How One Condition Leads to Another

A mechanism is the physical story of how the first condition produces the second. Not that they appear together. How.

Diabetes is the clearest example. Blood sugar that stays high for years damages the smallest blood vessels, and the smallest blood vessels are the ones feeding the longest nerves and the back of the eye. That is why the damage shows up in the feet before the hands. The pattern follows the plumbing.

That pathway is textbook physiology rather than the finding of any one paper, and the two are worth keeping apart. What the papers add is scale. A pooled analysis of 29 studies covering 50,112 people found nerve damage in about 30 percent of people with diabetes, higher in type 2 than in type 1 (PMID 31917119). That study counted people. It did not trace the pathway, and this page does not pretend it did.

Sleep apnea has a mechanism too, and it is mechanical. Breathing stops, oxygen drops, the body wakes itself enough to restart it, and this happens dozens of times an hour every night. Blood pressure and heart rhythm are dragged along with it.

Now compare that to a list. A list says PTSD and high blood pressure. It does not say what connects them, so it cannot tell you whether anything does. Most of what circulates about secondary conditions is lists, which is why so much of it is wrong in ways nobody can check.

It means the two things turned up together more often than chance would explain. That is all it means. It is not a small finding and it is not the same finding as one causing the other.

Researchers are careful about this in a way that popular writing is not. A paper that found people with sleep apnea developed diabetes more often over ten years says exactly that, and puts a number on it: a relative risk of 1.40, rising step by step with how severe the apnea was (PMID 31872550). It does not say sleep apnea causes diabetes. The distance between those two sentences is where most of the bad information in this category lives.

There is a phrase worth learning to spot: risk factor. Researchers use it when a study followed people forward in time, the first thing came before the second, and more of the first went with more of the second. The Wisconsin Sleep Cohort study measured people overnight and then measured their blood pressure four years later, and even then the authors wrote that the findings suggest sleep-disordered breathing is likely to be a risk factor (PMID 10805822). Suggest. Likely. Those words are load bearing and they get deleted every time this study is quoted in an advertisement.

Three questions sort almost everything.

First, what kind of study is it? A meta-analysis pools many separate studies into one result and is the strongest thing on most of these pages. A single large cohort followed forward is next. A cross-sectional study photographs one moment, which means it can count how many people have both conditions but cannot say which came first.

Second, how many people? A pooled analysis of 338,912 people carries more than a single study of 117. Both can be worth reading. They are not worth the same.

Third, and this is the one almost nobody asks: do the studies agree? A review of sleep apnea and depression found nothing convincing when it pooled the cross-sectional studies, and a doubled risk when it pooled the ones that followed people forward (PMID 33158487). Both results are in the same paper. A list that prints one line for this pair has thrown away the most interesting thing about it.

When a finding is repeated across many studies, followed forward in time, and has a mechanism that explains the direction, it is about as established as this field gets. When it rests on one small study, or on studies that disagree, that is worth knowing before repeating it.

Only some designs can, and knowing which is most of the skill.

Counting people at one moment cannot. If a survey finds that veterans with tinnitus are more likely to be depressed, that is real and it is silent about direction. Tinnitus is exhausting and could produce depression. Depression makes intrusive noise harder to tolerate and could make tinnitus register as worse. Both could come from a third thing.

Following people forward does better. A Danish study tracked nearly 2.5 million people from birth records and found that a first head injury roughly doubled the rate of epilepsy afterwards, and a second one raised it to about four and a half times (PMID 33439977). The injuries came first. That is built into the design rather than assumed.

Some designs go further still. A study of Vietnam era twins compared pairs where one twin had PTSD and the other did not, which holds genetics and childhood steady between the two. Heart disease turned up in 22.6 percent of the twins with PTSD and 8.9 percent of those without (PMID 23810885).

Because most of them are copies of each other, and none of the copies went back to the literature.

The clearest case is the one about limping. The story is that a bad knee changes how you walk, which throws weight onto the other side, which wears that side out. It sounds obviously true. The best study found for it followed 117 people who had one knee replaced, and about 37 percent had the other knee replaced within ten years, so the pattern is real. But what predicted it was how much arthritis the other knee already showed on its own X-ray before the first surgery. Age, sex, weight and which side was operated on predicted nothing (PMID 12913941). The pattern is real and the explanation attached to it is not the one the data supports.

Plantar fasciitis is worse. Forums routinely treat it as the start of a chain running up through the knees, hips and back. Two PubMed searches for that relationship, run on 21 August 2026 and printed in full on the page about folklore lists, returned no results at all. Meanwhile the review that does exist on plantar fasciitis, covering 51 studies, states that there is a lack of evidence for the dogma of clinical and mechanical measures of foot and ankle function (PMID 26644427). The mechanical story the whole chain rests on is the part the evidence does not support.

And sometimes the correction runs the other way. Tinnitus and hearing loss are usually treated as one thing. A study of 596,905 people in the Million Veteran Program compared their genetics and concluded that tinnitus is a distinct disorder separate from hearing difficulties (PMID 38242899). They travel together because inner ear damage often produces both, not because one produces the other.

No, and this is the hardest thing on the page to hold onto.

Every number in this field describes a group. When a meta-analysis of 32 studies reports that head injury goes with about a 66 percent higher rate of dementia, that is a statement about populations (PMID 36305374). It is not a statement about anyone in particular, and it cannot be turned into one by reading it harder.

The same paper makes the point better than any warning could. Being a veteran did not itself carry a higher rate. What the authors found was that the share of dementia in the United States attributable to head injury is about twice as high among veterans, purely because far more veterans have had one. Same risk per person, different picture across a population. Those are different facts and they get collapsed constantly.

The other direction matters too. Some studies find nothing where a list confidently says there is something. In nearly 10,000 Vietnam era veterans, PTSD went with high blood pressure in the men and did not in the women (PMID 40353294). Anyone summarizing that as one flat fact about PTSD is reporting something the study did not find.

What a mechanism can do is tell you what to watch for and what to mention to a clinician. What it cannot do is tell you what is happening inside one body. That takes an examination.

It sits downstream of all of it, and it is a separate question.

38 CFR 3.310 is the rule that treats a disability caused by, or made worse by, an already service-connected condition as connected itself. The glossary entry on secondary conditions covers what the regulation says and is the right place to read it.

Two things worth keeping straight. The medicine and the regulation are answering different questions: the literature asks whether a pathway exists in general, and the VA asks what a particular record shows about a particular person. A well-established mechanism is not a finding about anyone's claim, and a claim can succeed on a record without any of this.

The second is that the regulation covers being made worse, not only being caused. That prong gets dropped from most guides, and a great many real situations live there. It has its own page.

Less than people expect, and something more useful than they expect.

A clinician cannot act on a printed list. They can act on a symptom.

So the practical version of everything above is short. If you recognize something in yourself, describe it to your provider as a symptom. Use your own words. Leave the diagnosis off. Numbness that started in the toes. Waking up with headaches. Choking awake at night.

That gets examined, and if it is real it gets written down. What is written down is what a record contains, and a record is what everything downstream is built from. A free VSO can help with the paperwork side at no cost, and they are accredited to do it.

There is no trick here and no shortcut worth paying for. The mechanism tells you what is worth mentioning. The examination decides what is true.