Yiliang Zhu

Prognostic Performance of Metabolic Indexes in Predicting Onset of Type 1 Diabetes

OBJECTIVE In this investigation we evaluated nine metabolic indexes from intravenous glucose tolerance tests (IVGTTs) and oral glucose tolerance tests (OGTTs) in an effort to determine their prognostic performance in predicting the development of type 1 diabetes in those with moderate risk, as defined by familial relation to a type 1 diabetic individual, a positive test for islet cell antibodies and insulin autoantibody, but normal glucose tolerance. RESEARCH DESIGN AND METHODS Subjects (n = 186) who had a projected risk of 25–50% for developing type 1 diabetes within 5 years were followed until clinical diabetes onset or the end of the study as part of the Diabetes Prevention Trial–Type 1. Prognostic performance of the metabolic indexes was determined using receiver operating characteristic (ROC) curve and survival analyses. RESULTS Two-hour glucose from an OGTT most accurately predicted progression to disease compared with all other metabolic indicators with an area under the ROC curve of 0.67 (95% CI 0.59–0.76), closely followed by the ratio of first-phase insulin response (FPIR) to homeostasis model assessment of insulin resistance (HOMA-IR) with an area under the curve value of 0.66. The optimal cutoff value for 2-h glucose (114 mg/dl) maintained sensitivity and specificity values >0.60. The hazard ratio for those with 2-h glucose ≥114 mg/dl compared with those with 2-h glucose <114 mg/dl was 2.96 (1.67–5.22). CONCLUSIONS The ratio of FPIR to HOMA-IR from an IVGTT provided accuracy in predicting the development of type 1 diabetes similar to that of 2-h glucose from an OGTT, which, because of its lower cost, is preferred. The optimal cutoff value determined for 2-h glucose provides additional guidance for clinicians to identify subjects for potential prevention treatments before the onset of impaired glucose tolerance.

Promoting Physical Activity Among Youth Through Community‐Based Prevention Marketing

BACKGROUNDnCommunity-based prevention marketing (CBPM) is a program planning framework that blends community-organizing principles with a social marketing mind-set to design, implement, and evaluate public health interventions. A community coalition used CBPM to create a physical activity promotion program for tweens (youth 9-13 years of age) called VERB Summer Scorecard. Based on the national VERB media campaign, the program offered opportunities for tweens to try new types of physical activity during the summer months.nnnMETHODSnThe VERB Summer Scorecard was implemented and monitored between 2004 and 2007 using the 9-step CBPM framework. Program performance was assessed through in-depth interviews and a school-based survey of youth.nnnRESULTSnThe CBPM process and principles used by school and community personnel to promote physical activity among tweens are presented. Observed declines may become less steep if school officials adopt a marketing mind-set to encourage youth physical activity: deemphasizing health benefits but promoting activity as something fun that fosters spending time with friends while trying and mastering new skills.nnnCONCLUSIONSnCommunity-based programs can augment and provide continuity to school-based prevention programs to increase physical activity among tweens.

Preventing eye injuries among citrus harvesters: the community health worker model.

OBJECTIVESnAlthough eye injuries are common among citrus harvesters, the proportion of workers using protective eyewear has been negligible. We focused on adoption of worker-tested safety glasses with and without the presence and activities of trained peer-worker role models on harvesting crews.nnnMETHODSnObservation of 13 citrus harvesting crews established baseline use of safety eyewear. Nine crews subsequently were assigned a peer worker to model use of safety glasses, conduct eye safety education, and treat minor eye injuries. Safety eyewear use by crews was monitored up to 15 weeks into the intervention.nnnRESULTSnIntervention crews with peer workers had significantly higher rates of eyewear use than control crews. Intervention exposure time and level of worker use were strongly correlated. Among intervention crews, workers with 1 to 2 years of experience (odds ratio [OR] = 2.89; 95% confidence interval [CI] = 1.11, 7.55) and who received help from their peer worker (OR = 3.73; 95% CI = 1.21, 11.57) were significantly more likely to use glasses than were other intervention crew members.nnnCONCLUSIONSnAdaptation of the community health worker model for this setting improved injury prevention practices and may have relevance for similar agricultural settings.

