Dietary Risks
- Relative risk estimation
- Linking FAO food groups to nutrient composition
- Estimating SDs
- Estimating Missing Country-Years of FAO & Sales Data (ST-GPR)
- Age Splitting of FAO & Sales Data
- Adjustment to Optimal Assessment Method (Mixed Effects Regression)
- Estimating missing country year data (ST-GPR)
- Age pattern estimation based on metabolic indicators
- Global Age Pattern of Consumption (DisMod)
- TMREL
- SBP-CVD RR by age
- Ensemble Distribution Fitting
- Estimating Mean-SD relationship
- PAF Calculation
Drug Use
High fasting plasma glucose
- Risk factor estimation
- Directly calculated population attributable fractions
High LDL Cholesterol
- Age-sex splitting
- Calculate PAFs
- BRFSS crosswalk
- LDL crosswalk
- Standard deviation model
- Covariate selection process
- Meta-analysis of relative risks
- ST-GPR
- TMREL
High systolic blood pressure
- Age-sex splitting
- Calculate PAFs
- BRFSS crosswalk
- Standard deviation model
- Covariate selection process
- Usual blood pressure adjustment
- ST-GPR
- Meta-analysis of relative risks
- TMREL
Household Air Pollution from Solid Fuels
- Exposure
- Calculate proportion of individuals using solid fuel and fuel types
- Calculate proportion of households using solid fuel and fuel types
- Run MR-BRT crosswalk calculation to generate a multiplier of shift due to family size adjustment
- Crosswalk data for family size adjustment
- ST-GPR
- Squeeze proportion of individuals using fuel types to proportion of individuals using solid fuel
- Mapping
- Subtract off estimated ambient PM2.5 exposure by location year
- Run a mixed effects regression to estimate country-, year-specific PM2.5 concentration for females and living areas
- Calculate the ratio of personal exposure to female exposure for both men and children
- Scale female PM2.5 exposure for male and child exposures
- Cataract
- Relative risk
- PAF
Impaired kidney function
- Albuminuria DisMod-MR 2.1
- Calculate PAFs
- Interpolation
- Prepare data for MR-BRT
- Relative risks with MR-BRT
Intimate Partner Violence
- Directly calculated population attributable fractions
- Exposure
- Dismod-MR 2.1: Time series of ever been partnered
- ST-GPR
- ST-GPR model with no age-smoothing
- MR-BRT Network Analysis to calculate adjustment factors for alternate case definitions
- Adjust data with alternate case definitions
- Adjustment to data points from sources where population is ever partnered or currently partnered women only, by multiplying estimates by age-specific fraction of women in a relationship
- Age-split data reported in aggregate age groups
- Holt Method to fore/back-cast time series of intial ST-GPR draws
- Risk Factor Estimation
- Cumulative risk approach of HIV\AIDS due to IPV
- Relative Risks
Iron deficiency
- RF calculation of exposure based on anemia causal attribution
- PAF Calculation
- PAF Mediation adjustment
Lead Exposure
- Application of mediation factors
- Calculate PAFs
- Fit ensemble distribution on microdata
- Age-sex splitting
- Dismod ODE
- Calculate CLBI
- Calculate shift in IQ
- Crosswalk lead to arithmetic means
- ST-GPR
Low birthweight and prematurity
- Mean birthweight is regressed on birthweight
- predicts 1000 draws of mean gestational age and birthweight for every l/y/s, for one age group.
- ST-GPR
- mean STGPR models and paired with ga and bw results to get optimal SD
- helper functions so that mean STGPR models and paired with ga and bw results to get optimal SD
- MLE: Mean GA from
- Joint GA + BW microdata is used to model a global copula
- Helper functions so that joint GA + BW microdata can be used to model a global copula
- This script takes the univariate weights, mean, SD, and global copula to produce 100 draws of the joint distribution for every l/y/s for birth
- Rakes the joint distributions to exactly match the STGPR prevalences at
- Uses Relative Risks to age birth cohort and produce prevalence in a 500g x 2wk grid for birth, EN, LN and 28 days
- Aggregates the 500g x 2wk bins into
- Meta-analyze literature and microdata relative risk surfaces to create location-specific RRs
- Calculate all-cause mortality odds using logistic regression with dummy variables for each 500g and 2wk category, by sex and location and Smooth mortality odds categories using Gaussian Process Regression
- Divide all categories by lowest mortality risk to calculate relative risk per 500g and 2wk category, by sex and location
- crosswalks data types for gestational age 28 weeks
- crosswalks data types for gestational age 37 weeks
- squares the dataset so that all observations for all lbw, ga points are present for all other points
- Calculate PAFs using exposure, relative risks, and TMREL
