Intracranial Hypotension and Spinal CSF Leaks

Intracranial hypotension is a condition in which the pressure of cerebrospinal fluid (CSF) inside the skull drops too low, almost always because CSF is leaking somewhere along the spine. The hallmark symptom is a headache that worsens when you stand up and improves when you lie down, though the full picture can be far stranger and harder to pin down than that simple description suggests. Once thought to be exceptionally rare, better imaging and growing clinical awareness have pushed estimated incidence figures up to roughly 5 per 100,000 people per year, and many experts suspect the true number is higher still.

Why CSF Pressure Drops

Your brain floats in cerebrospinal fluid, which acts as a cushion and keeps intracranial pressure stable. The Monro-Kellie principle describes the skull as a fixed-volume container: the combined volume of brain tissue, blood, and CSF must stay roughly constant. When CSF leaks out through a spinal defect, the lost volume has to be compensated somehow. Blood vessels inside the skull engorge to fill the gap, the meninges (the membranes wrapping the brain) thicken, and the brain itself can sag downward under gravity. Many of the symptoms and imaging findings in intracranial hypotension follow directly from this compensatory response.

CSF can escape through the spine in a few distinct ways. A large classification study of over 500 patients identified three main leak types. About a quarter of patients had a dural tear, a straightforward hole in the membrane surrounding the spinal cord, with the vast majority of these tears located on the front (ventral) side of the spine. The most common leak type, found in over 40% of patients, involved meningeal diverticula, which are small outpouchings of the spinal membrane that can rupture and allow CSF to seep out. A third, less common type is the CSF-venous fistula, in which cerebrospinal fluid drains directly into a spinal vein. Around 29% of patients in that study had an indeterminate leak source even after testing.

Who Develops It

Spontaneous intracranial hypotension (SIH) affects women roughly twice as often as men, and symptoms most commonly begin around age 40, though cases span from early childhood to the late eighties. A systematic review and meta-analysis found a mean age of about 42.5 years at presentation, with women making up about 63% of cases.

Early estimates from the 1990s put the prevalence at about 1 per 50,000. More recent data from a community-based study in Beverly Hills found an average annual incidence of about 3.7 per 100,000, with rates slightly higher in women than men.p>⁴ An earlier emergency department study estimated the annual incidence at roughly 5 per 100,000, about half as common as spontaneous subarachnoid hemorrhage.⁵ The gap between these estimates likely reflects differences in study design, diagnostic criteria, and the reality that many cases go unrecognized for months or years.

The Headache and Beyond

The classic symptom is an orthostatic headache, one that gets worse when upright and better when lying flat. In a study of 137 SIH patients, about 76% had this positional headache pattern.

⁶ The typical version starts mild upon waking, worsens within a few hours of standing, and fades within about 90 minutes of lying back down. But this “textbook” pattern does not appear in every case. A quarter of patients in that study lacked clearly orthostatic headaches, and some present with headaches that are continuous, worse later in the day, or triggered by exertion rather than posture alone. These atypical presentations are a major reason SIH gets missed.

Beyond headache, the neurological effects can be surprisingly varied. Neck stiffness, nausea, changes in hearing or ringing in the ears, and dizziness are common. In more severe cases, when the brain sags far enough to compress structures at the base of the skull, patients may develop double vision, facial numbness, or difficulty with balance. One particularly dramatic presentation is “frontotemporal brain sagging syndrome,” in which downward displacement of the brain mimics the behavioral and cognitive changes seen in frontotemporal dementia, including personality shifts, impaired judgment, and altered mental status.

⁷

How It Shows Up on Imaging

Brain MRI with contrast is the standard first-line imaging study. When CSF volume drops, a predictable set of changes appears: diffuse, smooth thickening and enhancement of the dura (the tough outer membrane); downward displacement of the brain and cerebellar tonsils (brain sag); engorgement of the venous sinuses; enlargement of the pituitary gland; and sometimes subdural fluid collections or small bleeds.

