In depth
IgG4-related disease and blood tests
The blood test this disease is named for is a useful one, and it isn't the whole answer to anything. How useful it is depends entirely on where the line gets drawn. This page gives you both of the figures rather than picking whichever one of the two reads better.
What the research found.
One review pooled 23 studies and 6048 people with IgG4-related disease. Serum IgG4 caught 85 percent of cases, on an interval from 78 to 90 percent, and it correctly cleared 93 percent of people without the disease, from 90 to 95 percent. The area under the curve was 0.95, from 0.93 to 0.97.
One review pooled 9 case-control studies, 1235 people with the disease, and 5696 controls. At the lower cut-off, serum IgG4 caught 87.2 percent of cases, on an interval from 85.2 to 89.0. It cleared 82.6 percent of people without the disease, from 81.6 to 83.6.
That review doubled the cut-off and ran the numbers through again. Catching fell to 63 percent, from 60.0 to 66.0, while clearing rose to 94.8 percent, from 94.1 to 95.4. Its authors write that high specificity comes with a significant sacrifice in sensitivity, which is their own wording.
Hao and colleagues, Medicine, 2016
One review searched 167 publications for a usable clinical practice guideline in this disease, and it found one. It named four things that don't exist here, which are shared classification criteria, formal guidance on diagnosis, treatment recommendations backed by evidence, and any measure of disease activity or damage. That's thinner than any other disease covered on this site.
What the number means.
| Test | What it measures | How it performs | What it can't do |
|---|---|---|---|
| Serum IgG4, pooled across 23 studies | The blood level of one antibody subclass | Caught 85 percent, cleared 93 percent, area under the curve 0.95 | Stand alone. About one case in seven reads normal |
| Serum IgG4, at the lower cut-off | That test with the line drawn lower | Caught 87.2 percent, cleared 82.6 percent | Clear people confidently. It reads positive in some who don't have the disease |
| Serum IgG4, at double that cut-off | That test with the line drawn higher | Caught 63 percent, cleared 94.8 percent | Catch cases. More than a third of them read normal |
| Biopsy | A piece of tissue looked at under a microscope | This site holds no source on how it performs | Be described here until a reference is supplied |
| Any activity or damage score | A number for how active the disease is | One review reports that no such score exists | Be quoted to you, because there isn't one |
A reference range is not a target. Where a range on this page differs from the one your laboratory prints, the difference is explained above and sourced below. Where no trial has tested a target, this page says so rather than offering one.
The test this disease is named for
IgG4-related disease is named after an antibody, and that antibody is one subclass of the immunoglobulin G family. A blood test measures how much of it you have circulating. So the first thing most people do is look up that test, which is a fair place to start.
It's also the place where people expect a good deal more certainty than the test can give them. This page sets out what two reviews found about it. One of the two is more favorable than the other, and both are here, because picking the better-looking one would be a choice rather than a summary.
What the larger review found
One review pooled 23 studies, which between them held 6048 people with IgG4-related disease, making it the larger of the two here. The test caught 85 percent of cases, where the true value ranges from 78 to 90 percent. That's the headline figure, and it's the friendlier of the two reviews.
It correctly cleared 93 percent of people without the disease, on an interval from 90 to 95 percent. It also reports an area under the curve of 0.95, from 0.93 to 0.97, which is a single number for how well a test separates two groups across every possible cut-off. A perfect test scores 1.0 and a useless one scores 0.5.
Read those three figures together. This is a good test, and it isn't a test that settles the question by itself. Catching 85 percent means that about one case in seven reads normal, which is the part people miss.
What the second review found
One review pooled 9 case-control studies, which between them held 1235 people with the disease and 5696 controls. It asked the question the first review asked and drew the line in two different places. At the lower cut-off, the test caught 87.2 percent of cases, where the true value ranges from 85.2 to 89.0.
It cleared 82.6 percent of people without the disease, on an interval from 81.6 to 83.6. Then it doubled the cut-off and ran the numbers through again, where catching fell to 63 percent, from 60.0 to 66.0. Clearing rose to 94.8 percent, from 94.1 to 95.4.
Its authors put the conclusion plainly, and their wording is worth having. They call serum IgG4 a modestly effective marker, and they write that high specificity comes with a significant sacrifice in sensitivity. That's the whole argument of this page in one sentence of theirs.
