A sensitivity analysis approach for the causal hazard ratio in randomized and observational studies.

The hazard ratio (HR) is often reported as the main causal effect when studying survival data. Despite its popularity, the HR suffers from an unclear causal interpretation. As already pointed out in the literature, there is a built-in selection bias in the HR, because similarly to the truncation by death problem, the HR conditions on post-treatment survival. A recently proposed alternative, inspired by the Survivor Average Causal Effect, is the causal HR, defined as the ratio between hazards across treatment groups among the study participants that would have survived regardless of their treatment assignment. We discuss the challenge in identifying the causal HR and present a sensitivity analysis identification approach in randomized controlled trials utilizing a working frailty model. We further extend our framework to adjust for potential confounders using inverse probability of treatment weighting. We present a Cox-based and a flexible non-parametric kernel-based estimation under right censoring. We study the finite-sample properties of the proposed estimation methods through simulations. We illustrate the utility of our framework using two real-data examples.

Comparisons in polypharmacy over a decade in community-dwelling older adults-findings from Israel national health and nutrition surveys.

BACKGROUND: Polypharmacy increases with age and is associated with serious health and economic costs. This study reports changes over a decade in medication-use patterns and polypharmacy, in Israeli community-dwelling older adults aged ≥ 65 years. METHODS: Demographic and health data from two representative national health cross-sectional surveys - MABAT ZAHAV 1 (MZ1) in 2005-2006, and MZ2 in 2014-2015 were analyzed. Polypharmacy was defined as use of ≥ 5 medications. Risk factors for polypharmacy were estimated by multivariable logistic regression with adjusted odds ratios (aOR) and their 95% confidence intervals (CI). RESULTS: Self-reported data on medications taken were available for 1647 participants (91.5%) in MZ1, and for 833 participants (80.2%) in MZ2, 55% women, and about 20% aged ≥ 80, in both surveys. The prevalence of polypharmacy was significantly lower in MZ2 than in MZ1: 64.2% versus 56.3%, p = .0001; with an aOR (95%CI) of 0.64 (0.52, 0.80). The most commonly taken drugs were for hypertension (27.0%, 25.3%), dyslipidemia (9.7%, 12.4%) and anticoagulation (9.2%, 9.8%). For approximately 10% of drugs, indications were either unknown or incorrect. Polypharmacy was significantly associated with poor self-health assessment 2.47 (1.99, 3.06), ≥ 4 versus 1-3 chronic illnesses 6.36 (3.85, 10.50), and age ≥ 80 versus younger 1.72 (1.32, 2.24). Similar associations were observed with major polypharmacy of ≥ 8 medications. CONCLUSION: Polypharmacy, although reduced in the last decade, requires constant attention, especially concerning lack of knowledge of indications which leads to poor adherence and adverse side effects. Health-care teams should carry out regular medicine reconciliation in at-risk elderly patients.

Energy Intake Is Highly Associated with Handgrip Strength in Community-Dwelling Elderly Adults.

BACKGROUND: Handgrip strength (HGS) is used to assess muscle strength, which is an important indicator of health status in older adults. OBJECTIVE: We evaluated associations of demographic, anthropometric, clinical, and nutritional factors with muscle strength in community-dwelling elderly adults. METHODS: This population-based cross-sectional study employed the 2014-2015 Israeli Health and Nutrition Survey in the Elderly (n = 1039, age ≥65 y, 46.7% males, 9.9% Arabs). Trained personnel performed face-to-face interviews that focused on health and nutrition status, as well as anthropometric measurements and examinations. HGS was measured in the dominant hand 3 times using a digital grip strength dynamometer; the highest result was reported. Dietary intake data were collected using single-day 24-h recall. Multivariable logistic regressions were used to explore factors associated with low HGS (<27 kg for men and <16 kg for women). RESULTS: HGS measurements were completed by 704 participants. Following adjustment for several factors, higher prevalence of low HGS was significantly associated with age (OR: 1.14; 95% CI: 1.11, 1.18), whereas decreased prevalence was associated with higher levels of education (OR: 0.55; 95% CI: 0.32, 0.94) and meeting physical activity recommendations (OR: 0.53; 95% CI: 0.31, 0.88); P < 0.05 for all. Incremental increases of 100 kcal/d in energy intake and of 1 cm in midarm circumference were associated with decreased prevalence of low HGS (OR: 0.95; 95% CI: 0.91, 0.99 and OR: 0.91; 95% CI: 0.85, 0.97, respectively; P < 0.01 for both). Associations were not found of low HGS with ethnicity, comorbidity, BMI, smoking, or alcohol consumption or with protein, carbohydrate, or fat intakes. CONCLUSION: Energy intake, physical activity, midarm circumference, and education are associated with HGS in elderly Israeli adults. Further cohort studies are necessary to assess possible causal relations between these factors and HGS. Modifiable factors should be targeted in planning public health strategies for promoting a healthy aging population.

