Before you read this: My name is Landon. I'm a 100% P&T vet and a data guy, not a VSO or a lawyer. I work a full time job as a Police Officer in Arizona and develop/manage Claim Raven in any spare minute I get. On my lunch breaks at work I'm answering emails and pushing out updates. A lot of you here were some of the first people to back what I was building, and I have not forgotten it. This one is me trying to give some of that back.

I built Claim Raven. It is a set of AI tools that read your medical records, answer questions about your claim, and help you put together things like your personal statement. The part that makes it different is what it runs on: real Board of Veterans' Appeals decisions, nearly half a million of them in the database, with more than 100,000 broken down case by case.

You do not need to buy anything to use what is in this post. It is all right here. If it helps one person file smarter, that is the point.

Now the data....

I pulled every Board decision in my dataset where a veteran claimed obstructive sleep apnea as secondary to a service-connected psychiatric condition. 1,034 of them. Then I sorted those same decisions by one thing: whether the file held a real, reasoned medical opinion tying the two together.

Strong nexus opinion: granted 95 percent of the time. Weak or missing: 3 percent.

That is not a gap. That is a canyon.

This sits inside a larger set of 101,518 analyzed BVA decisions I work from, and the sleep apnea numbers are not a fluke. Across the whole corpus, a strong nexus opinion grants at 89.5 percent and a weak one grants at 3.3 percent. Sleep apnea secondary to PTSD just shows the cliff in its sharpest form, because the medicine and the law both hang almost entirely on that one document.

Why this matters to you: if you are sitting on a sleep apnea claim tied to your PTSD, depression, or another rated mental health condition, and you are wondering whether the private nexus opinion is worth the money and the hassle, this is the post that answers it. Not with my opinion. With what the Board actually did across a thousand decisions.

TL;DR

• The nexus opinion is the whole ballgame. Sleep apnea secondary to PTSD: strong nexus granted 95%, adequate 87%, weak 3%, missing 0%.
• Who writes that opinion decides the case. Backed by a private IME or treating physician, granted around 86%. Resting on the VA's exam alone, 22%. Nothing, 12%.
• These get remanded more than they get granted. 38% remanded, 27% granted, 32% denied. A remand is not a loss, but it is years.
• The VA's own exams are the problem. In 74% of these cases, the C&P exam was inadequate. That is the single biggest reason the Board keeps shipping them back.
• The leading reason for outright denial was a nexus gap, then no proven primary service connection, then missing diagnosis.
• Sleep apnea is rated 50% the moment it requires a CPAP, which is $1,132.90 a month, tax-free, for a veteran with no dependents.

The Nexus Cliff: 95% vs 3%

Here is the breakdown for sleep apnea claimed secondary to PTSD, sorted by the quality of the medical opinion in the file:

• Strong nexus opinion: 175 cases, granted 94.9%
• Adequate nexus opinion: 111 cases, granted 87.4%
• Weak nexus opinion: 322 cases, granted 2.8%
• No nexus opinion at all: 87 cases, granted 0.0%

Read those last two lines again. When the nexus was weak, 9 grants out of 322. When it was missing, zero out of 87.

The thing almost no one tells you is what was usually NOT in dispute in these cases. Whether the veteran had PTSD: already service-connected, settled. Whether the veteran had sleep apnea: confirmed by sleep study, settled. The fight, over and over, was the bridge between them. One document. A doctor explaining why one caused or worsened the other.

That is the entire case. Everything else is table-setting.

Who Writes the Opinion Decides It

The nexus cliff is about how good the opinion is. This next part is about where it came from, and it is just as stark. Same 1,034 cases, sorted by who authored the nexus opinion:

• Private IME (independent opinion the veteran obtained): 138 cases, granted 86.2%
• Treating physician (your own doctor): 35 cases, granted 85.7%
• VA examiner only (the C&P exam): 302 cases, granted 21.9%
• No medical opinion at all: 559 cases, granted 11.6%

There is a massive self-selection issue here, and I want to be upfront about it. Veterans who go out and pay for a private opinion usually already believe they have a real, developable case, so the private-IME group is not a random slice. The 86 percent is not "buy an opinion and you have an 86 percent shot."