Using Community-Based Prevention Marketing to Improve Farm Worker Safety

Community-based prevention marketing (CBPM) combines a powerful planning framework, social marketing, with community organization principles to design behavior change programs. In southwest Florida, a coalition comprised of citrus workers and their employers, health providers, and academic researchers is using CBPM to identify occupational health issues among agricultural laborers, conduct community-based participatory research, and design culturally appropriate interventions. This article describes how this coalition was able to apply CBPM successfully to develop and implement an occupational safety program to prevent eye injuries among migrant farm workers. Lessons learned from this project and implications for designing and disseminating occupational safety programs for other agricultural workers are discussed.

Comment on: Effects of Decabrominated Diphenyl Ether (PBDE 209) Exposure at Different Developmental Periods on Synaptic Plasticity in the Dentate Gyrus of Adult Rats In Vivo

Xing et al. (2009) exposed Wistar rats to decabromodiphenyl ether (PBDE 209; CASRN 1163-19-5) during five developmental periods: pregnancy, lactation via mother’s milk (indirect dosing), lactation via gavage (direct dosing), postweaning (direct dosing), and prenatal to life (indirect dosing during gestation and lactation and direct dosing postweaning). All animals were evaluated on postnatal day 60. The authors reported that exposure during lactation was the most sensitive period for causing effects on postsynaptic cell excitability. In the Materials and Methods section, the authors stated: ‘‘[t]he animals in each group consisted of three to four litters with both sexes (the ratio of males to females was ~1:1) to avoid the effects of litters and sexes.’’ For the input/output functions, paired-pulse reactions, long-term potentiation, and PBDE 209 concentrations in the hippocampus, however, the authors reported an ‘‘n’’ of 5, 7, 8, 9, 10, or 13 for the respective treatment groups (Xing et al., 2009). Thus, Xing et al. (2009) utilized an average of 1–2 (n 1⁄4 5), 1–3 (n 1⁄4 7), 2–3 (n 1⁄4 8), 2–3 (n 1⁄4 9), 2–4 (n 1⁄4 10), or 3–4 (n 1⁄4 13) littermates as independent values per treatment group. This study is based on a clustered experimental design, and data analysis must control for litter effects. The U.S. Environmental Protection Agency recently cosponsored two expert working groups on statistical considerations for use in developmental neurotoxicity testing and for direct dosing preweaning animals (Holson et al., 2008; Moser et al., 2005; Zoetis and Walls, 2003). These expert groups concluded that the litter must be used as the experimental and statistical unit regardless of whether animals are dosed indirectly or directly. Based on the number of litters used by Xing et al. (2009), their reported sample size should be an ‘‘n’’ of 3 or 4, with values from littermates treated as replicates within each litter. Failing to control for litter effects can have a profound impact on data analysis and resulting conclusions. For example, the rate of type I or false-positive errors has been reported to triple the nominal 0.05 alpha level when as few as two pups per litter were used as independent measures (Holson and Pearce, 1992). Because Xing et al. (2009) utilized varying numbers of pups from the same litter per treatment group as independent values, their results may be reflective of litter effects, inflating the significance of test article–related effects. It is noteworthy that the in vivo studies cited by Xing et al. (2009) as support that PBDEs cause developmental neurotoxicity also failed to control for litter effects (Eriksson et al., 2001, 2002; Fischer et al., 2008; Viberg et al., 2003a,b 2004, 2007, 2008). Furthermore, PBDE 209 was evaluated using the most recent validated test guideline for developmental neurotoxicity (OECD, 2007). This study utilized doses ranging from 1 mg/kg/day to 1000 mg/kg/day and was unable to replicate the effects reported by Viberg et al. (2003b, 2007) (Hardy et al., 2009). In conclusion, Xing et al. (2009) failed to control for litter effects in analyzing data arising from a clustered experimental design. If the authors tracked which pups came from which litters per treatment group, their data may be reanalyzed to determine to what degree the effect they attributed to the treatment was a result of uncontrolled litter effects. In the absence of such a reanalysis, data by Xing et al. (2009) are of questionable relevance for risk assessment.

