+ Reply to Thread
Results 1 to 4 of 4

Thread: Path analysis, mediational model (including repeated measures)?

  1. #1
    Points: 17, Level: 1
    Level completed: 33%, Points required for next Level: 33

    Posts
    2
    Thanks
    1
    Thanked 0 Times in 0 Posts

    Path analysis, mediational model (including repeated measures)?




    Hello, everyone!
    I am having an interesting research question but unfortunately, I cannot come up with the right statistical model to test it. I hope someone can help me out.

    Here is some background information:
    Typically, comorbid depression in Alzheimer's disease (AD) was thought of as secondary, stemming from the difficulties that come along with the dementia symptoms. Now, there is evidence that describes depression as a primary symptom of AD. This line of evidence tries to link a specific etiological biomarker directly to depression.

    What I like to do in this project is to see if an intervention that reduces depression also A) reduces the concentration of the biomarker, and B) if this reduction then also results in an overall decrease in dementia symptoms. To not over-complicate things I have attached an image depicting the exact relations. Though I have never encountered path analysis in any of my classes this seems to be a fitting scenario.

    So again, the question I cannot answer is how do I best test this with a statistical model and how can I introduce repeated measures into this design?

    Any advice is appreciated! Thanks in advance.
    -Lukas
    Attached Images  

  2. #2
    Omega Contributor
    Points: 38,253, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,989
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Path analysis, mediational model (including repeated measures)?

    Many questions but little time.


    1.) All patients will have AD?
    2.) you would intervene after DX of AD and Depression?
    3.) You can quantify Dementia?
    4.) You can have a sufficient sample of subjects not on anti-dep as a control group?
    5.) Does early onset of AD be controlled for or sampled around?
    6.) Any genetic testing
    7.) Criteria for determining baseline mild cognitive issues etc.?
    8.) Is it possible to randomize intervention to balance treatment group characteristics?
    9.) Anti-dep has direct arrow to biomarker, is that legit and if so it has no sign.
    10.) Assuming Amyloid symptoms influence depression and vice versa, that will be extremely hard to address!
    11.) There are some many anti-dep, do you have a medicinal pathophysiologic mechanism why one should work?
    12.) How do you make sure there is drug adhere and track it?
    13.) How do you make sure subjects attend visits and missing data are not dependent on study variables?
    14.) Do you have a rationale for timing of repeated visits and how long would it take to conduct such a study?
    15.) Is it easy to use standard depression screening tools on AD / demented subjects?


    Great graph it definitely helped to explain your study!
    Stop cowardice, ban guns!

  3. The Following User Says Thank You to hlsmith For This Useful Post:

    Lukas123 (09-07-2017)

  4. #3
    Points: 17, Level: 1
    Level completed: 33%, Points required for next Level: 33

    Posts
    2
    Thanks
    1
    Thanked 0 Times in 0 Posts

    Re: Path analysis, mediational model (including repeated measures)?

    Hi hlsmith and thanks for the reply!

    Here are the answers to your questions:
    1.) All patients will have AD?
    -> Half the sample has AD and half the sample has normal cognition.
    2) You would intervene after DX of AD and Depression?
    -> The study is retrospective. Some participants may have been treated for depression others may have not. Same story for AD.
    3) You can quantify Dementia?
    -> Yes, I can quantify dementia with certain assessment tools pretty reliably.
    4) You can have a sufficient sample of subjects not on anti-dep as a control group?
    -> Yes, I believe there should be enough participants not on any anti-depressant medication.
    5) Does early onset of AD be controlled for or sampled around?
    -> I can control for early-onset AD as the APOE genotype is also recorded.
    6) Any genetic testing
    -> Yes, APOE is recorded for most participants I assume.
    7) Criteria for determining baseline mild cognitive issues etc.?
    -> There are assessment tools used that can give me pretty decent indications of the cognitive function subject are in.
    8) Is it possible to randomize intervention to balance treatment group characteristics?
    -> No, as it is retrospective I cannot randomize intervention unfortunately.
    9) Anti-dep has direct arrow to biomarker, is that legit and if so it has no sign.
    -> This arrow is not too important. The idea was that SSRIs may also have a direct effect on the biomarker that cannot be explained through the effect of alleviating depression only.
    10) Assuming Amyloid symptoms influence depression and vice versa, that will be extremely hard to address!
    -> I agree. I guess for the sake of this study I would have to rely on a unidirectional relationship where depression influences amyloid and not vice versa.
    11) There are some many anti-dep, do you have a medicinal pathophysiologic mechanism why one should work?
    -> There is preliminary evidence that SSRI's have an effect on amyloid. Why they work I am not sure.
    12) How do you make sure there is drug adhere and track it?
    -> I believe this is also recorded however I would need to look into that.
    13) How do you make sure subjects attend visits and missing data are not dependent on study variables?
    -> The sample I am working with is not optimal. At this point, I cannot make any statements regarding missingness patterns. I believe that I will have to deal with that once I have taken a closer look at the data and after investigating correlations.
    14) Do you have a rationale for timing of repeated visits and how long would it take to conduct such a study?
    -> The timing is also very crude. Subjects are assessed annually for a maximum of 6 times.
    15) Is it easy to use standard depression screening tools on AD / demented subjects?
    -> I wouldn't say it is easy but there are standard tools that one can use and make comparisons with.

    I hope this clarifies things a bit more. Thanks a lot for now and if there is any further information you need please let me know.

  5. #4
    Omega Contributor
    Points: 38,253, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,989
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Path analysis, mediational model (including repeated measures)?


    You need to switch anti-dep and depression in the graph, since del occurs before anti-dep.

    Also, you will need a big sample to control for all of the potential confounders.
    Stop cowardice, ban guns!

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats