1. Using the key below, classify each of the following descriptions of epidemiologic studies according to (40 marks):
a. Study type (1 mark each)
• Descriptive Cross-sectional Study
• Analytic Ecological Study
• Analytic Cross-Sectional Study
• Case-Control Study
• Retrospective Cohort Study
• Prospective Cohort Study
• Longitudinal study
b. Give reasons for your choice of study type (2 marks each).
c. Name the measure of frequency (incidence or prevalence including type of prevalence (point or period) and type of incidence (cumulative or rate)) or the measure of association (rate ratio, odds ratio etc) that would be appropriate to use for that study type (1 mark each).
i. A study examined whether changes in the share of people living in relative poverty in a municipality were linked to changes in the use of antidepressants based on data from prescription register of the Social Insurance Institution of Finland in several Finnish municipalities between 1995 and 2010. Findings indicate more young adult females used antidepressants in municipalities where relative poverty had increased, suggesting an association between relative poverty and antidepressant use.
ii. Analysis of existing data from the Health Insurance Database for Catastrophic Illness Patients in Taiwan found women with breast cancer who were morphine users were 1.24 times more likely to suffer from type two diabetes mellitus than non-morphine users (95% CI = 1.04–1.49). The study included 31,112 women with breast cancer without a history of type two diabetes mellitus during the period 2000–2005. Women were divided into groups based on their exposure status - those who received morphine and those who did not receive morphine as a part of their breast cancer treatment to determine the risk of type two diabetes mellitus.
iii. A study investigated whether living near a high-hazard dump site compared to a low-hazard dump site increased the risk of having a low birth weight child. Women who had a low birth weight child were recruited and matched to women who had normal birth weight child. Data was then collected on proximity to high and low hazard dump sites. The study found that for women living near a high-hazard dump site while pregnant, 181 had a low birth weight child and 4,268 had a normal weight child. For women living near a low-hazard dump site while pregnant, 126 had a low birth weight child and 4,268 had a normal weight child.
iv. A study examining the rate of congenital birth defects in a large maternity hospital in Melbourne collected data on all babies who were born at the hospital in 2013. Of the 2,000 babies born, 70 had a congenital defect.
v. A sample of 75,236 healthy male construction workers without a history of ischemic stroke or CHD was followed from 1989 to 2004. Events of ischemic stroke and CHD were identified through health or death records at five year intervals. The study found 1,884 new cases of CHD and 739 new cases of ischemic stroke.
vi. Researchers were interested in the association between parity and breast cancer. To investigate the possible association they recruited 1,500 women who were mothers and 1,500 women who were not mothers at age 30. Both groups of women were then followed for 30 years to assess the rate of breast cancer.
vii. To study a presumed inverse relationship between black tea consumption and cardiovascular disease, investigators randomly selected a sample of 13,500 subjects, 50-64 years of age, and queried them about tea consumption and current treatment for cardiovascular disease. No significant association was found between drinking black tea and having cardiovascular disease.
viii. Researchers investigated the consumption of chocolate and whether or not it was positively associated with overweightness among elderly men aged 65 years or more. The sample included 250 men who were classified as overweight or obese and 250 men who were classified as within normal weight range. Men who were classified as underweight were excluded from the study. The men were asked about their diet, including the consumption of sugary sweets such as chocolates and cakes, and their participation in physical activity.
ix. One thousand healthy women aged between 45 and 50 years underwent annual health screening. The women were followed for eight years. At the end of the first year, 15 women had developed high blood pressure.
x. A total of 825 insulation workers employed between 1941 and 1944 were identified from the personnel records of three large insulation manufacturing plants in the south eastern U.S. in 1975. During the period 1941-1975, 26 deaths from lung cancer were discovered among the workers. Only six lung cancer deaths, however, were reported among a comparable group of 700 co-workers who did not work with insulation during the same time period. The investigators had postulated that exposure to the insulation material increases the risk of lung cancer.