Organophosphate pesticide exposure in children in Israel: Dietary associations and implications for risk assessment.

BACKGROUND: Human biomonitoring (HBM) data is increasingly being compared to risk-based screening values to assess human health risk. However, as screening values have not been established for assessing biomarker concentrations of organophosphate (OP) pesticide metabolites, there are few studies using HBM data on urinary OP concentrations to assess human health risk. The purpose of the current study was to measure OP exposure in a sample of children in Israel; to explore associations between dietary patterns and OP exposure; and to assess risk of OP pesticides using urinary metabolite concentrations. METHODS: We recruited 103 children in Israel and collected demographic and dietary data and urinary samples, and measured creatinine and dialkyl phosphate (DAP) concentrations. We compared urinary DAP concentrations to international populations and analysed associations between fruit and vegetable consumption and urinary DAP concentrations. Using urinary DAP concentrations, we calculated estimated daily intakes (EDI) of OP pesticides in each child and compared those to the acceptable daily intake (ADI). RESULTS: Concentrations of several dialkyl phosphate metabolites (dimethylphosphate (DMP) and dimethylthiophosphate (DMTP)) were higher in our study population of Israeli children (geometric mean concentrations of DMP and DMTP were 6.6 µg/L and 7.6 µg/L, respectively) compared to children in the US, Canada, Spain, and Denmark. We found positive correlations between total fruit consumption and creatinine adjusted log transformed urinary DMP, DMTP, diethylthiophopshate (DETP), total dimethyl (DM) and total DAP concentrations (p < 0.05), positive correlations between cucumber consumption and diethylphosphate (DEP), DETP and diethyl (DE) concentrations (p < 0.05), and positive correlations between apple consumption and DETP concentrations (p = 0.02). Based on urinary DAP concentrations, we found that a portion of the children in our study had EDIs above the ADI, ranging from 2.9% to 79.4% of the children, depending on the active OP ingredient. CONCLUSIONS: We found that Israeli children in our study are widely exposed to OP pesticides; that levels of dimethyl metabolites were high compared to other international populations; and that fruit consumption was associated with higher urinary DAP levels. Using urinary DAP concentration data, we found that a portion of the children in our study may be exposed to OP pesticides at levels above those considered safe.

Dietary Patterns among Adolescents Are Associated with Growth, Socioeconomic Features, and Health-Related Behaviors.

Dietary patterns (DPs), usually established in adolescents, are important modifiable risk factors in the etiology of malnutrition and chronic diseases. This study aimed to identify DPs of adolescents and examine their associations with growth, sociodemographic, and lifestyle characteristics. A nationally representative, school-based, cross-sectional study was conducted in Israeli adolescents aged 11-18 years during 2015-2016. A self-administered survey queried sociodemographics, health behaviors, and diet. Weight and height were measured, and WHO height z-scores and BMI cutoffs were calculated. Food frequency questionnaire data were analyzed using principal components analysis (PCA) to identify DPs. Associations between growth, lifestyle, and sociodemographic characteristics and DPs were modeled using multivariable logistic regressions. A total of 3902 adolescents (46% males, mean age 15.2 ± 1.6 years) completed the survey. PCA identified five DPs, accounting for 38.3% of the total variance. The first two prominent DPs were the 'plant-based food' DP, which was associated with the female sex, higher socioeconomic status, overweight/obesity, and healthy lifestyle and the 'junk food' DP, which was associated with lower SES, unhealthy lifestyle, and lower height z-scores. Our results elucidate major DPs that strongly correlate with lifestyle risk behaviors and suboptimal growth among adolescents. Implementing screening for DPs should be further examined to identify higher risk health factors among youth.

