Dear body: What the fuck?

[kiya]

A week ago my TSH was four.

On the recheck to evaluate meds levels, it's two.

My hypochondriacal tendencies are Not Assisted by this sort of thing.

Comments

[ardaniel]

I would be wondering about the quality control of the lab in question at that point.

[kiya]

Yeah, that's what keshwyn said.

I shall be, well, keeping an eye on this.

[leanne-opaskar.livejournal.com]

I think I agree -- and you might want to get a second opinion at this point. Particularly since I recall that the last time you had testing, it varied pretty widely too.

Hope all stays well. *hugs*

[chinders.livejournal.com]

*offers hugs*

[whispercricket]

If you're just venting, then *hug*.

If not...I was curious, so I did some research (yeah, surprise :} ).

First question is whether they used the same lab both times, and whether they have the same ranges.

Second, here are a few interesting things I found (note, I don't know how reliable these sites are, because I just pulled them up quickly between lunch and a meeting :} ):
Metafilter's thread on fluctuating TSH:
http://ask.metafilter.com/95783/Fluctuating-TSH-thyroid-normal-or-weird

- Do you know if you've been tested for the Hashimoto's disease antibodies (peroxidase and thyroglobin, it seems)?

Recommended lab tests and a discussion of fluctuating thyroid (really don't know how reliable this site is at all):
http://www.stopthethyroidmadness.com/recommended-labwork/
(Search down also for the section on TSH that talks about fluctuations and limitations of the test)

Must run, but the search on "tsh fluctuation week" seemed to be getting some information that might be helpful.

[whispercricket]

Also, wasn't your first test in the morning and the second later in the day? That supposedly may make a difference (get tested in the morning).

ETA:Okay, now I found another site that says the levels don't fluctuate, so...time for a real expert on that one. :)

[ardaniel]

None of MeFi's resident doctors appear in that thread, at least the ones I know of, so I'm not sure exactly how valid the information is. (Most of the docs tend to stick to their specialties, though, and I don't think we have any endocrine specialists on board. One neurologist and a handful of GPs...)

[whispercricket]

*nod* I tend to take this stuff as potential questions to ask doctors / specialists, more than anything else, but it might be better than just getting back "we have no idea". :)

However, I am finding some stuff on PubMed about circadian rhythm of TSH, so that part at least may be credible (links just in case someone is interested):

- http://www.ncbi.nlm.nih.gov/pubmed/2588938
- http://www.ncbi.nlm.nih.gov/pubmed/17201800 (interesting - they found a better update of l-thyroxine when taken at night than in the morning)
- http://www.ncbi.nlm.nih.gov/pubmed/16148346 (I think this one supports looking at free T3 / T4 over TSH more than just not changing the TSH reference range, myself, but IANAD)
-- "INTERVENTION: The time of phlebotomy is important, because the TSH level varies throughout the day, with early morning values greater than later ones, and is accentuated by sleep deprivation, strenuous exercise, or working during the night or evening shifts. Repeated measurements in the same individual vary considerably over months."
- http://jcem.endojournals.org/cgi/content/full/86/7/3284 (interesting study, tiny control group)

Granted, I just have access to abstracts and not full studies, so I can't see how well any of these were done.

[syndramise.livejournal.com]

Testing only a week later to check levels seems wonky to begin with. According to my endocrinologist, it takes at least six weeks to get any reliable readings after an adjustment because the levels won't stabilize for at least a month. That could be the entirety of the problem right there.

But, since TSH levels *should* be lower than 3, at least the fluctuation is in the right direction. That's a yay, right?

[kiya]

I slightly misstated -- this is all unmedicated, she wanted to get a second check to see what my meds *should* be.