2. A longitudinal study of the onset of arthritis followed 2,000 initially healthy women, 70-74 years of age, over several years. Every two years the investigators interviewed the women to see if they had had a diagnosis of arthritis since the previous interview. The results were as follows:
• 7 new arthritis cases were diagnosed between the baseline and the first interview,
• 8 new arthritis cases were diagnosed between the first and second interviews,
• 9 new arthritis cases were diagnosed between the second and third interviews, and
• 10 new arthritis cases were diagnosed between the third and fourth interviews.
In addition, the investigators noted that:
• 8 women withdrew between the first and second interviews
• 10 women withdrew between the second and third interviews (17 marks).
a. Based on this information calculate the cumulative incidence of arthritis among this population (4 marks).
b. Based on this information, calculate the incidence rate of arthritis and interpret your answer (10 marks)
c. Of the two measures calculated, which is the most appropriate measure of frequency and why? (3 marks)
3. Researchers were interested in investigating the health effects, and in particular primary brain cancer, of mobile phone use. They recruited 1,500 people age 18 years who then underwent a health check and 10 were found to have primary brain cancer. Ten years later (at age 28) all 1,500 people attended a second health check and another 43 people had developed primary brain cancer. The 10 people initially identified with primary brain cancer have received ongoing treatment for their primary brain cancer (16 marks).
a. What was the prevalence of primary brain cancer among the sample at age 18? (4 marks)
b. How many people were at risk of developing primary brain cancer at the start of the 10 year period? (1 mark)
c. What was the incidence of primary brain cancer among this sample? (4 marks)
d. Is this a measure of cumulative incidence or incidence rate? (2 marks)
e. On the basis of these findings, the researchers decided to test the hypothesis that using handheld mobile phones is related to the risk of primary brain cancer. A case-control study was conducted in five medical centres between 2007 and 2010 using a structured questionnaire. A total of 469 men and women aged 18 to 80 years with primary brain cancer and 422 matched controls without brain cancer were recruited into the study. The table below presents the study findings.
i. Name the measure of association that would be most appropriate to see if there is an association between mobile phone use and primary brain cancer (1 mark).
ii. Calculate the appropriate measure of association (3 marks).
iii. Interpret your result (1 mark).
4. In an industry employing 20,000 people, 5,000 were employed in areas where they were exposed to pesticides, while the remaining 15,000 were not exposed. At the beginning of the study, all employees were free from disease. The entire population of 20,000 was followed for 20 years to determine whether exposure to pesticides increased the risk of developing a particular disease. (18 marks).
a. Calculate the appropriate measure of frequency of the disease for (8 marks):
i. The exposed workers (3 marks)
ii. The unexposed workers (2 marks)
iii. All the workers (2 marks)
iv. Interpret your results for i - iii together (1 mark)
b. Calculate the appropriate measure of association and interpret your answer (5 marks)
c. How much disease in the exposed workers could be due to their pesticide exposure? Interpret your answer (5 marks).
5. A causal relationship between cigarette smoking and lung cancer was first suspected in the 1920s on the basis of clinical observations. To test this apparent association, Richard Doll and Austin Bradford Hill conducted a case-control study in Britain in 1947 comparing the smoking habits of lung cancer patients with the smoking habits of other patients. Data for the case-control study were obtained from hospitalised patients in London and vicinity. Initially, 20 hospitals, and later more, were asked to notify the investigators of all patients admitted with a new diagnosis of lung cancer. These patients were then interviewed concerning smoking habits, as were controls selected from patients with other disorders (primarily non-malignant) who were hospitalised in the same hospitals at the same time. What makes this a case-control study? (3 marks)
6. What is the difference between attributable risk and attribute risk percent? (2 marks)
7. Researchers are interested in assessing a possible association between pet ownership and a reduction in depressive symptoms among people living in urban cities. What is the best study design to assess this possible association? (1 mark)
8. Name three characteristics of observational studies. (3 marks)