But strip the bias away and the direction still holds, and it lines up exactly with the nexus-quality numbers: a case carried by a reasoned private or treating-physician opinion wins far more often than one resting on the VA's exam, and the VA's exam alone wins less than a quarter of the time. Here is what one of those winning private opinions sounded like, in the Board's words:

>The private opinions submitted by the Veteran found that the Veteran's service-connected disabilities caused him to gain weight, his obstructive sleep apnea was due to his weight gain, and but for his service-connected disabilities the Veteran would not have developed obstructive sleep apnea.

(BVA decision A24041665)

Whether the private opinion is worth it for your situation is a conversation for you, your doctor, and your accredited rep. The data just tells you why that conversation is the one that matters.

Why "Strong" and "Weak" Are Legal Terms, Not Vibes

This is not me grading opinions on feel. The Court of Appeals for Veterans Claims has been specific about what makes a medical opinion worth anything.

Under Stefl v. Nicholson (2007), an opinion has to support its conclusion with reasoning the Board can actually weigh. Under Nieves-Rodriguez v. Peake (2008), the value of an opinion comes from the analysis behind it, not the letterhead it is printed on or who asked for it. A one-line "it is at least as likely as not related to service," with nothing behind it, is close to worthless to the Board. They have a name for it. They call it conclusory, and they throw it out.

So when I say "strong," I mean an opinion that walks through the mechanism. Here is what that reads like, in the Board's own words, from a case they granted:

>the Board is satisfied that the evidence supports a conclusion that the Veteran's sleep apnea, if not caused by it, was at least as likely as not aggravated by his service-connected psychiatric condition, due to weight gain caused by that service-connected disability

(BVA decision A24041700)

Notice what that opinion did. It did not just assert a link. It named the pathway: service-connected psychiatric condition, then weight gain, then aggravation of the sleep apnea. Cause and effect, spelled out.

And here is the Board explaining, in another granted case, exactly why it gave an opinion full weight:

>The November 2024 private opinion contains a clear conclusion with a supporting rationale and relied upon consideration of the Veteran's medical history, the Veteran's lay statements, and medical literature. The Board gives this opinion great probative value.

(BVA decision A25007145)

That sentence is a checklist. Clear conclusion. Supporting rationale. Built on the medical history, the lay statements, and the literature. That is what 95 percent is made of. A weak opinion says "related to service" and stops.

Why These Get Remanded More Than They Get Granted

The outcome split surprised even me. For sleep apnea secondary to PTSD:

• Remanded: 38.4%
• Denied: 32.4%
• Granted: 27.1%

More cases got sent back than got granted. To understand why, look at one number: in 74 percent of these cases, the C&P exam was inadequate.

That is not a typo. Three out of four times, the exam the VA itself ordered was not good enough to decide the case on. And across the full 101,518-decision corpus, when the exam was inadequate, the Board ended up granting or remanding 81.8 percent of the time, mostly remanding. An inadequate exam is a remand machine.

Here is the part that should make you angry, and then make you patient. A remand is not a denial. It means the Board looked at the VA's own work and said "this is not good enough, do it again." It is often the system catching the system's mistake. But it costs you. In my data, remanded cases ran a median of roughly 1,959 days from claim to that decision. Denied cases, about 1,256. The remand is a second chance that you pay for in years.

So if your sleep apnea claim got remanded, you did not lose. You got the most common outcome there is, and the clock is the price.

What Actually Kills These Claims

For the cases that were flat denied, the leading coded reasons were:

1. Nexus gap (the missing or weak medical opinion, same villain as above)
2. No proven primary service connection (you cannot go secondary to a condition that is not service-connected yet)
3. Missing diagnosis

That third one is quieter but it ends claims cold. The Board will not connect a condition you have not actually been diagnosed with, and they are specific about what counts as a diagnosis for sleep apnea. From a denied case:

>The Board cannot accept an obstructive sleep apnea diagnosis based solely on the findings of a STOP-Bang questionnaire, especially when the Veteran expressly denied snoring loudly or having any observed apneas... No polysomnography (sleep study) had been conducted.