Why the cut-off changes everything
This is the part worth taking away from the page and remembering. Every blood test draws a line somewhere, and where that line sits decides which mistake the test makes more often. There's no way to escape that trade, in this disease or in any other one anybody has studied.
Draw the line low and the test catches nearly everybody who has the disease, while calling more healthy people positive. Draw it high and fewer healthy people test positive, while more real cases get missed. Those are the only two directions available to anybody who has to draw it.
The second review shows that trade happening inside one single dataset of its own. The catching figure fell from 87.2 percent to 63 percent, while the clearing figure rose from 82.6 percent to 94.8 percent. One test, one set of people, and a different line drawn through them.
So a sensible question at your appointment is which cut-off your own laboratory uses. That question isn't a challenge to anybody who ordered the test. It's the single thing that decides what your own result means to you, and it's easy to ask.
Why the two reviews differ
The two reviews disagree most on clearing, which is the figure people quote least often. One says 93 percent and the other says 82.6 percent at the lower of its two cut-offs. That's a real difference, and it deserves stating plainly rather than being smoothed into an average the study never reported at all.
They pooled different studies, and that's the likeliest reason for the difference between the two figures. The first took 23 studies and 6048 patients, while the second took 9 case-control studies with 5696 controls between them. Different sets of studies give different answers, and that isn't a fault in either.
Neither figure is the right one and the other the wrong one. Treat the pair of them as the range this test lives in, and don't choose between the two. That's a more honest reading than picking whichever one happens to read better to you.
What is still missing here
One review searched 167 publications for a clinical practice guideline in this disease and found a single usable document. That one is a consensus statement written by experts drawn from several different specialties. It's the only document of its kind anywhere in the field at present.
It then named four things that do not exist in this disease at all. There are no shared criteria for sorting cases, no formal guidance on diagnosis, and no treatment recommendations backed by evidence. There is also no score for how active the disease is or how much damage it has done.
That last absence has a practical consequence worth spelling out. Without an activity score, no study can measure whether something made the disease better or worse. So anything sold to you as treating this illness has cleared no bar, because there isn't a bar here yet.
This site also holds no source on how a biopsy performs in this disease. Tissue is how the diagnosis usually gets confirmed, so that's a real absence and it's recorded in the review notes on this page. A visible absence is more use to you than a confident sentence with nothing behind it.
What to do with a number you already have
If you're holding a result in your hand right now, the useful questions are short ones. Ask which cut-off it was measured against, ask what the scan showed, and ask whether any tissue was taken from you. Those three settle a good deal more between them than the number does.
A raised level with a matching scan and matching tissue is a different situation from a raised level on its own. So is a normal level in somebody whose scan looks typical for this particular disease. The test is one input among three rather than the answer to the question by itself.
Nothing on this page tells you what your own result means for you in particular. It tells you what the test does across several thousand people, which is a different question from what yours means. Your team is the one that answers the first.
Common misconceptions.
Myth. A normal IgG4 level rules this disease out.
Reality. It doesn't rule it out, and the size of the problem depends on the cut-off used. One review pooled 23 studies and 6048 patients and found the test catching 85 percent of cases, so about one case in seven read normal. At the higher cut-off in the other review, catching fell to 63 percent, and more than a third of cases read normal there. A normal result makes the diagnosis less likely without closing the question.
Myth. A raised IgG4 level makes the diagnosis.
Reality. It doesn't make the diagnosis on its own. One review found the test clearing 93 percent of people without the disease, on an interval from 90 to 95, and the other found 82.6 percent at the lower cut-off. Neither cleared everybody, so some people without this disease do test positive, which is why the result gets read beside the scan and the biopsy.
Myth. There's one agreed threshold for the test.
Reality. One review ran its pooled data at two different cut-offs and got two different tests out of it. At the lower one it caught 87.2 percent and cleared 82.6 percent, and at double that it caught 63 percent and cleared 94.8 percent. Its authors call serum IgG4 a modestly effective marker, and they say high specificity comes with a significant sacrifice in sensitivity. So ask which cut-off your own laboratory used.
Myth. The two reviews disagree, so one of them must be wrong.
Reality. They pooled different studies, which is where the difference comes from. One took 23 studies and 6048 patients, and the other took 9 case-control studies with 5696 controls. Different sets of studies give different answers, and that spread is information rather than an error, so treat 93 percent and 82.6 percent as the range this test lives in.
Myth. There must be a guideline for a disease with a name like this.
