Emerging Themes in Epidemiology最新文献

Background: There are both genetic and environmental factors which contribute to a child's chances of being obese. When low birth weight (LBW) has been specifically evaluated relative to its association with childhood obesity, the results have produced conflicting findings. This study aims to describe the relationship between birth weight and childhood obesity and investigate the influence that residence and household wealth has on this relationship.

Methods: I performed a secondary analysis on the 2013 Riskesdas (or Basic Health Research), a cross-sectional, nationally representative survey of the Indonesian population. Height, weight, information regarding child's birth weight, and basic characteristics of the study population were collected from parents with children aged 0 to 5 years (n = 63,237) in 2013. The exposure was child's birth weight and the outcomes were child's current weight, BMI z-score, and obesity. Data were analyzed by using multiple linear regression and multiple logistic regression.

Results: I found a significant increase in the weight, BMI z-score, and risk of childhood obesity to be associated with LBW. LBW children in rural area were associated with higher BMI z-score (mean ± standard error: 1.44 ± 0.02) and higher odds (odds ratio (95% confidence interval): 7.46 (6.77-8.23)) of obesity than those in urban area. LBW children from low class families were associated with higher BMI z-score (1.79 ± 0.04) and had higher odds (14.79 (12.47-17.54)) of obesity than those from middle class and wealthy families.

Conclusions: Effective prevention and intervention to childhood obesity as early as possible are imperative. As far as this study was concerned, efforts, policies, and targets are required to reduce the prevalence of LBW. Children born of LBW, who live in a rural area and from low income families, should be emphatically intervened as early as possible.

Multiple imputation is a recommended method for handling incomplete data problems. One of the barriers to its successful use is the breakdown of the multiple imputation procedure, often due to numerical problems with the algorithms used within the imputation process. These problems frequently occur when imputation models contain large numbers of variables, especially with the popular approach of multivariate imputation by chained equations. This paper describes common causes of failure of the imputation procedure including perfect prediction and collinearity, focusing on issues when using Stata software. We outline a number of strategies for addressing these issues, including imputation of composite variables instead of individual components, introducing prior information and changing the form of the imputation model. These strategies are illustrated using a case study based on data from the Longitudinal Study of Australian Children.

Background: To illustrate the public health potential of linking individual bedside data with community-based household data in a poor rural setting, we estimated excess pediatric mortality risk after discharge from St Francis Designated District Hospital in Ifakara, Tanzania.

Methods: Linked data from demographic and clinical surveillance were used to describe post-discharge mortality and survival probability in children aged < 5 years, by age group and cause of admission. Cox regression models were developed to identify risk factors.

Results: Between March 2003 and March 2007, demographic surveillance included 28,910 children aged 0 to 5 years and among them 831 (3%) were admitted at least once to the district hospital. From all the children under the demographic surveillance 57,880 person years and 1381 deaths were observed in 24 months of follow up. Survivors of hospital discharge aged 0-5 years were almost two times more likely to die than children of the same age in the community who had not been admitted (RR = 1.9, P < 0.01, 95% CI 1.6, 2.4). Amongst children who had been admitted, mortality rate within a year was highest in infants (93 per 1000 person years) and amongst those admitted due to pneumonia and diarrhoea (97 and 85 per 1000 person years respectively). Those who lived 75 km or further from the district hospital, amongst children who were admitted and survived discharge from hospital, had a three times greater chance of dying within one year compared to those living within 25 km (adjusted HR 3.23, 95% CI 1.54,6.75). The probability of surviving the first 30 days post hospitalization was 94.4% [95% CI 94.4, 94.9], compared to 98.8% [95% CI 97.199.5] in non-hospitalized children of the same age in the commuity.

Conclusion: This study illustrates the potential of linking health related data from facility and household levels. Our results suggest that families may need additional support post hospitalization.

Background: In South Africa, age-disparate to sexual relationships where the age difference between partners is 5 years or greater is an important contributor to the spread of HIV. However, little is known about the predictors of age-disparate sexual relationships. This study investigates factors associated with age-disparate sexual relationships among males and females in South Africa.

Methods: This analysis used the 2012 nationally representative population-based household survey conducted using multi-stage stratified cluster sampling design. Multivariate multinomial stepwise logistic regression models were used to determine factors associated with age-disparate sexual relationships.

