Stephen Weng

Systematic review and meta-analyses of risk factors for childhood overweight identifiable during infancy

Objective To determine risk factors for childhood overweight that can be identified during the first year of life to facilitate early identification and targeted intervention. Design Systematic review and meta-analysis. Search strategy Electronic database search of MEDLINE, EMBASE, PubMed and CAB Abstracts. Eligibility criteria Prospective observational studies following up children from birth for at least 2 years. Results Thirty prospective studies were identified. Significant and strong independent associations with childhood overweight were identified for maternal pre-pregnancy overweight, high infant birth weight and rapid weight gain during the first year of life. Meta-analysis comparing breastfed with non-breastfed infants found a 15% decrease (95% CI 0.74 to 0.99; I2=73.3%; n=10) in the odds of childhood overweight. For children of mothers smoking during pregnancy there was a 47% increase (95% CI 1.26 to 1.73; I2=47.5%; n=7) in the odds of childhood overweight. There was some evidence associating early introduction of solid foods and childhood overweight. There was conflicting evidence for duration of breastfeeding, socioeconomic status at birth, parity and maternal marital status at birth. No association with childhood overweight was found for maternal age or education at birth, maternal depression or infant ethnicity. There was inconclusive evidence for delivery type, gestational weight gain, maternal postpartum weight loss and ‘fussy’ infant temperament due to the limited number of studies. Conclusions Several risk factors for both overweight and obesity in childhood are identifiable during infancy. Future research needs to focus on whether it is clinically feasible for healthcare professionals to identify infants at greatest risk.

Systematic review of randomised controlled trials of interventions that aim to reduce the risk, either directly or indirectly, of overweight and obesity in infancy and early childhood

Abstract The risk factors for childhood overweight and obesity are known and can be identified antenatally or during infancy, however, the majority of effective interventions are designed for older children. This review identified interventions designed to reduce the risk of overweight/obesity that were delivered antenatally or during the first 2 years of life, with outcomes reported from birth to 7 years of age. Six electronic databases were searched for papers reporting randomised controlled trials of interventions published from January 1990 to September 2013. A total of 35 eligible studies were identified, describing 27 unique trials of which 24 were behavioural and three were non‐behavioural. The 24 behavioural trials were categorised by type of intervention: (1) nutritional and/or responsive feeding interventions targeted at parents of infants, which improved feeding practices and had some impact on child weight (n = 12); (2) breastfeeding promotion and lactation support for mothers, which had a positive effect on breastfeeding but not child weight (n = 5); (3) parenting and family lifestyle (n = 4); and (4) maternal health (n = 3) interventions that had some impact on feeding practices but not child weight. The non‐behavioural trials comprised interventions manipulating formula milk composition (n = 3). Of these, lower/hydrolysed protein formula milk had a positive effect on weight outcomes. Interventions that aim to improve diet and parental responsiveness to infant cues showed most promise in terms of self‐reported behavioural change. Despite the known risk factors, there were very few intervention studies for pregnant women that continue during infancy which should be a priority for future research.

Can machine-learning improve cardiovascular risk prediction using routine clinical data?

Background Current approaches to predict cardiovascular risk fail to identify many people who would benefit from preventive treatment, while others receive unnecessary intervention. Machine-learning offers opportunity to improve accuracy by exploiting complex interactions between risk factors. We assessed whether machine-learning can improve cardiovascular risk prediction. Methods Prospective cohort study using routine clinical data of 378,256 patients from UK family practices, free from cardiovascular disease at outset. Four machine-learning algorithms (random forest, logistic regression, gradient boosting machines, neural networks) were compared to an established algorithm (American College of Cardiology guidelines) to predict first cardiovascular event over 10-years. Predictive accuracy was assessed by area under the ‘receiver operating curve’ (AUC); and sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) to predict 7.5% cardiovascular risk (threshold for initiating statins). Findings 24,970 incident cardiovascular events (6.6%) occurred. Compared to the established risk prediction algorithm (AUC 0.728, 95% CI 0.723–0.735), machine-learning algorithms improved prediction: random forest +1.7% (AUC 0.745, 95% CI 0.739–0.750), logistic regression +3.2% (AUC 0.760, 95% CI 0.755–0.766), gradient boosting +3.3% (AUC 0.761, 95% CI 0.755–0.766), neural networks +3.6% (AUC 0.764, 95% CI 0.759–0.769). The highest achieving (neural networks) algorithm predicted 4,998/7,404 cases (sensitivity 67.5%, PPV 18.4%) and 53,458/75,585 non-cases (specificity 70.7%, NPV 95.7%), correctly predicting 355 (+7.6%) more patients who developed cardiovascular disease compared to the established algorithm. Conclusions Machine-learning significantly improves accuracy of cardiovascular risk prediction, increasing the number of patients identified who could benefit from preventive treatment, while avoiding unnecessary treatment of others.