⁸

Not all of these signs are permanent. One study found that patients who had been symptomatic longer were actually less likely to show dural enhancement on MRI, with average symptom duration of about 45 weeks in those without enhancement compared to about 15 weeks in those with it.

⁹ This means a normal-looking brain MRI does not rule out SIH, especially in someone who has had symptoms for months. The imaging signs can fade even while the leak persists.

Another diagnostic pitfall: CSF pressure itself can be normal. A study measuring opening pressure in SIH patients found that about 55% had clearly low pressure, but about 44% had pressure in the normal range, and a couple even had elevated pressure.

¹⁰ This is a crucial point: “intracranial hypotension” is named for low pressure, but the actual problem is CSF volume loss, and the two do not always line up. A normal lumbar puncture reading should not be used to dismiss the diagnosis.

Finding the Leak

Once brain imaging points toward SIH, the next challenge is finding where the CSF is escaping along the spine. This is often the hardest part of the diagnostic process. Conventional MRI of the spine can sometimes show extradural fluid collections that suggest a leak location, but frequently the results are inconclusive.

More advanced techniques involve injecting contrast dye into the spinal fluid and watching where it goes. Digital subtraction myelography (DSM) is a fluoroscopy-based method that can pinpoint leak sites in real time and is especially useful for detecting CSF-venous fistulas, the leak type in which CSF drains directly into a vein.

¹¹ Dynamic CT myelography uses CT scanning after intrathecal contrast injection and is particularly good at identifying dural tears and ruptured diverticula, though radiation exposure can be a concern with repeated scans.

¹²

Patient positioning during these studies turns out to matter enormously. A study comparing digital subtraction myelography done with the patient lying on their side versus face-down found CSF-venous fistulas in 74% of patients in the lateral position versus only 15% in the prone position.

¹³ A separate head-to-head study comparing lateral decubitus CT myelography with lateral decubitus digital subtraction myelography found that the CT method identified substantially more fistulas, suggesting it may be the more sensitive approach for this specific leak type.

¹⁴

Treatment, From Conservative to Invasive

Initial treatment is usually conservative: bed rest, plenty of fluids, and caffeine, which may help stimulate CSF production. For many patients with mild symptoms, these measures are enough. When they are not, the next step is an epidural blood patch (EBP). In this procedure, a small volume of the patient’s own blood is injected into the epidural space in the spine, where it clots and ideally seals the leak.

¹⁵

Blood patches can be done “blind” (injected at a standard lumbar level without knowing the exact leak site) or targeted to the area where imaging has localized a leak. One study found site-directed patches were significantly more effective, but a recent systematic review and meta-analysis concluded that when both approaches were compared across the available literature, there was no statistically significant difference in outcomes.

¹⁶¹⁷ The meta-analysis suggested that a non-targeted blood patch may be a reasonable first step regardless of whether the leak has been precisely located, which matters practically because the imaging workup to localize a leak can be time-consuming, expensive, and not available at every center.

When blood patches fail, options get more specialized. Epidural injection of fibrin glue, a surgical adhesive, has been used successfully in cases where blood alone does not hold.

¹⁸ For dural tears that resist patching, open surgical repair may be needed. And for CSF-venous fistulas, an increasingly popular treatment is transvenous embolization, in which a catheter is threaded into the draining vein and used to close the fistula. In one series of 22 patients, embolization was successful in every case, with 90% experiencing improvement and about two-thirds reporting complete recovery. Headache impact scores dropped dramatically and quality-of-life ratings climbed from very low levels to near-normal.

¹⁹ A larger series of 100 patients treated with transvenous embolization reported a 100% technical success rate, with 95% reporting significant improvement or complete resolution of symptoms.²⁰

Subdural Hematomas and the “Leak First” Approach

One of the more serious complications of prolonged SIH is the development of subdural hematomas, collections of blood between the brain and its outer membrane. When the brain sags away from the skull, the bridging veins that span that gap get stretched and can tear. In one retrospective series, about 16% of SIH patients developed subdural hematomas.

