Individual-level Estimates Research Articles

Model-based estimation of individual-level social determinants of health and its applications in All of Us.

We introduce a widely applicable model-based approach for estimating individual-level Social Determinants of Health (SDoH) and evaluate its effectiveness using the All of Us Research Program. Our approach utilizes aggregated SDoH datasets to estimate individual-level SDoH, demonstrated with examples of no high school diploma (NOHSDP) and no health insurance (UNINSUR) variables. Models are estimated using American Community Survey data and applied to derive individual-level estimates for All of Us participants. We assess concordance between model-based SDoH estimates and self-reported SDoHs in All of Usand examine associations with undiagnosed hypertension and diabetes. Compared to self-reported SDoHs, the area under the curve for NOHSDP is 0.727 (95% CI, 0.724-0.730) and for UNINSUR is 0.730 (95% CI, 0.727-0.733) among the 329074 All of Us participants, both significantly higher than aggregated SDoHs. The association between model-based NOHSDP and undiagnosed hypertension is concordant with those estimated using self-reported NOHSDP, with a correlation coefficient of 0.649. Similarly, the association between model-based NOHSDP and undiagnosed diabetes is concordant with those estimated using self-reported NOHSDP, with a correlation coefficient of 0.900. The model-based SDoH estimation method offers a scalable and easily standardized approach for estimating individual-level SDoHs. Using the All of Us dataset, we demonstrate reasonable concordance between model-based SDoH estimates and self-reported SDoHs, along with consistent associations with health outcomes. Our findings also underscore the critical role of geographic contexts in SDoH estimation and in evaluating the association between SDoHs and health outcomes.

Computing levels of nutrient inadequacy from household consumption and expenditure surveys: A case study

This paper presents an approach to estimate the between-subject variability in nutrient intake (through the coefficient of variation [CV]) and a method to estimate the prevalence of nutrient inadequacy (PoNI) (for eight micronutrients) using household consumption and expenditure survey (HCES) data. Prevalence values are compared to individual-level estimates derived using the National-Cancer-Institute method. Data come from the 2015 Bangladesh Integrated-Household-Survey, which conducted a household-level 7-day recall (7DR) and two rounds of individual-level 24-hour recall (24HR), filled by one respondent on behalf of all members, for the same rural households. The PoNI values based on 7DR are lower than those calculated from 24HR data, due to the larger average intake estimates from 7DR data. After controlling for differences in average intake estimates and adjusting household-level data for random measurement errors, the PoNI values from 7DR and 24HR data are remarkably close. This highlights the potential use of HCES data (conducted according to international agreed standards) for estimating the level of between-subject variability in usual nutrient intake in a population. The CVs from HCES could be used to compute the PoNI using average intake estimates from individual-level data; and the inadequacy of global nutrient supply using Supply and Utilization Accounts data.

Comparing cadence-based and machine learning based estimates for physical activity intensity classification: The UK Biobank

ObjectivesCadence thresholds have been widely used to categorize physical activity intensity in health-related research. We examined the convergent validity of two cadence-based intensity classification approaches against a machine-learning-based intensity schema in 84,315 participants (≥40 years) with wrist-worn accelerometers. DesignValidity study. MethodsBoth cadence-based methods (one-level cadence, two-level cadence) calculated intensity-specific time based on cadence-thresholds while the two-level cadence identified stepping behaviors first. We used an overlapping plot, mean absolute error, and Spearman's correlation coefficient to examine agreements between the cadence-based and machine-learning methods. We also evaluated agreements between methods based on practically-important-difference (moderate-to-vigorous-physical activity: ±20 min/day, moderate-physical activity: ±15, vigorous-physical activity: ±2.5, light-physical activity: ±30). ResultsThe group-level (median) minutes of moderate-to-vigorous- and moderate-physical activity estimated by one-level cadence were within the range of practically-important-difference compared to the machine-learning method (bias of median: moderate-to-vigorous-physical activity, −3.5, interquartile range [−15.8, 12.2]; moderate-physical activity, −6.0 [−17.2, 4.1]). The group-level vigorous- and light-physical activity minutes derived by two-level cadence were within practically-important-difference range (vigorous-physical activity: −0.9 [−3.1, 0.5]; light-physical activity, −1.3 [−28.2, 28.9]). The individual-level differences between the cadence-based and machine learning methods were high across intensities (e.g., moderate-to-vigorous-physical activity: mean absolute error [one-level cadence: 24.2 min/day; two-level cadence: 26.2]), with the proportion of participants within the practically-important-difference ranging from 8.4 % to 61.6 %. ConclusionsOne-level cadence showed acceptable group-level estimates of moderate-to-vigorous and moderate-physical activity while two-level cadence showed acceptable group-level estimates of vigorous- and light-physical activity. The cadence-based methods might not be appropriate for individual-level intensity-specific time estimation.