Bayesian inference for two-part mixed-effects model using skew distributions, with application to longitudinal semicontinuous alcohol data

Semicontinuous data featured with an excessive proportion of zeros and right-skewed continuous positive values arise frequently in practice. One example would be the substance abuse/dependence symptoms data for which a substantial proportion of subjects investigated may report zero. Two-part mixed-effects models have been developed to analyze repeated measures of semicontinuous data from longitudinal studies. In this paper, we propose a flexible two-part mixed-effects model with skew distributions for correlated semicontinuous alcohol data under the framework of a Bayesian approach. The proposed model specification consists of two mixed-effects models linked by the correlated random effects: (i) a model on the occurrence of positive values using a generalized logistic mixed-effects model (Part I); and (ii) a model on the intensity of positive values using a linear mixed-effects model where the model errors follow skew distributions including skew-t and skew-normal distributions (Part II). The proposed method is illustrated with an alcohol abuse/dependence symptoms data from a longitudinal observational study, and the analytic results are reported by comparing potential models under different random-effects structures. Simulation studies are conducted to assess the performance of the proposed models and method.

VERB™ Summer Scorecard: Increasing Tween Girls' Vigorous Physical Activity

OBJECTIVEnWe assessed changes in the frequency of self-reported physical activity (PA) among tween girls exposed and not exposed to the VERB™ Summer Scorecard (VSS) intervention in Lexington, Kentucky, during 2004, 2006, and 2007.nnnMETHODSnGirls who reported 0-1u2009day per week of PA were classified as having little or no PA. Girls who reported 2-3u2009days of PA were classified as low PA performers; 4-5u2009days of PA were labeled as moderate performers; and 6-7u2009days of PA were identified as high performers. Logit regression analysis of survey data from girls identified trends in PA frequency across time.nnnRESULTSnIn 2004, participant girls were more likely than girls unfamiliar with VSS (reference group girls) to report high frequency of PA (ORu2009=u20091.44, CIu2009=u20091.18, 1.70). In 2006, participants were statistically less likely than reference group girls to report low frequency of PA (ORu2009=u20091.75, CIu2009=u20091.33, 2.21). In 2007, VSS participants were consistently more likely to report moderate frequency (ORu2009=u20091.56, CIu2009=u20091.35, 1.77) and high frequency of PA (ORu2009=u20091.44, CIu2009=u20091.24, 1.64) than reference group girls.nnnCONCLUSIONnAn innovative, community-driven intervention demonstrated promise for increasing PA among tween girls. VSS may have transportability to other communities to help reverse the secular trend of declining PA for this population segment.

Bayesian Inference on Bivariate Semi-continuous Mixed-effects Models with Application to Longitudinal Substance Use Data

Multivariate (bivariate) correlated data encountered frequently in longitudinal studies are often analyzed using a multivariate linear mixed-effects model with normality assumption. Semi-continuous data in the form of a mixture of high proportion of zeros and right-skewed positive values bring special challenges to the field of multivariate modeling. In this paper, we propose a Bayesian approach to analyze bivariate semi-continuous outcomes by jointly modeling a generalized logistic mixed-effects model on zero-inflation in either response and a bivariate linear mixed-effects model (BLMM) on the positive values given both responses occurred through a correlated randomeffects structure. Multivariate skew distributions including skew-t and skew-normal distributions are used to relax the normality assumption in BLMM. The proposed models are illustrated with an application to the correlated alcohol and drug uses data from a longitudinal observational study. A simulation study is conducted to evaluate the performance of the proposed models.

Bayesian Bivariate Linear Mixed-Effects Models with Skew-Normal/Independent Distributions, with Application to AIDS Clinical Studies

Bivariate correlated (clustered) data often encountered in epidemiological and clinical research are routinely analyzed under a linear mixed-effected (LME) model with normality assumptions for the random-effects and within-subject errors. However, those analyses might not provide robust inference when the normality assumptions are questionable if the data set particularly exhibits skewness and heavy tails. In this article, we develop a Bayesian approach to bivariate linear mixed-effects (BLME) models replacing the Gaussian assumptions for the random terms with skew-normal/independent (SNI) distributions. The SNI distribution is an attractive class of asymmetric heavy-tailed parametric structure which includes the skew-normal, skew-t, skew-slash, and skew-contaminated normal distributions as special cases. We assume that the random-effects and the within-subject (random) errors, respectively, follow multivariate SNI and normal/independent (NI) distributions, which provide an appealing robust alternative to the symmetric normal distribution in a BLME model framework. The method is exemplified through an application to an AIDS clinical data set to compare potential models with different distribution specifications, and clinically important findings are reported.