Our research

When Your Symptoms Have No Diagnosis

Some veterans carry a pile of symptoms that never adds up to one named illness: exhaustion that sleep does not fix, pain that moves around the body, a stomach that will not settle, memory and concentration that slip. Doctors run tests and find nothing. Claims like these have their own rule, 38 CFR 3.317, and their own pattern at the Board of Veterans' Appeals. We read every published Board decision to find out what that pattern is.

The short version: claims pled as an undiagnosed illness are more likely to be sent back for more medical development than to be decided at all. That is not a prediction about any one claim. It is what the record shows, and the sections below give the exact counts and say what each one is a share of.

What the Board actually did

We counted every issue in the published Board record where the claim itself was pled as an undiagnosed illness or a medically unexplained chronic multisymptom illness, and separately the three conditions the regulation names by title. They are counted as issues, not as veterans: one Board decision often rules on several issues at once.

Claimed as an undiagnosed illness

The condition itself was pled under 38 CFR 3.317 as undiagnosed or medically unexplained.

54.5%
sent back for more development, of all 28,448 issues the Board handled
Granted2,196
Denied10,107
Sent back (remanded)15,501
Other (withdrawn, dismissed)644
17.8% granted counting only the 12,303 issues the Board actually decided (granted or denied). That is a different measure from the one above: it leaves every remand out.

The three named conditions

Chronic fatigue syndrome, fibromyalgia and irritable bowel syndrome: diagnosable, but 3.317 treats them as medically unexplained.

36.2%
sent back for more development, of all 38,505 issues the Board handled
Granted7,311
Denied14,694
Sent back (remanded)13,926
Other (withdrawn, dismissed)2,574
33.2% granted counting only the 22,005 issues the Board actually decided (granted or denied). That is a different measure from the one above: it leaves every remand out.

Those three, pled as undiagnosed

The overlap between the first two columns. It is counted separately and never added to either.

39.3%
sent back for more development, of all 3,666 issues the Board handled
Granted492
Denied1,631
Sent back (remanded)1,441
Other (withdrawn, dismissed)102
23.2% granted counting only the 2,123 issues the Board actually decided (granted or denied). That is a different measure from the one above: it leaves every remand out.

Three groups, three separate counts. They overlap on purpose, so they are never added together, and no combined figure exists anywhere on this page. Every number counts ISSUES, not veterans: one Board decision often rules on several.

Baseline, current as of the last data build: across 28,448 issues pled as an undiagnosed illness, 54.5 percent were sent back to the VA for more development rather than decided. Of the ones the Board did decide, 17.8 percent were granted. The three named conditions, claimed under their own diagnostic codes, were granted at 33.2 percent of decided issues. Those two figures use different denominators and are not two readings of the same thing.

Why these claims get sent back

A remand is not a loss. It means the Board could not decide the case on the file in front of it and ordered the VA to go get something first. Because more than half of this group ends there, what the Board asks for is the most useful thing in the whole record. We read the remand instructions themselves and grouped what they ordered.

What the Board orderedShare of these remandsDecisions
Medical opinion or addendum ordered 76.7% 4,715
New VA examination ordered 64.7% 3,975
VA treatment records to obtain 19.8% 1,219
Service records to obtain 18.8% 1,153
Private treatment records to obtain 16.1% 988
Corrective notice to send (VCAA) 11.4% 698
Hearing to be scheduled 7.7% 473
Prior remand not complied with (Stegall) 7.4% 454
Social Security records to obtain 3.1% 193
Statement of the case to issue 2% 120
Evidence requested from the veteran 1.7% 103

Read from the remand instructions in 6,145 decisions that sent one of these claims back. One remand usually orders several things at once, so these do not add up to 100 percent, and 528 remands ordered something this reader could not label.

Medical development dominates: about three quarters of these remands ordered a medical opinion and about two thirds ordered a new examination. The Board is not usually sending these back over paperwork. It is sending them back because nobody has yet written down a medical answer it can rely on.

What the denials turned on

When the Board did deny one of these claims, its reasoning usually failed at one of three points: no current disability, no qualifying service or in-service event, or no medical link between the two.

What the denial turned onShareIssues
No medical link to service (nexus) 57%5,002
No current diagnosis or qualifying disability 31.4%2,754
No qualifying in-service event or service 11.7%1,024

Of 10,107 denied issues, 8,780 carried a readable reason; the shares above are of those 8,780, not of all denials and not of all claims.

The medical link is the wall these claims hit most often, which lines up with what the remands ask for. Categories here are the reason the Board itself gave, read from the decision, and the shares are of denials where a reason could be read, not of every claim.

Which conditions these claims end up rated under

An undiagnosed illness has no code of its own. VA rates it by analogy, under whichever existing condition on the rating schedule comes closest to the symptoms. That is why the list below is a mix of fatigue, headache, stomach, skin, sleep and joint codes.