Updated Analysis of Radon Exposure and Lung Cancer Mortality in the Cohort of Newfoundland Fluorspar Miners (1950-2016).

The commercial mining of fluorspar in St. Lawrence Newfoundland began in 1933. Miners who worked underground were exposed to high levels of radon progeny, especially before ventilation was introduced into the mines in 1960. The mean cumulative radon exposure for underground miners in this cohort was 380.9 working level months (WLM). A series of studies of this cohort have characterized the increased risks of lung cancer mortality due to radon. We have extended the follow-up of this cohort an additional 15 years to provide additional insights on the risks of low levels of radon exposure, and the modifying effects of time since exposure, age at first exposure, attained age, duration of exposure, and cigarette smoking. The cohort consisted of 1,735 underground and 315 male surface miners who, combined, accrued 81,650 person-years of follow-up. The mortality experience of the cohort was determined from 1950-2016 through record linkage to Canadian national death data. Individual-level estimates of exposure to radon progeny, in WLMs, were determined for each year of employment. We compared the mortality experience of the underground miners to Newfoundland men using the standardized mortality ratio (SMR). Poisson regression models were fit to estimate excess relative risks (ERR) per 100 WLM. There were 236 lung cancer deaths identified, and of these, 221 occurred among underground workers. The SMR for lung cancer among underground miners compared to Newfoundland men was 2.67 (95% CI: 2.33, 3.04). The ERR per 100 WLM for lung cancer mortality, assuming a 5-year exposure lag, was 0.41 (95% CI: 0.23, 0.59). Attained age and time since exposure were important modifiers to the radon-lung cancer relationship. The joint relationship between smoking and radon on lung cancer risk was sub-additive, however, the smoking data were limited and available for only half of the cohort.

The Impacts of New York State's Paid Family Leave Policy on Parents' Sleep and Exercise.

To assess changes in young parents' health behaviors following implementation of New York State's Paid Family Leave Program (NYSPFL). We used synthetic control (N = 117,552) and difference-in-differences (N = 18,973) models with data from the nationally representative Behavioral Risk Factor Surveillance System (BRFSS) from 2011 to 2019 to provide individual-level estimates of the effects of NYSPFL on self-reported exercise in the past month and average daily sleep of adults aged 21-30 years living with one or more children under 18 years of age in New York and comparison states. Synthetic control model results indicate that the NYSPFL increased the likelihood of exercise in the past month among mothers, single parents, and low-income parents by 6.3-10.3% points (pp), whereas fathers showed a decrease in exercise (7.8 pp). Fathers, single parents, and those with two or more children showed increases in daily sleep between 14 and 21min per day. State paid family and medical leave laws may provide benefits for health behaviors among young parents with children under 18, particularly those in low-income and single-parent households.

Understanding the impact of radiotherapy fractionation on overall survival in a large head and neck squamous cell carcinoma dataset: a comprehensive approach combining mechanistic and machine learning models.