Smoking and absence from work: systematic review and meta-analysis of occupational studies.

AIMS This study aimed to assess the association between smoking and absenteeism in working adults. METHODS A systematic review and meta-analysis was performed by electronic database searches in MEDLINE, EMBASE, CAB Abstracts, PubMed, Science Direct and National Health Service Economic Evaluation Database (February 2012). Longitudinal, prospective cohorts or retrospective cohorts were included in the review. Summary effect estimates were calculated using random-effects meta-analysis. Heterogeneity was assessed by I(2) and publication bias was investigated. RESULTS A total of 29 longitudinal or cohort studies were included. Compared with non-smokers, current smokers had a 33% increase in risk of absenteeism [95% confidence interval (CI): 1.25-1.41; I(2)  = 62.7%; 17 studies]. Current smokers were absent for an average of 2.74 more days per year compared with non-smokers (95% CI: 1.54-3.95; I(2)  = 89.6%; 13 studies). Compared with never smokers, ex-smokers had a 14% increase in risk of absenteeism (95% CI: 1.08-1.21; I(2)  = 62.4%; eight studies); however, no increase in duration of absence could be detected. Current smokers also had a 19% increase in risk of absenteeism compared with ex-smokers (95% CI: 1.09-1.32, P < 0.01, eight studies). There was no evidence of publication bias. The total cost of absenteeism due to smoking in the United Kingdom was estimated to be £1.4 billion in 2011. CONCLUSIONS Quitting smoking appears to reduce absenteeism and result in substantial cost-savings for employers.

Improving identification of familial hypercholesterolaemia in primary care: Derivation and validation of the familial hypercholesterolaemia case ascertainment tool (FAMCAT)

OBJECTIVE Heterozygous familial hypercholesterolaemia (FH) is a common autosomal dominant disorder. The vast majority of affected individuals remain undiagnosed, resulting in lost opportunities for preventing premature heart disease. Better use of routine primary care data offers an opportunity to enhance detection. We sought to develop a new predictive algorithm for improving identification of individuals in primary care who could be prioritised for further clinical assessment using established diagnostic criteria. METHODS Data were analysed for 2,975,281 patients with total or LDL-cholesterol measurement from 1 Jan 1999 to 31 August 2013 using the Clinical Practice Research Datalink (CPRD). Included in this cohort study were 5050 documented cases of FH. Stepwise logistic regression was used to derive optimal multivariate prediction models. Model performance was assessed by its discriminatory accuracy (area under receiver operating curve [AUC]). RESULTS The FH prediction model (FAMCAT), consisting of nine diagnostic variables, showed high discrimination (AUC 0.860, 95% CI 0.848-0.871) for distinguishing cases from non-cases. Sensitivity analysis demonstrated no significant drop in discrimination (AUC 0.858, 95% CI 0.845-0.869) after excluding secondary causes of hypercholesterolaemia. Removing family history variables reduced discrimination (AUC 0.820, 95% CI 0.807-0.834), while incorporating more comprehensive family history recording of myocardial infraction significantly improved discrimination (AUC 0.894, 95% CI 0.884-0.904). CONCLUSION This approach offers the opportunity to enhance detection of FH in primary care by identifying individuals with greatest probability of having the condition. Such cases can be prioritised for further clinical assessment, appropriate referral and treatment to prevent premature heart disease.

