Hidden Meteosensitivity? Looking For Sleep Diaries To Analyze

Discussion in 'Possible causes and predisposing factor discussion' started by Circadino, Jan 3, 2023.

  1. Circadino

    Circadino Established Member

    Messages:
    4
    Hi there,

    I'm diagnosed with a circadian rhythm disorder, non-24, cfs, the whole bunch.

    Recently, I analyzed my sleep diary for possible connections with weather conditions. To my surprise, I found very strong correlations, but they were cyclical, which is why I haven't noticed them, yet.

    Then I analyzed the sleep diary of two and in one case also the body temperature and found even stronger correlations of a very regular cyclical nature. Both I and the others were surprised. You can follow the debate in the N24 subreddit, where you will also find several graphs. Here's one of the most striking ones:

    [​IMG]

    It does kind of look like as if I had struck gold with the meteosensitivity. But I'm all but sure about it. For instance, I don't know yet how a healthy person's cycle is correlated to the weather. The initial suspicion though is clearly there. Besides the analysis of healthy individual's sleep patterns, I'd like to also analyze the patterns more cfs/n24 patients: Sleep/wake times, body temperature, heart rate etc.

    In case you do keep track of your vitals and sleep patterns and you're interested in knowing more about a possible hidden meteosensitivity, I would be happy to analyze your data. All I need is a table with the numbers and the location(s) where you were at the time.

    Thank you very much!
     
    DokaGirl and Peter Trewhitt like this.
  2. Trish

    Trish Moderator Staff Member

    Messages:
    55,414
    Location:
    UK
    Hi @Circadino, welcome to the forum.
    Perhaps you could share more about how you analysed your data so others can analyse their own in the same way if they want to.

    I can think of all sorts of possible confounding factors that may confuse the picture for many people with ME/CFS. Just a few to start the discussion:
    Medication effects, especially sleep meds.
    Menstrual cycles and/or hormone use in people who menstruate.
    Inaccuracies in recording sleep data - for example my fitbit usually gets my sleep wrong, adding hours when I am resting but not asleep.
    Light exposure deliberately reduced because of ME/CFS light sensitivity.

    Also we generally advise members not to share personal, including medical, data, and not to reveal things like email addresses and location unless they are completely happy for these to be public.
     
    Sean and Circadino like this.
  3. Circadino

    Circadino Established Member

    Messages:
    4
    Of course. Here's how to replicate it:
    1. You need your own data and for the same time period the weather data from your location or region.
    2. Both data sets go as colums into the same table of a spreadsheet with each row having matching time/date stamps.
    3. Now create moving averages for both your numbers and the weather data; the number of values per average is not so important.
    4. When you have that, you can create moving correlations between the moving averages. The correlation has to span over at least 30 values; the more, the better.
    When you have that, then you can play around with the size of the averages and correlations as well as different metrics, until you found the one with the most striking correlations and freuence etc. The by far most time consuming part is (2), because the time stamps of the data usually does not fit by default. You have to go through that manually.

    Thank you for your input. So far, the sample size is rather small and male only, which makes the following statements preliminary:
    • Medication indeed shows up in the graphs. Ritalin intake appears to have inverted the correlation in one case. Overall and given the body's natural needs, it seems impossible to cover all ends with medication. This means that either the sleep times or the wake-up times will give it away.
    • Inaccuracies of the data recording can pose a problem, but so far the results have shown to be robust since it's mostly about averages with an emphasis on phase inversions. Meaning, it's not about learning everything about the entire dataset, but only about its crucial segments.
    • The graph which I posted above comes from a patient who developed his on light therapy and is applying it on himself. The yellow segment on the graph shows when he applied the light therapy. It had a profound impact on the correlation between his sleep times and daily sun hours.
    Of course, I do respect the privacy of any data that I receive.
     
    Trish likes this.
  4. Creekside

    Creekside Senior Member (Voting Rights)

    Messages:
    1,218
    There are confounding factors too. Maybe on cloudy days you eat more of some kind of food, and on hot days avoid some. Activity levels may vary with the weather too. Weather affects airborne spore and pollen counts, and biting insect activity. Scents differ too, which can have physiological effects, as does the spectral quality and intensity of the light. So, you might see a correlation between weather and sleep, but precisely what factor is responsible will take more work.
     

Share This Page