Treating head and neck squamous cell carcinomas (HNSCC), especially human papillomavirus negative (HPV-) and locally advanced cases, remains difficult. Our previous analyses of radiotherapy-only HNSCC clinical trials data using mechanistically-motivated models of tumor repopulation and killing by radiotherapy predicted that hyperfractionation with twice-daily fractions, or hypofractionation involving increased doses/fraction and reduced treatment durations, both improve tumor control and reduce late normal tissue toxicity, compared with standard protocols using 35×2 Gy. Here we further investigated the validity of these conclusions by analyzing a large modern dataset on 3,346 HNSCC radiotherapy patients from the University Health Network in Toronto, Canada, where 42.5% of patients were also treated with chemotherapy. We used a two-step approach that combines mechanistic modeling concepts with state-of-the-art machine learning, beginning with Random Survival Forests (RSF) for an exploratory analysis and followed by Causal Survival Forests (CSF) for a focused causal analysis. The mechanistic concept of biologically effective dose (BED) was implemented for the standard dose-independent (DI) tumor repopulation model, our alternative dose-dependent (DD) repopulation model, and a simple model with no repopulation (BEDsimp). These BED variants were included in the RSF model, along with age, stage, HPV status and other relevant variables, to predict patient overall survival (OS) and cause-specific mortality (deaths from the index cancer, other cancers or other causes). Model interpretation using Shapley Additive Explanations (SHAP) values and correlation matrices showed that high values of BEDDD or BEDDI, but not BEDsimp, were associated with decreased patient mortality. Targeted causal inference analyses were then performed using CSF to estimate the causal effect of each BED variant on OS. They revealed that high BEDDD (>61.8 Gy) or BEDDI (>57.6 Gy), but not BEDsimp, increased patient restricted mean survival time (RMST) by 0.5-1.0 years and increased survival probability (SP) by 5-15% several years after treatment. In addition to population-level averages, CSF generated individual-level causal effect estimates for each patient, facilitating personalized medicine. These findings are generally consistent with those of our previous mechanistic modeling, implying the potential benefits of altered radiotherapy fractionation schemes (e.g. 25×2.4 Gy, 20×2.75 Gy, 18×3.0 Gy) which increase BEDDD and BEDDI and counteract tumor repopulation more effectively than standard fractionation. Such regimens may represent potentially useful hypofractionated options for treating HNSCC.

Integration of datasets for individual prediction of DNA methylation-based biomarkers

BackgroundEpigenetic scores (EpiScores) can provide biomarkers of lifestyle and disease risk. Projecting new datasets onto a reference panel is challenging due to separation of technical and biological variation with array data. Normalisation can standardise data distributions but may also remove population-level biological variation.ResultsWe compare two birth cohorts (Lothian Birth Cohorts of 1921 and 1936 — nLBC1921 = 387 and nLBC1936 = 498) with blood-based DNA methylation assessed at the same chronological age (79 years) and processed in the same lab but in different years and experimental batches. We examine the effect of 16 normalisation methods on a novel BMI EpiScore (trained in an external cohort, n = 18,413), and Horvath’s pan-tissue DNA methylation age, when the cohorts are normalised separately and together. The BMI EpiScore explains a maximum variance of R2=24.5% in BMI in LBC1936 (SWAN normalisation). Although there are cross-cohort R2 differences, the normalisation method makes a minimal difference to within-cohort estimates. Conversely, a range of absolute differences are seen for individual-level EpiScore estimates for BMI and age when cohorts are normalised separately versus together. While within-array methods result in identical EpiScores whether a cohort is normalised on its own or together with the second dataset, a range of differences is observed for between-array methods.ConclusionsNormalisation methods returning similar EpiScores, whether cohorts are analysed separately or together, will minimise technical variation when projecting new data onto a reference panel. These methods are important for cases where raw data is unavailable and joint normalisation of cohorts is computationally expensive.