Estimating Overweight Risk in Childhood From Predictors During Infancy

OBJECTIVE: The aim of this study was to develop and validate a risk score algorithm for childhood overweight based on a prediction model in infants. METHODS: Analysis was conducted by using the UK Millennium Cohort Study. The cohort was divided randomly by using 80% of the sample for derivation of the risk algorithm and 20% of the sample for validation. Stepwise logistic regression determined a prediction model for childhood overweight at 3 years defined by the International Obesity Task Force criteria. Predictive metrics R2, area under the receiver operating curve (AUROC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. RESULTS: Seven predictors were found to be significantly associated with overweight at 3 years in a mutually adjusted predictor model: gender, birth weight, weight gain, maternal prepregnancy BMI, paternal BMI, maternal smoking in pregnancy, and breastfeeding status. Risk scores ranged from 0 to 59 corresponding to a predicted risk from 4.1% to 73.8%. The model revealed moderately good predictive ability in both the derivation cohort (R2 = 0.92, AUROC = 0.721, sensitivity = 0.699, specificity = 0.679, PPV = 38%, NPV = 87%) and validation cohort (R2 = 0.84, AUROC = 0.755, sensitivity = 0.769, specificity = 0.665, PPV = 37%, NPV = 89%). CONCLUSIONS: Using a prediction algorithm to identify at-risk infants could reduce levels of child overweight and obesity by enabling health professionals to target prevention more effectively. Further research needs to evaluate the clinical validity, feasibility, and acceptability of communicating this risk.

Validation, Optimal Threshold Determination, and Clinical Utility of the Infant Risk of Overweight Checklist for Early Prevention of Child Overweight

BACKGROUND Previous research has demonstrated the predictive validity of the Infant Risk of Overweight Checklist (IROC). This study further establishes the predictive accuracy of the IROC using data from the Avon Longitudinal Study of Parents and Children (ALSPAC) and examines the optimal threshold for determining high risk of childhood overweight. METHODS Using the IROC algorithm, we calculated the risk of being overweight, based on International Obesity Task Force criteria, in the first year of life for 980 children in the ALSPAC cohort at 5 years. Discrimination was assessed by the area under the receiver operating curve (AUC c-statistic). Net reclassification index (NRI) was calculated for risk thresholds ranging from 2.5% to 30%, which determine cutoffs for identifying infants at risk of becoming overweight. RESULTS At 5 years of age, 12.3% of boys and 19.6% of girls were categorized overweight. Discrimination (AUC c-statistic) ranged from 0.67 (95% confidence interval [CI], 0.62-0.72) when risk scores were calculated directly to 0.93 (95% CI, 0.88-0.98) when the algorithm was recalibrated and missing values of the risk factor algorithm were imputed. The NRI showed that there were positive gains in reclassification using risk thresholds from 5% to 20%, with the maximum NRI being at 10%. CONCLUSIONS This study confirms that the IROC has moderately good validity for assessing overweight risk in infants and offers an optimal threshold for determining high risk. The IROC algorithm has been imbedded into a computer program for Proactive Assessment of Obesity Risk during Infancy, which facilitates early overweight prevention through communication of risk to parents.

Feasibility of improving identification of familial hypercholesterolaemia in general practice: intervention development study

Objectives To assess the feasibility of improving identification of familial hypercholesterolaemia (FH) in primary care, and of collecting outcome measures to inform a future trial. Design Feasibility intervention study. Setting 6 general practices (GPs) in central England. Participants 831 eligible patients with elevated cholesterol >7.5 mmol/L were identified, by search of electronic health records, for recruitment to the intervention. Intervention Educational session in practice; use of opportunistic computer reminders in consultations or universal postal invitation over 6 months to eligible patients invited to complete a family history questionnaire. Those fulfilling the Simon-Broome criteria for possible FH were invited for GP assessment and referred for specialist definitive diagnosis. Outcome measures Rates of recruitment of eligible patients, identification of patients with possible FH, referral to specialist care, diagnosis of confirmed FH in specialist care; and feasibility of collecting relevant outcome measures for a future trial. Results Of 173 general practices, 18 were interested in participating and 6 were recruited. From 831 eligible patients, 127 (15.3%) were recruited and completed family history questionnaires: 86 (10.7%) through postal invitation and 41 (4.9%) opportunistically. Among the 127 patients, 32 (25.6%) had a possible diagnosis of FH in primary care. Within 6 months of completing recruitment, 7 patients had had specialist assessment confirming 2 patients with definite FH (28.6%), and 5 patients with possible FH (71.4%). Potential trial outcome measures for lipid tests, statin prescribing and secondary causes of hypercholesterolaemia were extracted using automated data extraction from electronic records alone without recourse to other methods. Conclusions The intervention is feasible to implement in GP, and facilitates recruitment of patients with raised cholesterol for targeted assessment and identification of FH. Extracting data directly from electronic records could be used to evaluate relevant outcome measures in a future trial.