(BVA decision A25004918)

A screening questionnaire is not a diagnosis. The Board wants the sleep study. That is not a trick, it is the standard, and knowing it ahead of time is worth more than most of what gets posted about sleep apnea claims.

The second reason matters for sequencing. Secondary service connection runs on 38 CFR 3.310, and it is built on top of something already granted. If the PTSD is not service-connected yet, the sleep apnea claim hanging off it has nothing to stand on.

There Is More Than One Door In

The other thing the data makes obvious: "sleep apnea secondary to PTSD" is not one theory, it is a family of them. The granted cases came in through different doors.

Weight gain, and the rule nobody tells you about. This was the most common winning pathway: the service-connected condition (PTSD, depression, a wrecked back or knee that killed your activity level) drives weight gain, and the weight drives the sleep apnea. Here is the part most veterans have never heard. The VA's own rules treat obesity as a valid link in that chain, even though obesity by itself is not a disability. The Board said it plainly in a granted case:

>obesity is not a disability for purposes of VA benefits; hence, it cannot be the subject of service connection. However, obesity may act as an 'intermediate step' between a service-connected disability and a current disability that may be service-connected on a secondary basis

(BVA decision A24042368)

That is the VA agreeing, in writing, that PTSD to weight gain to sleep apnea is a legitimate route. It is not a stretch. It is in their playbook.

Aggravation, not just causation. This is the other one people miss. Under Allen v. Brown (1995) and 38 CFR 3.310(b), you do not have to prove your service-connected condition CAUSED the sleep apnea from scratch. You can prove it made an existing sleep apnea worse than it would have been on its own. One granted decision:

>Resolving reasonable doubt in his favor, the Veteran's sleep apnea has been aggravated beyond the natural progression as a result of his service-connected posttraumatic stress disorder (PTSD) and septal deviation.

(BVA decision A24039169)

A different physical route entirely. Sleep apnea also won as secondary to service-connected sinusitis, on a pure airway-mechanics theory:

>Medical evidence shows that the Veteran's obstructive sleep apnea is causally related to his service-connected sinusitis, due to pathologic changes in airflow velocity and resistance, which result from nasal obstruction.

(BVA decision A25000260)

Direct, in-service onset. Snoring, witnessed apneas, waking up unrefreshed during service, documented by the people who were there. Sleep studies were rarely done on active duty, so lay statements from the folks who shared a barracks or a bed carry real weight here.

The VA fights back on these, especially the weight-gain route. In one denied case, the examiner argued weight gain is "75 percent diet and 25 percent exercise" to break the chain. Knowing they will attack the pathway is half of being ready for it.

The Sentence the Board Keeps Granting On

If you read enough of these grants, the winning opinions start to sound the same. They follow a three-step chain, and they say it out loud:

1. The service-connected condition caused the weight gain.
2. The weight gain, the obesity, was a substantial factor in causing the sleep apnea.
3. But for that weight gain, the sleep apnea would not have developed.

Here is that exact structure, from three different granted decisions, three different veterans, three different service-connected conditions:

>had the Veteran not suffered from his service-connected musculoskeletal disabilities, he would likely not have suffered from substantial weight gain, and had he not suffered from substantial weight gain, he likely would not have developed OSA.

(BVA decision A24041681)

>the Veteran's service-connected knee disability caused the Veteran to become obese, the obesity was a substantial factor in causing OSA, and OSA would not have occurred but for the obesity caused by service-connected knee disability.

(BVA decision A25019658)

>The Veteran's obstructive sleep apnea is but-for due to his service-connected disabilities, with obesity as an intermediate step.

(BVA decision A25001144)

Same skeleton every time. Service-connected condition, then weight gain, then "but for," then sleep apnea. That is not a coincidence. That is the structure the law rewards. An opinion that does not connect all three links is the one that lands in the 3 percent.