Individual- and supply-level macronutrient intakes are well correlated over a 50-year period (1961–2011) in 18 countries in Asia, North America, and Europe

Reliable information on dietary trends is essential. We compared individual-level dietary estimates for total energy, carbohydrate, fat, and protein intake over time with national supply data from the Global Expanded Nutrient Supply Model (186 paired estimates from 1961 to 2011, 18 countries). We hypothesized that supply data would overestimate individual measures and that the two measures would be weakly correlated. Individual- and supply-level estimates were compared using Spearman correlation coefficients and linear mixed-effect models were used to estimate the differences between measures. Overall, the correlations between individual- and supply-level measures were moderate for energy (rs = 0.34) and carbohydrate (rs = 0.39), strong for fat (rs = 0.85), and protein (rs = 0.69). Trends in total energy measured by individual-level surveys and total energy supply were positively correlated in 38.9% of countries, whereas trends in macronutrients aligned between estimates in most countries. Supply-level dietary data overestimated individual-level intakes, especially in higher income countries in Europe and in the United States. In the United States, supply-level data exceeded individual-level estimates by 26.3% to 29.9% for energy, carbohydrate, and fat, whereas protein estimates were similar between measures. In Europe, supply-level estimates overestimated individual-level intake by 19.9% for energy, 17.0% for carbohydrate, 13.7% for fat, and 7.7% for protein, whereas estimates for energy and macronutrients were similar in Asia. In Asia and lower income countries, our findings generally support the use of supply-level data in the absence of individual-level data, though this finding may be related to smaller sample size and differences in underlying national statistics that inform supply data.

Motor Performance During Sensorimotor Training for Airway Protection in Parkinson's Disease.

Cough dysfunction is highly prevalent in Parkinson's disease (PD) and associated with pneumonia, a leading cause of death. Although research suggests that cough can be volitionally upregulated, patterns of improvements that occur during cough skill training and potential correlates remain unexamined. Therefore, we sought to characterize changes to peak flow during cough skill training, examine whether early variability predicted motor performance trajectories during treatment, and explore the relationship between peak flow during cough skill training and motor learning on a similar but untrained task (i.e., reflex cough testing). This secondary analysis of treatment data from a randomized controlled trial included 28 individuals with PD who participated in five sessions of sensorimotor training for airway protection (smTAP). During this novel cough skill training, participants completed 25 repetitions of coughs targeting peak flow 25% above their baseline. Reflex and voluntary cough testing was performed pre- and posttreatment. Bayesian multilevel growth curve models provided group and individual-level estimates of peak flow during training. The magnitude and consistency of peak flow increased during cough skill training. Variability in peak flow during the first treatment session was associated with greater improvements to peak flow in later sessions. There was no relationship between changes to peak flow during cough skill training and motor learning. Individuals with PD improved the strength and variability of cough peak flow during cough skill training. These findings provide a clinically relevant characterization of motor performance during cough skill training and lend insight into potential correlates to guide future treatment paradigms.

Characterizing the emergence of amyloid and tau burden in Down syndrome.

Almost all individuals with Down syndrome (DS) will develop neuropathological features of Alzheimer's disease (AD). Understanding AD biomarker trajectories is necessary for DS-specific clinical interventions and interpretation of drug-related changes in the disease trajectory. A total of 177 adults with DS from the Alzheimer's Biomarker Consortium-Down Syndrome (ABC-DS) underwent positron emission tomography (PET) and MR imaging. Amyloid-beta (Aβ) trajectories were modeled to provide individual-level estimates of Aβ-positive (A+) chronicity, which were compared against longitudinal tau change. Elevated tau was observed in all NFT regions following A+ and longitudinal tau increased with respect to A+ chronicity. Tau increases in NFT regions I-III was observed 0-2.5 years following A+. Nearly all A+ individuals had tau increases in the medial temporal lobe. These findings highlight the rapid accumulation of amyloid and early onset of tau relative to amyloid in DS and provide a strategy for temporally characterizing AD neuropathology progression that is specific to the DS population and independent of chronological age. Longitudinal amyloid trajectories reveal rapid Aβ accumulation in Down syndrome NFT stage tau was strongly associated with A+ chronicity Early longitudinal tau increases were observed 2.5-5 years after reaching A.