Menarche at an Earlier Age: Results from Two National Surveys of Israeli Youth, 2003 and 2016.

STUDY OBJECTIVE: To assess emergent changes in the age at menarche and investigate associated factors in Israeli adolescents in 2003 and 2016. DESIGN: Cross-sectional study. SETTING: Two national representative school-based surveys (first and second "Mabat Youth"). PARTICIPANTS: Both surveys included female students in 7th-12th grades (ages 11-19 years). The first (N = 3328) was conducted between the years 2003 and 2004, and the second (N = 2535) from 2015 to 2016. INTERVENTIONS: The survey questionnaire was self-administered and anthropometric measurements were performed by trained personnel. MAIN OUTCOME MEASURES: The current age at menarche in Israeli girls was determined and independent factors (demographic, clinical, and lifestyle) examined. Changes that occurred since the past national survey more than a decade ago were documented. RESULTS: The estimated median age at menarche declined from 13.0 (interquartile range, 12.0-14.0) years in 2003-2004 to 12.5 (interquartile range, 12.0-13.0) years in 2015-2016 (P < .0001). Jewish girls reached menarche earlier than Arab girls, but both populations experienced a similar downward trend in the past approximately 14 years. Greater body mass index, higher socioeconomic status, and immigrant status were associated with younger menarche onset (P < .001). Age at menarche remained lower in 2015-2016 vs 2003-2004, even after adjustment for these potential confounders, with a high hazard ratio (HR), which decreased as a function of survival time (t): HRt = 15.417 × 0.813t. CONCLUSION: This study confirms the decline in age at menarche in Israel. Findings were associated with body mass index and population group but also indicated that other factors are likely involved.

Socioeconomic inequalities in exposure to environmental tobacco smoke in children in Israel.

BACKGROUND: Environmental tobacco smoke (ETS) exposure in infants and children causes more frequent and severe asthma attacks, respiratory infections, ear infections, and sudden infant death syndrome. The aim of this study was to measure ETS exposure in children in Israel (ages 4-11 years) using urinary cotinine measurements, in order to compare exposure levels to other international populations, and to assess predictors of ETS exposure in children in Israel. METHODS: A subset of children who participated in the National Health and Nutrition Survey (RAV- MABAT) in 2015-2016 were invited to participate in the Second Israel Biomonitoring Survey. We analyzed urinary cotinine and creatinine concentrations in 103 children. Parents of study participants were interviewed in person on children's exposure to ETS at home and in other environments and on sociodemographic variables. We calculated creatinine-adjusted and unadjusted urinary cotinine geometric means in children and analyzed associations in univariable and multivariable analyses, between sociodemographic variables and parental - reported exposure, and urinary cotinine concentrations. RESULTS: Based on urinary creatinine measurement, over 60% of children are exposed to ETS (compared to <40% based on parental report). Linear regression showed a positive association between urinary cotinine concentration and reported ETS exposure (p = 0.001). Mean cotinine concentration among children whose parents reported that they are exposed to ETS at home (5.1 µg/l) was significantly higher than the concentration among children whose parents reported they are not exposed to ETS at home (1.6 µg/l, p < 0.001). There was an inverse relationship between total family income and urinary cotinine concentration (p < 0.05). In a multivariable model adjusted for ethnicity and other factors, family income was a significant predictor of urinary cotinine level (p = 0.04, slope = -0.49). Geometric mean creatinine adjusted concentrations in children in the current study were higher than in children in Canada and selected European countries. CONCLUSIONS: We found evidence of widespread exposure to ETS in children in the study. There is an urgent need to protect children in Israel from exposure to ETS.