) respectively. Therefore childhood HIV/TB mortality could be reduced if mothers are better catered for during pregnancy as this can reduce mother-to-child transmissions and contribute to improved birth weights. The INLA and SPDE approaches are computationally good alternatives in modelling large multilevel spatiotemporal GMRF data structures.

Estimating HIV incidence over a decade in Zimbabwe: A comparison of the catalytic and Farrington models.

Over the years, numerous modelling studies have been proposed to estimate HIV incidence. As a result, this study aimed to evaluate two alternative methods for predicting HIV incidence in Zimbabwe between 2005 and 2015. We estimated HIV incidence from seroprevalence data using the catalytic and Farrington-2-parameter models. Data were obtained from 2005-06, 2010-11, and 2015 Zimbabwe Demographic Health Survey (ZDHS). These models were validated at the micro and macro-level using community-based cohort incidence and empirical estimates from UNAIDS EPP/SPECTRUM, respectively. The HIV incidence for the catalytic model was 0.32% (CI: 0.28%, 0.36%), 0.36% (CI: 0.33%, 0.39%), and 0.28% (CI: 0.26%, 0.30%), for the years 2005-06, 2010-11, and 2015, respectively. The HIV incidence for the Farrington model was 0.21% (CI: 0.16%, 0.26%), 0.22% (CI: 0.20%, 0.25%), and 0.19% (CI: 0.16%, 0.22%), for the years 2005-06, 2010-11, and 2015, respectively. According to these findings, the catalytic model estimated a higher HIV incidence rate than the Farrington model. Compared to cohort estimates, the estimates were within the observed 95% confidence interval, with 88% and 75% agreement for the catalytic and Farrington models, respectively. The limits of agreement observed in the Bland-Altman plot were narrow for all plots, indicating that our model estimates were comparable to cohort estimates. Compared to UNAIDS estimates, the catalytic model predicted a progressive increase in HIV incidence for males throughout all survey years. Without a doubt, HIV incidence declined with each subsequent survey year for all models. To improve programmatic and policy decisions in the national HIV response, we recommend the triangulation of multiple methods for incidence estimation and interpretation of results. Multiple estimating approaches should be considered to reduce uncertainty in the estimations from various models.

Predicting HIV infection in the decade (2005-2015) pre-COVID-19 in Zimbabwe: A supervised classification-based machine learning approach.

The burden of HIV and related diseases have been areas of great concern pre and post the emergence of COVID-19 in Zimbabwe. Machine learning models have been used to predict the risk of diseases, including HIV accurately. Therefore, this paper aimed to determine common risk factors of HIV positivity in Zimbabwe between the decade 2005 to 2015. The data were from three two staged population five-yearly surveys conducted between 2005 and 2015. The outcome variable was HIV status. The prediction model was fit by adopting 80% of the data for learning/training and 20% for testing/prediction. Resampling was done using the stratified 5-fold cross-validation procedure repeatedly. Feature selection was done using Lasso regression, and the best combination of selected features was determined using Sequential Forward Floating Selection. We compared six algorithms in both sexes based on the F1 score, which is the harmonic mean of precision and recall. The overall HIV prevalence for the combined dataset was 22.5% and 15.3% for females and males, respectively. The best-performing algorithm to identify individuals with a higher likelihood of HIV infection was XGBoost, with a high F1 score of 91.4% for males and 90.1% for females based on the combined surveys. The results from the prediction model identified six common features associated with HIV, with total number of lifetime sexual partners and cohabitation duration being the most influential variables for females and males, respectively. In addition to other risk reduction techniques, machine learning may aid in identifying those who might require Pre-exposure prophylaxis, particularly women who experience intimate partner violence. Furthermore, compared to traditional statistical approaches, machine learning uncovered patterns in predicting HIV infection with comparatively reduced uncertainty and, therefore, crucial for effective decision-making.

Spatial heterogeneity in relationship between district patterns of HIV incidence and covariates in Zimbabwe: a multi-scale geographically weighted regression analysis.

A study was conducted to investigate the district-level patterns of incidence of the human immunodeficiency virus (HIV) in Zimbabwe in the period 2005-2015 and explore variations in the relationship between covariates and HIV incidence across different districts. Demographic health survey data were analysed using hotspot analysis, spatial autocorrelation, and multi-scale geographically weighted regression (MGWR) techniques. The analysis revealed hotspots of the HIV epidemic in the southern and western regions of Zimbabwe in contrast to the eastern and northern regions. Specific districts in Matabeleland South and Matabeleland North provinces showed clusters of HIV incidence in 2005-2006, 2010-2011 and 2015. Variables studied were multiple sex partners and sexually transmitted infections (STI) condom use and being married. Recommendations include implementing targeted HIV prevention programmes in identified hotspots, prioritising interventions addressing the factors mentioned above as well as enhancing access to HIV testing and treatment services in high-risk areas, strengthening surveillance systems, and conducting further research to tailor interventions based on contextual factors. The study also emphasizes the need for regular monitoring and evaluation at the district level to inform effective responses to the HIV epidemic over time. By addressing the unique challenges and risk factors in different districts, significant progress can be made in reducing HIV transmission and improving health outcomes in Zimbabwe. These findings should be valuable for policymakers in resource allocation and designing evidence-based interventions.

Population-based temporal trends and ethnic disparity in breast cancer mortality in South Africa (1999-2018): Joinpoint and age-period-cohort regression analyses.

Globally, breast cancer is the leading cause of cancer deaths, accounting for 15.5% of female cancer deaths in 2020. Breast cancer is also the leading cause of female cancers in South Africa. The rapid epidemiological transition in South Africa may have an impact on the trends in breast cancer mortality in the country. We therefore evaluated the trends in the breast cancer mortality in SA over 20 years (1999-2020). Methods: Joinpoint regression analyses of the trends in crude and age-standardized mortality rates (ASMR) of breast cancer among South African women were conducted from 1999 to 2018 using mortality data from Statistics South Africa. Age-period-cohort regression analysis was then conducted to evaluate the independent effect of age, period, and cohort on breast cancer mortality, and analysis was stratified by ethnicity. Results: The mortality rate of breast cancer (from 9.82 to 13.27 per 100,000 women) increased at around 1.4% per annum (Average Annual Percent Change (AAPC): 1.4%, 95% CI:0.8-2.0, P-value< 0.001). Young women aged 30-49 years (1.1%-1.8%, P-value< 0.001) had increased breast cancer mortality. The risk of breast cancer mortality increased among successive birth cohorts from 1924 to 1928 but decreased among recent cohorts born from 1989 to 1993. In 2018, the breast cancer mortality rate among Blacks (9.49/100,000 women) was around half of the rates among the non-Blacks. (Coloreds: 18.11 per 100,000 women; Whites: 17.77/100,000 women; Indian/Asian: 13.24 per 100,000 women). Conclusions: Contrary to the trends in high- and middle-income countries, breast cancer mortality increased in South Africa especially among young women. Breast cancer prevention programs should be intensified and should also target young women. The marked disparity in ethnic burden of breast cancer should be considered during planning and implementation of interventions.