Elevated decision uncertainty and reduced avoidance drives in depression, anxiety and substance use disorders during approach-avoidance conflict: a replication study.

Decision-making under approach-avoidance conflict (AAC; e.g., sacrificing quality of life to avoid feared outcomes) may be affected in multiple psychiatric disorders. Recently, we used a computational (active inference) model to characterize information processing differences during AAC in individuals with depression, anxiety and/or substance use disorders. Individuals with psychiatric disorders exhibited increased decision uncertainty (DU) and reduced sensitivity to unpleasant stimuli. This preregistered study aimed to determine the replicability of this processing dysfunction. A new sample of participants completed the AAC task. Individual-level computational parameter estimates, reflecting decision uncertainty and sensitivity to unpleasant stimuli ("emotion conflict"; EC), were obtained and compared between groups. Subsequent analyses combining the prior and current samples allowed assessment of narrower disorder categories. The sample in the present study included 480 participants: 97 healthy controls, 175 individuals with substance use disorders and 208 individuals with depression and/or anxiety disorders. Individuals with substance use disorders showed higher DU and lower EC values than healthy controls. The EC values were lower in females, but not males, with depression and/or anxiety disorders than in healthy controls. However, the previously observed difference in DU between participants with depression and/or anxiety disorders and healthy controls did not replicate. Analyses of specific disorders in the combined samples indicated that effects were common across different substance use disorders and affective disorders. There were differences, although with small effect size, in age and baseline intellectual functioning between the previous and current sample, which may have affected replication of DU differences in participants with depression and/or anxiety disorders. The now robust evidence base for these clinical group differences motivates specific questions that should be addressed in future research: can DU and EC become behavioural treatment targets, and can we identify neural substrates of DU and EC that could be used to measure severity of dysfunction or as neuromodulatory treatment targets?

Behavioral economic relationship between cannabis flower and concentrates: Evidence from simulated purchase tasks.

Cannabis users use different forms of cannabis, which are associated with distinct public health concerns. Policies that aim to regulate one specific form may have unintended impacts on other forms. This study examined the behavioral economic relationship between flower and concentrates, the two most common forms of cannabis. We surveyed 605 adult cannabis users (21+) who lived in one of the U.S. states that had legalized recreational cannabis by the time of interview in 2019. The participants completed simulated purchase tasks, which asked how much cannabis flower and concentrates they would purchase in the next 30 days at escalating prices. We estimated (a) demand indices and own-price elasticities using nonlinear exponential demand models and (b) group- and individual-level cross-price elasticities using log-linear demand models. The estimated rate of change in demand elasticity (α) was 0.00066 for cannabis flower (SE = 0.00002, p < .001) and 0.00058 for cannabis concentrate (SE = 0.00002, p < .001). Group-level cross-price elasticity estimate (slope = -0.075, SE = 0.0135, p < .001) indicated that cannabis flower and concentrates were weak complements. Individual-level cross-price elasticity estimates showed that flower and concentrates were treated as independent by 76.2% of the users, as complements by 19.0% of the users, and as substitutes by 4.8% of the users. The findings suggested that cannabis flower and concentrates were overall weak complements and for most adult cannabis users were treated as independent of each other. Price and tax policies regulating either cannabis form may have minimal impacts on the other form. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Distribution of gestational age at birth by maternal and infant characteristics in U.S. birth certificate data: Informing gestational age assumptions when clinical estimates are not available.