Condition it was rated underIssuesSent backGranted, of decided
Chronic fatigue syndrome (CFS) DC 6354 2,53760.7% 19.6%
Migraine DC 8100 1,95057.2% 18.6%
Dermatitis or eczema DC 7806 1,22157.7% 11.7%
Sleep Apnea Syndromes (Obstructive, Central, Mixed) DC 6847 1,06860.4% 9.7%
Irritable bowel syndrome (IBS) DC 7319 86566.6% 41.7%
Fibromyalgia (fibrositis, primary fibromyalgia syndrome) DC 5025 74255.8% 27.1%
Lumbosacral or cervical strain DC 5237 54479% 14%
Gastroesophageal reflux disease DC 7206 54366.7% 12.9%
Asthma, bronchial DC 6602 47573.1% 17.7%
Posttraumatic stress disorder DC 9411 44565.6% 15.7%
Knee, other impairment of DC 5257 40883.1% 21.7%
Degenerative arthritis, other than post-traumatic DC 5003 35464.1% 6.5%
Bursitis DC 5019 33283.4% 26.4%
Hypertensive vascular disease (hypertension and isolated systolic hypertension) DC 7101 32356.3% 8.5%
Unspecified depressive disorder DC 9435 31953.3% 13.6%
Chronic obstructive pulmonary disease DC 6604 29560.7% 15%
Other specified anxiety disorder DC 9410 29470.4% 11.4%
Emphysema, pulmonary DC 6603 25968.7% 30.7%
Major or mild neurocognitive disorder due to Alzheimer's disease DC 9312 19397.4% not enough decided
Arteriosclerotic heart disease (coronary artery disease) DC 7005 19171.2% 12.2%

The codes claims pled as an undiagnosed illness were rated under, from 54 codes with at least 50 issues. A code needs at least 25 decided issues before a grant rate is shown at all, because a code can carry hundreds of issues that are nearly all remands.

Has it changed over the years?

Yes, and not in the direction most people expect. The share of these claims that gets sent back instead of decided has gone up, not down, across the three eras of the record.

When the Board decided itIssuesShare sent back
Before 201012,129 49.1%
2010 to 20187,951 55.8%
2019 onward (AMA era)8,368 61%

These links are not tied to one era the way some Board statistics are: the rule has been in place for decades, so a decision from the 1990s still shows how the Board treats this kind of claim. The rising remand share is worth knowing because it says the medical-development problem has not been solved by time.

The symptom clusters are recognized, not imagined

Veterans in this group are often told, in one way or another, that nothing is wrong with them. The federal medical literature says otherwise. The pattern has a name, a case definition, and a research history.

Chronic multisymptom illness

The CDC case definition describes a chronic pattern in two or more of these areas: fatigue, mood and thinking (including memory and concentration trouble), and musculoskeletal pain. Symptoms have to persist for at least six months. It is defined by the pattern, not by a lab result.

What VA recognizes

VA's own public-health material treats medically unexplained illnesses in Gulf War veterans as real and service-related under the law, and names chronic fatigue syndrome, fibromyalgia and functional gastrointestinal disorders such as irritable bowel syndrome among them.

Researchers have also published narrower case definitions, notably the Kansas criteria developed by Steele, which sort veterans by symptom domain and exclude other explanations. Those are published in the medical literature; we describe that they exist and point to them rather than reproducing them here.

Nothing on this page is a diagnosis or a screening tool. Whether any of these definitions fits a particular person is a medical question for a clinician.

How the rule is built (38 CFR 3.317)

The regulation creates two separate doors for veterans with qualifying Southwest Asia service, and the numbers at the top of this page show they do not behave alike.

  • Door one, a true undiagnosed illness: objective signs or other verifiable indications of a chronic disability that no doctor can attribute to a known diagnosis. It gets rated by analogy to whichever condition on the schedule fits the symptoms.
  • Door two, the three named conditions: chronic fatigue syndrome (DC 6354), fibromyalgia (DC 5025) and irritable bowel syndrome (DC 7319). These have names and diagnostic codes, but the regulation still treats them as medically unexplained.

Two requirements sit underneath both doors: the disability has to be chronic, generally meaning six months or longer, and it has to reach at least a 10 percent level. The rule also carries a manifestation deadline that VA has extended more than once, so the current date belongs to the regulation, not to this page.

The classic reason a door-one claim fails is that a doctor does attribute the symptoms to a known diagnosis, which takes the claim out of 3.317 entirely and sends it back to ordinary service connection. Read the regulation in full on our 38 CFR 3.317 page, and see the filing side of it in the Gulf War Illness claims guide.

Common questions

My claim was remanded. Is that bad?
No. A remand means the Board wants more evidence before it decides, most often a medical opinion or a new examination. It is the most common single result in this group. It does add time.
Why is the grant rate different in the two places you show it?
Because they count different things. One is a share of every issue, including the ones sent back and withdrawn. The other counts only issues the Board actually granted or denied. We show both and label which is which rather than picking the flattering one.
Do I have to have served in the Gulf War?
38 CFR 3.317 applies to veterans with qualifying service in the Southwest Asia theater, and separately to certain service in Afghanistan, Syria, Djibouti and Uzbekistan. The regulation itself defines the locations and dates, so check it directly.
Is there a test that proves an undiagnosed illness?
No, and that is the point of the rule. It exists precisely because the usual proof, a diagnosis with a cause, is not available. What the record shows is that a written medical opinion still ends up being the thing these claims turn on.
Can you tell me whether my claim will be granted?
No. Nothing on this page is a prediction for an individual claim, and we are not permitted to prepare or advise on one. These are counts of what the Board has done across published decisions. A VA-accredited representative can help with your own case for free.

Sources and method: outcome counts are our own reading of the published Board of Veterans' Appeals record, at issue grain, with each denominator named where it is used. Remand groupings come from the remand instructions in the decisions themselves. Case-definition material comes from the VA Public Health and CDC public-domain material and from 38 CFR 3.317; published research criteria are described and cited, never reproduced. This page reports what the Board did. It is educational, not legal or medical advice, and not a prediction of any individual claim outcome. For help with your own claim, talk to a VA-accredited representative.