5) A ten-year follow-up was conducted to determine if someone experiencing food allergies is at increased risk of coronary heart disease. The following covariates were considered in the analysis: AGE = age at enrollment, BMI = body mass index, and SMK = smoking. The results for the standard, all possible subsets, and the crude are listed below. Covariates aRR 95% Cl Width AGE, BMI, SMK (std) 4.10 (1.7, 8.8) AGE, BMI 2.70 (1.6, 7.8) AGE, SMK 4.12 (1.4, 9.5) BMI, SMK 2.50 (1.8, 6.2) AGE 4.01 (1.6, 8.7) BMI 2.50 (1.4, 7.5)| SMK 4.16 (1.5, 8.6) None (crude) 1.02 (0.8, 1.2) a) Is there evidence of joint confounding? Explain b) Is there evidence of marginal confounding? Explain c) Besides the standard, list the candidate subgroups for control. d) Which of the candidate subgroups (including the standard) corresponds to the most valid estimate? Explain your choice. 6) In this cohort study we examined the impact on being unvaccinated on the risk of hospitalization due to severe COVlD-19. All participants were diagnosed with COVlD-19. Age is a potential confounding variable. The following tables give the distribution of hospitalized and non-hospitalized COVlD-19 subjects by vaccination status and age group. Conduct a complete stratied analysis including estimates of the crude, stratified, and adjusted estimates, the Breslow-Day test for interaction, as well as evaluation of criteria for potential confounding by age group. Briey summarize the results including what estimate you will present and why and a nal statement about the observed association. Aged 50+ Aged