We aimed to describe the distribution of gestational age at birth (GAB) to inform the estimation of GAB when clinical or obstetric estimates are not available for perinatal pharmacoepidemiology studies. We estimated GAB (median, mode, mean, and standard deviation) and percentage born at each gestational week in groups based on plurality and other variables for live births in CDC's U.S. birth data. In 2020, 3 617 213 newborns had birth certificates with nonmissing GAB. Among singletons (3 501 693), median and mode GAB were both 39 weeks. Births with lower median GAB were from women with eclampsia (37 weeks) or receiving intensive care (37 weeks); newborns receiving intensive care (37 weeks); newborns with birth weight <2500 g (35 weeks), <1500 g (28 weeks), or <1000 g (25 weeks); and newborns not discharged alive (23 weeks). Among twins (112 633), median GAB was 36 weeks (mode, 37 weeks). Additional noteworthy groups were women with 0-6 prenatal visits (median, 34 weeks) or 7-8 prenatal visits (median, 35 weeks) or aged 15-19 years (median, 35 weeks). Some maternal and infant groups had distinct GAB distributions in the United States. This information can be useful in estimating GAB when individual-level clinical estimates are not available, such as in database studies of medication use during pregnancy.

Decision Making About Disease-Modifying Treatments for Relapsing-Remitting Multiple Sclerosis: Stated Preferences and Real-World Choices.

People with relapsing-remitting multiple sclerosis can benefit from disease-modifying treatments (DMTs). Several DMTs are available that vary in their efficacy, side-effect profile and mode of administration. We aimed to measure the preferences ofpeople with relapsing-remitting multiple sclerosis for DMTs using a discrete choice experiment and to assess which stated preference attributes correlate with the attributes of the DMTs they take in the real world. Discrete choice experiment attributes were developed from literature reviews, interviews and focus groups. In a discrete choice experiment, participants were shown two hypothetical DMTs, then chose whether they preferred one of the DMTs or no treatment. A mixed logit model was estimated from responses and individual-level estimates of participants' preferences conditional on their discrete choice experiment choices calculated. Logit models were estimated with stated preferences predicting current real-world on-treatment status, DMT mode of administration and current DMT. A stated intrinsic preference for taking a DMT was correlated with currently taking a DMT, and stated preferences for mode of administration were correlated with the modes of administration of the DMTs participants were currently taking. Stated preferences for treatment effectiveness and adverse effects were not correlated with real-world behaviour. There was variation in which discrete choice experiment attributes correlated with participants' real-world DMT choices. This may indicate patient preferences for treatment efficacy/risk are not adequately taken account of in prescribing. Treatment guidelines must ensure they take into consideration patients' preferences and improve communication around treatment efficacy/risk.

Overestimation of school-based deworming coverage resulting from school-based reporting.

Soil Transmitted Helminths (STH) infect over 1.5 billion people globally and are associated with anemia and stunting, resulting in an annual toll of 1.9 million Disability-Adjusted Life Years (DALYs). School-based deworming (SBD), via mass drug administration (MDA) campaigns with albendazole or mebendazole, has been recommended by the World Health Organization to reduce levels of morbidity due to STH in endemic areas. DeWorm3 is a cluster-randomized trial, conducted in three study sites in Benin, India, and Malawi, designed to assess the feasibility of interrupting STH transmission with community-wide MDA as a potential strategy to replace SBD. This analysis examines data from the DeWorm3 trial to quantify discrepancies between school-level reporting of SBD and gold standard individual-level survey reporting of SBD. Population-weighted averages of school-level SBD calculated at the cluster level were compared to aggregated individual-level SBD estimates to produce a Mean Squared Error (MSE) estimate for each study site. In order to estimate individual-level SBD coverage, these MSE values were applied to SBD estimates from the control arm of the DeWorm3 trial, where only school-level reporting of SBD coverage had been collected. In each study site, SBD coverage in the school-level datasets was substantially higher than that obtained from individual-level datasets, indicating possible overestimation of school-level SBD coverage. When applying observed MSE to project expected coverages in the control arm, SBD coverage dropped from 89.1% to 70.5% (p-value < 0.001) in Benin, from 97.7% to 84.5% (p-value < 0.001) in India, and from 41.5% to 37.5% (p-value < 0.001) in Malawi. These estimates indicate that school-level SBD reporting is likely to significantly overestimate program coverage. These findings suggest that current SBD coverage estimates derived from school-based program data may substantially overestimate true pediatric deworming coverage within targeted communities. NCT03014167.

