Reference no: EM132519266

POPH90014 Epidemiology - The University of Melbourne

There is no need to cite the source.

Question 1. What percent of global Disability Adjusted Life Years (DALYs) did depression account for in 2017?

What percent of global Years Lived with Disability (YLDs) did depression account for in 2017?

Why are these figures so different?

The following questions relate to a Mendelian Randomization study of physical activity and incident depression.
Choi KW,Chen CY,Stein MB,Klimentidis YC,Wang MJ,Koenen KC,Smoller JW;Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Assessment of Bidirectional Relationships Between Physical Activity and Depression Among Adults: A 2-Sample Mendelian Randomization Study. JAMA Psychiatry. 2019 Apr 1;76(4):399-408
The journal article is available in the Assignment 2 folder of Readings Online. You will need to read the abstract and the first three paragraphs of the Introduction. You might also want to read the Discussion. The rest of the article is complex, and you don't need to read it to answer the questions.

In the section of the abstract labelled Importance, the authors wrote:
"Increasing evidence shows that physical activity is associated with reduced risk for depression, pointing to a potential modifiable target for prevention. However, the causality and direction of this association are not clear; physical activity may protect against depression, and/or depression may result in decreased physical activity."

Question 2. What types of studies would give rise to uncertainty about the "direction of this association"? Explain your answer.

Question 3. In the article's Introduction section, the authors state: "prospective studies have demonstrated associations between higher levels of physical activity and decreased risk for later depression". These prospective studies are cohort studies.

Why did the authors write the sentence that way (i.e., using the term "association") rather than writing something like "prospective studies have demonstrated that higher levels of physical activity decrease risk for later depression"?

We will return to this paper later in the assignment.

The next set of questions are about a systematic review of cohort studies of physical activity and incident depression. The journal article is available in the Assignment 2 folder of Readings Online. You will also need the supplementary material for the article.
Schuch FB, Vancampfort D, Firth J, et al. Physical activity and incident depression: a meta-analysis of prospective cohort studies. Am J Psychiatry. 2018;175:631-648.

Question 4. What is the research question in PICOS format? Paste the table below into your assignment and fill in the blanks. There is no need to use sentence format.

For this question, ignore the material about "moderators" (age at baseline, geographical location, sex, length of follow-up, study quality, number of covariates used in the model, study sample size, and total person-years) at the end of the introduction.

Question 5. Give two reasons why this eligibility criterion"who were free of depression or depressive symptoms at baseline"is important. Justify each reason.

Question 6. The authors restricted the review to studies with at least one year's follow-up. They said that less than one year's follow-up "was not considered a sufficient time frame for risk and protective factors toexert a meaningful influence on depressive symptoms".
What could be another reason to exclude these studies? Justify your answer. [2 marks]

Figure 1 shows odds ratio for each study comparing the category of highest level of physical activity for that study with the least active category for that study. This is frequently done in meta-analyses (the example used in the lecture did this too). The pooled odds ratio of 0.837 indicates that the most active people had lower odds of depression than the least active people (i.e. there was an inverse association).

Information bias

Question 7. In the discussion, the authors state that "self-report questionnaires are associated with recall bias". They have not used recall bias in the way you have been taught.
How do we know they have used the term differently?
What is the name for the type of measurement error in measuring physical activity in these cohort studies?
In which direction would this measurement error bias the pooled estimate from the meta-analysis - towards the null or away from the null?Comment on the implications of this bias with respect to the observed inverse association between physical activity and risk of depression.

Confounding

Question 8. Comment on the degree of difference between the crude pooled odds ratio (reported in Table 2) and the adjusted pooled odds ratio.Why are they different?

If you had to use the results to convince a government to invest in physical activity programs to prevent depression, which would you use and why?

Question 9. People with high BMI have increased risk of depression. Twelve of the studies adjusted for body mass index (BMI). Discuss the arguments for and against doing this. Explain your answer. Draw DAGs to illustrate your answer.

Selection bias

The authors use the term selection bias in a different way from what you have been taught. When they refer to selection bias, they are referring to the issues raised in the questions about eligibility criteria above. Unfortunately, differences in use of terminology are common in medical research.

Question 10. This question is about table S3 in the supplementary material. Is selection bias due to attrition(loss to follow up) likely? Explain your answer.

Question 11. What further information about attrition should the authors have included in table S3 to help you reach a judgment about the risk of bias due to attrition? Explain your answer.

Aspects of the systematic review

Question 12. Was the protocol for the systematic review registered?
If so, where was the registration reported?

Question 13. Is there a PRISMA flow diagram? If so, where was it reported?

Question 14. Just looking at Figure 1, do the odds ratios appear reasonably consistent?

Question 15. What measure of heterogeneity did they report?
How would you describe the degree of heterogeneity?

Question 16. The authors conducted statistical testing to assess possible publication bias (small study effects) and showed there was evidence of publication bias. They further showed using some statistical analyses that are beyond our scope that once this was taken into account, the pooled estimate of the odds ratio did not change much. They did not include a funnel plot. A funnel plot of the results presented in Figure 1 is shown below.

Is the funnel plot consistent with their conclusion that there was publication bias? Explain your answer. If you think there is publication bias, in which direction is the Pooled Estimate of the odds ratio biased?

Question 17. Based on this review, and your previous answers, how strong is the evidence that physical activity has a protective causal effect on risk of depression? Make sure you also consider the following items from Bradford Hill in your response - strength of association, consistency, and temporal sequence.

We return now to the Mendelian Randomisation (MR) paper
The study was described as bi-directional because they had two research questions:
a) Do genetic variants that predict physical activity reduce risk of depression and
b) Do genetic variants that predict depression affect physical activity.

Question 18. Draw two DAGs to illustrate the assumptions underlying this MR study. Each DAG should include these four variables: genetic variants, exposure, outcome and confounders.
a. In the first DAG, the genetic variants are variants that predict physical activity, the exposure is physical activity and the outcome is depression.
b. In the second DAG, the genetic variants are variants that predict depression, the exposure is depression and the outcome is physical activity.
In each DAG, you should assume that the genetic variants only affect the outcome through the exposure - this is one of the key assumptions of MR. How do your DAGs illustrate the strength of the MR approach?

Question 19. Did they find evidence that physical activity reduces the risk of incident depression?
Did they find evidence that depression affects the amount of physical activity?
What do these findings tell you about the directionality of the association between physical activity and depression?

Question 20. About 100,000 participants in the UK Biobank wore wrist-worn accelerometers to measure their physical activity. You can read about it here if you wish, but it isn't necessary. Why do you think that the analysis showed that physical activity measured by accelerometry lowered risk of depression, but self-reported physical activity did not?

Question 21. Triangulating the evidence from the two sources (cohort studies and the MR study), how strong is the evidence that physical activity can prevent depression?

Generalisability

Question 22. Comment on the generalisability of the results from the systematic review with respect to regions of the world and age.

Design a study

Question 23. What other type of study could be done to give more definitive evidence on whether physical activity can prevent depression? Explain your answer and give a brief description of the design of such a study including the research question, the eligibility criteria for participants, and other important design issues.You do not need to consider sample size. Your answer could be in dot points.

Referencing: APA

Note: Need Only Question 8, 16, 19, 20, 21 Question

Attachment:- Epidemiology.rar