The Chain Does Not Stop at Sleep Apnea

Here is what makes this more than a single claim. Once sleep apnea is service-connected, it becomes the primary condition for the NEXT one.

In my data, hypertension claimed as secondary to sleep apnea grants at 89.2 percent when the nexus opinion is strong. So the real chain looks like this:

PTSD, then weight gain, then sleep apnea, then hypertension.

Each link, properly documented, is its own rated condition. And the conditions that cluster around PTSD in my extracted set are exactly the ones you would expect: sleep apnea, hypertension, depression, and alcohol use disorder all show up together. The point is not to file everything. The point is that these conditions travel as a chain, and the Board grants them when the medical story connecting each link is told properly.

The Rating Reality

This is not a throwaway claim. Under Diagnostic Code 6847, obstructive sleep apnea is rated at:

• 0 percent if it is asymptomatic but documented
• 30 percent for persistent daytime sleepiness
• 50 percent the moment it requires a breathing assistance device, meaning a CPAP
• 100 percent for chronic respiratory failure or cor pulmonale, or if it requires a tracheostomy

The 50 percent tier is where most veterans land, because most diagnosed sleep apnea comes with a prescribed CPAP. As of the rates effective December 1, 2025, that 50 percent pays a veteran with no dependents $1,132.90 a month. That is $13,594.80 a year, and it does not show up on your tax return. With a spouse it is $1,241.90 a month. Run your own number for your dependent situation, but understand the stakes before you decide the nexus opinion is too expensive.

Everything above is a pattern across a thousand strangers' files. Here is what it means if the case is yours.

You are not fighting to prove you have PTSD. You probably already won that. You are not fighting to prove you have sleep apnea, if you have the sleep study. You are fighting over one document: the opinion that connects them, written by someone who explains the why and not just the conclusion. That is the hill. Everything the data shows points back to it.

Bottom Line

For a secondary sleep apnea claim, the diagnosis and the service-connected condition are the easy part. The reasoned nexus opinion is the case. Strong opinion from a doctor who explains the mechanism, the Board grants almost every time. Weak, conclusory, or missing, it almost never does. If you take one thing into the room with your accredited rep, take that.

Methodology and Limitations

Data source: Claim Raven's structured analysis of Board of Veterans' Appeals decisions. The headline numbers come from 1,034 decisions involving sleep apnea claimed secondary to a psychiatric condition, within a larger set of 101,518 analyzed BVA decisions. Case numbers cited are real and verifiable.

Three honest limits:

1. Selection bias, and it is big. These are cases that were denied or partially denied and then appealed all the way to the Board. They are not a random sample of all sleep apnea claims. Claims granted at the regional office never show up here, so these are Board grant rates, not your odds at the start. The same caution applies double to the by-provider numbers: veterans who buy a private opinion are not a random group.
2. "Nexus quality" is my classification, applied after the fact. It is consistent across the dataset, but it is an analytical judgment about the opinion in the file, not a label the Board assigned.
3. Patterns are not predictions. A correlation between strong opinions and grants is not a promise about your file. Your facts are your facts.

Disclaimer

I am not accredited by VA, not a lawyer, not a VSO. This is data analysis, not claim advice. Per 38 CFR 14.629, only VA-accredited representatives (VSOs, claims agents, or attorneys) may give personalized guidance on a specific claim. The patterns here are drawn from Claim Raven's analysis of 1,034 BVA decisions involving sleep apnea claimed secondary to PTSD. For your situation, work with an accredited VSO, claims agent, or attorney. You can find accredited representatives at VA.gov.

Where to go next

• https://claimraven.com

A note on where this came from, since I would want to know. I run the numbers like this because the information that decides claims is buried in more than 100,000 Board decisions nobody reads. I built Claim Raven to dig it out. The same analysis engine behind this post runs on any record you put through it. There is also a version built for the VSOs, claims agents, and attorneys who do this for a living, with the full theory bank and the Board pattern data for each condition side by side. If you carry a caseload, that side is built for you. If you are a veteran fighting your own claim, the rest of it is built for you.

-Landon
Founder, Claim Raven | U.S. Army Veteran