Prenatal exposure to tap water containing nitrate and the risk of small-for-gestational-age: A nationwide register-based study of Danish births, 1991–2015

BackgroundPrenatal nitrate exposure from household tap water has been associated with increased risk of fetal growth restriction, preterm birth, birth defects, and childhood cancer. We aim to examine the association between maternal consumption of drinking-water nitrate during pregnancy and small-for-gestational-age (SGA) in a nationwide study of Danish-born children, as only one prior study has examined this association. MethodsWe linked individual-level household estimates of nitrate in tap water and birth registry data to all live singleton Danish births during 1991–2015 from Danish-born parents where the mother resided in Denmark throughout the pregnancy. Exposure was both binned into four categories and modeled as an ln-transformed continuous variable. SGA was defined as the bottom 10% of births by birth weight per sex and gestational week. Multiple logistic regression models with generalized estimating equations were used to account for siblings born to the same mother while controlling for relevant confounders. ResultsIn the cohort of 1,078,892 births, the median pregnancy nitrate exposure was 1.9 mg/L nitrate. Compared to the reference group (≤2 mg/L), we found an increased risk of SGA in the second category (>2–5 mg/L) (OR = 1.04, 95% CI: 1.03–1.06) and third category (>5–25 mg/L) (OR = 1.02, 95% CI: 1.00–1.04) but not in the highest (>25 mg/L). There was strong (p = 0.002) evidence of an increase in SGA with nitrate in the model with continuous exposure (OR = 1.02, 95% CI: 1.01–1.04 per 10-fold increase in nitrate). Results were robust when restricting to households with nitrate levels at or below the current Danish and European Union regulatory drinking water standard (50 mg/L nitrate). ConclusionsOur findings suggest that exposure from nitrate in household tap water, even below current regulatory standards, may increase risk of SGA, raising concerns of whether current allowable nitrate levels in drinking water protect children from SGA.

Estimating individual-level pig growth trajectories from group-level weight time series using machine learning

Recent years have seen an increase in the use of advanced data collection systems to improve management in pig farms, however, the cost of implementing such systems remains a barrier to many. Some of the most useful data for farm management are on an individual pig level, which requires investment of labour and expense of systems such as RFID tagging. In this paper an alternative solution is proposed, where pen-level pig weight measurements (i.e. measurements not linked to a specific pig identity), which are significantly cheaper to obtain, are used to estimate individual-level growth trajectories using machine learning. We evaluated our method on group-level data that were simulated from individual-level weight data, which thus allowed for the verification of the accuracy of the trajectory predictions using a root mean squared error score. The on-average best performing predictor for this task was based on a Random Forest model which achieved a best-case score of 2.00 Kg per pig and a worst-case score of 2.45 Kg per pig depending upon the simulated conditions that were used, thus showing that these predicted trajectories could be accurate enough for commercial usage. The usefulness of the predicted trajectories was then evaluated through a growth prediction downstream task whereby the future weights of pigs two weeks into the future were predicted using a variety of machine learning models. It was found that, on average, the growth prediction models that utilised the predicted trajectories outperformed a method that utilised group-level weights directly by 1.42 Kg root mean squared error but were still outperformed by the models that utilised RFID growth trajectories by 0.83 Kg on average. These results show the potential of this method as an alternative to both RFID and photo-based systems for individual-level data estimation.

Probabilistic HIV recency classification-a logistic regression without labeled individual level training data.