Digital technology to facilitate Proactive Assessment of Obesity Risk during Infancy (ProAsk): a feasibility study

Objective To assess the feasibility and acceptability of using digital technology for Proactive Assessment of Obesity Risk during Infancy (ProAsk) with the UK health visitors (HVs) and parents. Design Multicentre, pre- and post-intervention feasibility study with process evaluation. Setting Rural and urban deprived settings, UK community care. Participants 66 parents of infants and 22 HVs. Intervention ProAsk was delivered on a tablet device. It comprises a validated risk prediction tool to quantify overweight risk status and a therapeutic wheel detailing motivational strategies for preventive parental behaviour. Parents were encouraged to agree goals for behaviour change with HVs who received motivational interviewing training. Outcome measures We assessed recruitment, response and attrition rates. Demographic details were collected, and overweight risk status. The proposed primary outcome measure was weight-for-age z-score. The proposed secondary outcomes were parenting self-efficacy, maternal feeding style, infant diet and exposure to physical activity/sedentary behaviour. Qualitative interviews ascertained the acceptability of study processes and intervention fidelity. Results HVs screened 324/589 infants for inclusion in the study and 66/226 (29%) eligible infants were recruited. Assessment of overweight risk was completed on 53 infants and 40% of these were identified as above population risk. Weight-for-age z-score (SD) between the infants at population risk and those above population risk differed significantly at baseline (−0.67 SD vs 0.32 SD). HVs were able to collect data and calculate overweight risk for the infants. Protocol adherence and intervention fidelity was a challenge. HVs and parents found the information provided in the therapeutic wheel appropriate and acceptable. Conclusion Study recruitment and protocol adherence were problematic. ProAsk was acceptable to most parents and HVs, but intervention fidelity was low. There was limited evidence to support the feasibility of implementing ProAsk without significant additional resources. A future study could evaluate ProAsk as a HV-supported, parent-led intervention. Trial registration number NCT02314494 (Feasibility Study Results)

The value of aspartate aminotransferase and alanine aminotransferase in cardiovascular disease risk assessment

Objective Aspartate aminotransferase to alanine aminotransferase (AST/ALT) ratio, reflecting liver disease severity, has been associated with increased risk of cardiovascular disease (CVD). The aim of this study was to evaluate whether the AST/ALT ratio improves established risk prediction tools in a primary care population. Methods Data were analysed from a prospective cohort of 29 316 UK primary care patients, aged 25–84 years with no history of CVD at baseline. Cox proportional hazards regression was used to derive 10-year multivariate risk models for the first occurrence of CVD based on two established risk prediction tools (Framingham and QRISK2), with and without including the AST/ALT ratio. Overall, model performance was assessed by discriminatory accuracy (AUC c-statistic). Results During a total follow-up of 120 462 person-years, 782 patients (59% men) experienced their first CVD event. Multivariate models showed that elevated AST/ALT ratios were significantly associated with CVD in men (Framingham: HR 1.37, 95% CI 1.05 to 1.79; QRISK2: HR 1.40, 95% CI 1.04 to 1.89) but not in women (Framingham: HR 1.06, 95% CI 0.78 to 1.43; QRISK2: HR 0.97, 95% CI 0.70 to 1.35). Including the AST/ALT ratio with all Framingham risk factors (AUC c-statistic: 0.72, 95% CI 0.71 to 0.74) or QRISK2 risk factors (AUC c-statistic: 0.73, 95% CI 0.71 to 0.74) resulted in no change in discrimination from the established risk prediction tools. Limiting analysis to those individuals with raised ALT showed that discrimination could improve by 5% and 4% with Framingham and QRISK2 risk factors, respectively. Conclusions Elevated AST/ALT ratio is significantly associated with increased risk of developing CVD in men but not women. However, the ratio does not confer any additional benefits over established CVD risk prediction tools in the general population, but may have clinical utility in certain subgroups.