E2040 Week 8 Practice 1. The 5-year incidence of cardiovascular disease (CVD) in relation to smoking status was determined in a population sample of 1000 men, 18 years and older. At baseline, 40% of the men were classified as being current smokers and 60% as being non-smokers. At the end of the follow-up, 80 CVD events had occurred among the smokers and 60 events among the non-smokers. a) Enter the information given in the example above into a 2x2 table CVD No CVD Total Smoking No smoking Total b) Calculate the cumulative incidence (incidence risk) of CVD for smokers and non-smokers c) Calculate and interpret the risk ratio 2. The Whitehall II study of British civil servants started in 1985. There have been nine waves of data collection to date. The numbers given below are for illustration only. a. Common mental disorder was captured by the GHQ30 – a 30-item self-completion tool that counts the number of symptoms of common mental disorder. Here, a score of 5 or more out of 30 is classified as a case of common mental disorder. The GHQ30 has been administered at every wave. The table below gives numbers of participants with CMD at waves 1 (1985-8) and 5 (1991-1993). No CMD at wave 5 CMD at wave 5 Total No CMD at wave 1 4277 781 5058 CMD at wave 1 1174 738 1912 Total 5451 1519 6970 a) Calculate the incidence risk (cumulative incidence) of common mental disorder. Hint: exclude the prevalent cases at baseline. b. The number of participants developing new CMD is broken down by employment grade in the following table. Employment grade Number with CMD / total N Incidence risk Risk Ratio High 235/1690 Medium 388/2496 Low 158/872 a) Calculate the incidence risk of CMD in each grade. b) With high grades as the reference group, calculate the risk ratio of developing CMD in the other grades. 3. At baseline (wave 1) the distribution of participants by civil service grade of employment (a marker of socioeconomic position) was as follows: Employment Grade Wave 1 number of participants (%) High grade 3028 (29) Medium grade 4943 (48) Low grade 2337 (23) By wave 5, the distribution was as follows: Employment Grade Wave 5 number of participants (%) High grade 1446 (41) Medium grade 1558 (44) Low grade 521 (15) Note that 477 participants were promoted from medium to high grade, 168 were promoted from low to medium grade and 7 were promoted from low to high grade. a) What do you notice about the number of participants in each grade at the two waves? b) And what do you notice about the distribution of participants by grade at the two waves? c) Besides promotion, what reasons can you suggest for these differences between waves? The Whitehall II study set out to examine whether people in lower socioeconomic positions developed poorer health. d) What is the implication of the loss to follow-up by grade for looking at this research question? 4. The table below shows participation in Whitehall II at wave 5 according to the health status at wave 1. The health status measure is self-rated health in the last year based on the question “How was your health in the last year?” . Health status Took part at wave 5 – Number of participants (%) Did not take part -Number of participants (%) Very good 2654 (81) 632 (19) Good 3236 (77) 973 (23) Average 1612 (71) 648 (29) Poor 319 (68) 152 (32) Very poor 27 (59) 19 (41) a) What does this table tell you? b) What is the implication of this for studying risk factors for poor health? EXTRA QUESTIONS 5. Read the abstract, then answer the questions below ‘Social avoidance’ was measured using binary responses (0,1) to the following four items: ‘I don’t speak until someone speaks to me first’ ‘I wish I were not as shy as I am’ ‘I stay by myself because I fear doing something wrong’ ‘I am often not in on the group’s gossip’ Participants were assigned to one of four groups with score 0, 1, 2, 3 or 4 Answer the following questions. a) Did the study address a clearly focused question? b) What was the population? c) Risk factor(s) over specified time period? d) What outcome(s) were studied? e) What was the measure of effect? f) What was the main result of the study? g) What further evidence would you want to see on this research question before reaching a conclusion on this association? Source: http://www.sciencedirect.com/science/article/pii/S1047279707001445# 6. The table below describes a hypothetical case-control study that examined the relationship between prostate cancer risk and beer consumption. 5 prostate cancer cases university clinics completed a questionnaire shortly after diagnosis. Controls were recruited from patients in the surgical outpatient clinics of the same hospitals. Cases and controls completed a questionnaire about the frequency and amount of beer consumed in the last year. Cases Controls Non-drinkers 51 88 Rarely drinking beers 53 101 Drinking few beers often 86 125 Drinking beer a lot and often 96 74 Total 286 388 a) Calculate the frequency of different levels of beer consumption for cases and controls. What is the interpretation? b) Calculate the odds ratio for each category of beer consumption compared to non-drinkers. Interpret. c) What other information would we need? d) Can the way the cases and controls were recruited affect the results? How and why? e) What other ways of recruiting probands for a similar study we can consider and what are the advantages and disadvantages?