Accurate HIV incidence estimation based on individual recent infection status (recent vs long-term infection) is important for monitoring the epidemic, targeting interventions to those at greatest risk of new infection, and evaluating existing programs of prevention and treatment. Starting from 2015, the Population-based HIV Impact Assessment (PHIA) individual-level surveys are implemented in the most-affected countries in sub-Saharan Africa. PHIA is a nationally-representative HIV-focused survey that combines household visits with key questions and cutting-edge technologies such as biomarker tests for HIV antibody and HIV viral load which offer the unique opportunity of distinguishing between recent infection and long-term infection, and providing relevant HIV information by age, gender, and location. In this article, we propose a semi-supervised logistic regression model for estimating individual level HIV recency status. It incorporates information from multiple data sources - the PHIA survey where the true HIV recency status is unknown, and the cohort studies provided in the literature where the relationship between HIV recency status and the covariates are presented in the form of a contingency table. It also utilizes the national level HIV incidence estimates from the epidemiology model. Applying the proposed model to Malawi PHIA data, we demonstrate that our approach is more accurate for the individual level estimation and more appropriate for estimating HIV recency rates at aggregated levels than the current practice - the binary classification tree (BCT).

PRSet: Pathway-based polygenic risk score analyses and software.

Polygenic risk scores (PRSs) have been among the leading advances in biomedicine in recent years. As a proxy of genetic liability, PRSs are utilised across multiple fields and applications. While numerous statistical and machine learning methods have been developed to optimise their predictive accuracy, these typically distil genetic liability to a single number based on aggregation of an individual's genome-wide risk alleles. This results in a key loss of information about an individual's genetic profile, which could be critical given the functional sub-structure of the genome and the heterogeneity of complex disease. In this manuscript, we introduce a 'pathway polygenic' paradigm of disease risk, in which multiple genetic liabilities underlie complex diseases, rather than a single genome-wide liability. We describe a method and accompanying software, PRSet, for computing and analysing pathway-based PRSs, in which polygenic scores are calculated across genomic pathways for each individual. We evaluate the potential of pathway PRSs in two distinct ways, creating two major sections: (1) In the first section, we benchmark PRSet as a pathway enrichment tool, evaluating its capacity to capture GWAS signal in pathways. We find that for target sample sizes of >10,000 individuals, pathway PRSs have similar power for evaluating pathway enrichment as leading methods MAGMA and LD score regression, with the distinct advantage of providing individual-level estimates of genetic liability for each pathway -opening up a range of pathway-based PRS applications, (2) In the second section, we evaluate the performance of pathway PRSs for disease stratification. We show that using a supervised disease stratification approach, pathway PRSs (computed by PRSet) outperform two standard genome-wide PRSs (computed by C+T and lassosum) for classifying disease subtypes in 20 of 21 scenarios tested. As the definition and functional annotation of pathways becomes increasingly refined, we expect pathway PRSs to offer key insights into the heterogeneity of complex disease and treatment response, to generate biologically tractable therapeutic targets from polygenic signal, and, ultimately, to provide a powerful path to precision medicine.

Who Is Leaving and Why? The Dynamics of High-Quality Human Capital Outflows

This study proposes a unified, dynamic framework based on Turnover Event Theory (TET) to evaluate the effects of dismissals, layoff announcements, and voluntary turnover on subsequent work unit voluntary turnover. Applying our approach to 1,620 retail stores over 22 months, we show that modeling exit events as a dynamic and interdependent system adds to our ability to predict subsequent human capital outflow. Dismissals had the weakest total effect on subsequent voluntary turnover, layoff announcements had the strongest and most immediate effects, and voluntary turnovers had moderate but lasting effects. We also find that these three exit reasons each exhibit a distinct pattern of subsequent turnover intensity and longevity. Based on the characteristics of work units in our setting, our results correspond to a cumulative worker-level average effect of 0.17 quits following dismissals, 0.23 quits following quits, and 2.2 quits following a layoff. We find that these “multiplier” effects are concentrated among workers of similar performance: high-performer exits beget high-performer quits, just as low-performer exits beget low-performer quits. Our findings suggest analyses offering individual-level estimates of turnover will generally underestimate the broader, work-unit level consequences of individual exit events on similar workers.