Results: Of 15,717 participants, who responded to the question on age-disparate sexual relationships, 62% males versus 58.5% females had partners within 5 years older or younger, 34.7% of males versus 2.7% of females had partners at least 5 years younger and 3.3% of males versus 38.8% of females had partners at least 5 years older. Among both males and females predictors of age-disparate sexual relationships were education, employment, socioeconomic status, locality type, age at sexual debut, condom use at last sexual act and HIV status while race was also an additional predictor for among females. Including unprotected sex and risk of HIV infection among adolescent girls and young women with sexual partners 5 years older their age.

Conclusions: This study suggest that there is a need for reprioritizing the combination of behavioural and structural interventions to address risky sexual behaviours, unprotected sex, poverty, limited education and gender inequitable norms related to age-disparate sexual relationships and HIV.

Background: Non-sensitive and non-specific observation of outcomes in time-to-event data affects event counts as well as the risk sets, thus, biasing the estimation of hazard ratios. We investigate how imperfect observation of incident events affects the estimation of vaccine effectiveness based on hazard ratios.

Methods: Imperfect time-to-event data contain two classes of events: a portion of the true events of interest; and false-positive events mistakenly recorded as events of interest. We develop an estimation method utilising a weighted partial likelihood and probabilistic deletion of false-positive events and assuming the sensitivity and the false-positive rate are known. The performance of the method is evaluated using simulated and Finnish register data.

Results: The novel method enables unbiased semiparametric estimation of hazard ratios from imperfect time-to-event data. False-positive rates that are small can be approximated to be zero without inducing bias. The method is robust to misspecification of the sensitivity as long as the ratio of the sensitivity in the vaccinated and the unvaccinated is specified correctly and the cumulative risk of the true event is small.

Conclusions: The weighted partial likelihood can be used to adjust for outcome measurement errors in the estimation of hazard ratios and effectiveness but requires specifying the sensitivity and the false-positive rate. In absence of exact information about these parameters, the method works as a tool for assessing the potential magnitude of bias given a range of likely parameter values.

Background: Worsening socioeconomic conditions in rural America have been fueling increases in chronic disease and poor health. The goal of this study was to identify cost-effective methods of deploying geographically targeted health surveys in rural areas, which often have limited resources. These health surveys were administered in New York's rural Sullivan County, which has some of the poorest health outcomes in the entire state.

Methods: Comparisons were made for response rates, estimated costs, respondent demographics, and prevalence estimates of a brief health survey delivered by mail and phone using address-based sampling, and in-person using convenience sampling at a sub-county level in New York's rural Sullivan County during 2017.

Results: Overall response rates were 27.0% by mail, 8.2% by phone, and 71.4% for convenience in-person surveys. Costs to perform phone surveys were substantially higher than mailed or convenience in-person surveys. All modalities had lower proportions of Hispanic respondents compared to Census estimates. Unadjusted and age-adjusted prevalence estimates were similar between mailed and in-person surveys, but not for phone surveys.

Conclusions: These findings are consistent with declining response rates of phone surveys, which obtained an inadequate sample of rural residents. Though in-person surveys had higher response rates, convenience sampling failed to obtain a geographically distributed sample of rural residents. Of modalities tested, mailed surveys provided the best opportunity to perform geographically targeted rural health surveillance.

A key aspect of the article by Lousdal on instrumental variables was a discussion of the monotonicity assumption. However, there was no mention of the history of the development of this assumption. The purpose of this letter is to note that Baker and Lindeman and Imbens and Angrist independently introduced the monotonicity assumption into the analysis of instrumental variables. The letter also places the monotonicity assumption in the context of the method of latent class instrumental variables.

We commend Nickerson and Brown on their insightful exposition of the mathematical algebra behind Simpson's paradox, suppression and Lord's paradox; we also acknowledge there can be differences in how Lord's paradox is approached analytically, compared to Simpson's paradox and suppression, though not in every example of Lord's paradox. Furthermore, Simpson's paradox, suppression and Lord's paradox ask the same contextual questions, seeking to understand if statistical adjustment is valid and meaningful, identifying which analytical option is correct. In our exposition of this, we focus on the perspective of context, which must invoke causal thinking. From a causal thinking perspective, Simpson's paradox, suppression and Lord's paradox present very similar analytical challenges.