Reality. There's almost nothing here, and one review shows how little. It searched 167 publications and found a single document usable as a clinical practice guideline, then named four things that don't exist. There are no shared criteria for sorting cases, no formal guidance on diagnosis, no treatment recommendations backed by evidence, and no score for disease activity.
Cautions specific to this condition.
- Take a normal result as one piece, not an all-clear. One review found the test missing about one case in seven, and the other missed more than a third at the higher cut-off.
- Ask which cut-off your laboratory used. The same test catches 87.2 percent at one line and 63 percent at another.
- A raised level on its own isn't a diagnosis. One review cleared 93 percent of people without the disease and the other cleared 82.6 percent, so neither cleared everybody.
- This page gives no threshold in units and names no treatment. Ask your team what your own number was measured against.
- Expect practice to vary more here than in commoner diseases. One review searched 167 publications and found a single usable guideline.
- Nothing on this page tells you what your own result means. Your team reads it beside the scan and the tissue.
Discuss any change with the rheumatologist who manages your care. Nothing here replaces that conversation.
Questions patients ask.
What is the IgG4 blood test?
It measures the blood level of one subclass of antibody, called IgG4, and it's the test this disease is named for. It's usually the first thing people look up, and it isn't the diagnosis. One review pooled 23 studies and 6048 patients and found it catching 85 percent of cases and clearing 93 percent of people without the disease.
How good is it?
Reasonable rather than decisive, on the two reviews here. One found it catching 85 percent and clearing 93 percent, with an area under the curve of 0.95, and another found 87.2 percent and 82.6 percent at the lower cut-off. Its authors describe serum IgG4 as a modestly effective marker, so it's a useful test that doesn't settle the question on its own.
Can I have this disease with a normal IgG4?
Yes, and it isn't at all rare here. In the larger review the test caught 85 percent of cases, so about one case in seven read normal, and at the higher cut-off in the second review catching fell to 63 percent. That means more than a third of people with the disease read normal there, so a normal result doesn't close the question.
Why does the cut-off change the answer so much?
Because every test trades one kind of mistake for another kind. Drawing the line higher means fewer healthy people test positive and more real cases get missed, and one review showed that happening directly. Doubling the cut-off moved clearing from 82.6 percent up to 94.8 percent, and moved catching from 87.2 percent down to 63 percent.
Are there agreed rules for diagnosing this?
Not yet, and one review says so plainly enough. It searched 167 publications and found a single document usable as a clinical practice guideline, then listed four things that don't exist here. There are no shared criteria for sorting cases, no formal guidance on diagnosis, no treatment recommendations backed by evidence, and no score for disease activity.
What else gets used alongside the blood test?
Imaging and a tissue sample, in most practice. This site holds no source on how well a biopsy performs in this disease, so this page won't put a number on it. What the reviews here do show is that the blood test alone isn't enough, and your team puts the result together with the scan and the tissue.
Somebody quoted me a cut-off number. Should I trust it?
Ask where it came from before you trust it. Cut-offs in this disease vary and the performance changes with them, which is the whole point of the second review here. A number with a paper behind it is worth having, and a number with nothing behind it is worth less than nothing, because it sounds settled when the field isn't.
References.
- Xu WL; Ling YC; Wang ZK et al. Diagnostic performance of serum IgG4 level for IgG4-related disease: a meta-analysis. Sci Rep. 2016;6:32035. 10.1038/srep32035Journal Article
- Hao M; Liu M; Fan G et al. Diagnostic Value of Serum IgG4 for IgG4-Related Disease: A PRISMA-compliant Systematic Review and Meta-analysis. Medicine (Baltimore). 2016;95:e3785. 10.1097/MD.0000000000003785Journal Article
- Iaccarino L; Talarico R; Scirè C et al. IgG4-related diseases: state of the art on clinical practice guidelines. RMD Open. 2019;4:e000787. 10.1136/rmdopen-2018-000787Systematic literature review for clinical practice guidelines in IgG4-related disease
This page gathers the published research on this subject into one place. The studies behind it were published between 1989 and 2026, and every figure links to the paper it came from. Those studies were peer reviewed. This summary of them was not. Dr. Sarah Luebker is reviewing these pages one at a time and has not reached this one yet, so it carries no medical review date and nothing here is her opinion or her advice to you. Each page gets updated as she reaches it. It is here in the meantime because the science is worth having in one organized place that is easy to find and easy to read. Talk to your own clinician before acting on any of it.
Across the site