²¹ The risk increases with longer symptom duration and with certain MRI findings like dural enhancement and engorgement of the venous sinuses.²²

Managing these hematomas is tricky because of a counterintuitive clinical trap. If the underlying CSF leak is not recognized and a surgeon simply drains the subdural collection, the low CSF pressure that caused the problem in the first place remains. The hematoma tends to recur, and the patient may worsen. The recommended approach for patients without acute neurological deterioration is to address the spinal leak first, often with a blood patch, and let the subdural collection resolve on its own. For small, asymptomatic hematomas under about 10 mm without significant compression of the brain, closing the CSF leak alone has been shown to resolve the hematoma completely without recurrence.

²³ When someone is deteriorating rapidly with signs of brain compression, surgeons may need to evacuate the hematoma first, but performing an early blood patch soon after surgery helps prevent recurrence.²⁴

Rebound High Pressure After Treatment

A frustrating twist for some patients is that after a successful blood patch, headaches come back but with different features. Instead of worsening when upright, the headache may now worsen when lying down, sometimes accompanied by nausea, blurred vision, or double vision. This is rebound intracranial hypertension, essentially an overshoot in the opposite direction. It has been reported in up to about 27% of patients after EBP.

²⁵

Recognizing this pattern matters because the treatment is the opposite of what you would do for persistent low pressure. Rebound high pressure may require medications that reduce CSF production (such as acetazolamide), while unresolved SIH may call for another blood patch. Researchers have suggested that SIH and high-pressure disorders like idiopathic intracranial hypertension may sit on a shared spectrum, with changes in venous pressure and the elasticity of the craniospinal system connecting the two ends.

²⁶

The Connective Tissue Connection

Because CSF leaks in SIH happen spontaneously, without trauma or surgery, researchers have long suspected that underlying weakness in connective tissue might be involved. A prospective study of 18 SIH patients found that 38% had signs of a systemic connective tissue disorder, including features of Marfan syndrome, Ehlers-Danlos syndrome, and unusual joint hypermobility with weakened fascia.

²⁷ However, a controlled study that compared SIH patients to matched controls found that most formal connective tissue criteria (joint hypermobility scores, skin signs of Ehlers-Danlos syndrome, skeletal features of Marfan syndrome) did not differ significantly between the two groups. One notable exception was dolichostenomelia, disproportionately long limbs, which was present in about 35% of SIH patients versus 9% of controls.²⁸

So the relationship is real but not straightforward. Having a recognized connective tissue disorder is neither necessary nor sufficient to develop SIH. There may be subtler variations in tissue strength, not captured by standard diagnostic checklists, that predispose certain people to spinal membrane weakness.

Misdiagnosis as Chiari Malformation

One of the most consequential diagnostic errors in SIH is mistaking it for Chiari malformation type 1, a condition in which the cerebellar tonsils protrude below the skull base. Both conditions produce tonsillar descent on brain MRI, and both can cause headaches, though the headache pattern typically differs. In a case series from Stanford, five patients with spinal CSF leaks were initially misdiagnosed as having Chiari malformation and underwent suboccipital decompression surgery, an operation that removes bone at the base of the skull. None of them improved from the surgery. All five were later correctly diagnosed and treated with epidural blood patches, which provided sustained relief.

²⁹

The lesson is that tonsillar descent on an MRI is not a diagnosis by itself. When it appears alongside a headache that worsens with upright posture, or alongside other signs of low CSF volume like dural enhancement or venous engorgement, a spinal leak should be investigated before anyone considers posterior fossa surgery.

Children and Adolescents

SIH in children is rare but not unheard of. A study of 24 pediatric patients (ages 2 to 19, average 14) found that the condition follows a broadly similar pattern to adult cases, with orthostatic headaches in nearly all of them and typical MRI findings in about 79%. But two features stood out. First, more than half of these children had a diagnosed connective tissue disorder, a much higher proportion than in adult series. Second, a larger fraction required escalation beyond blood patches: about a third needed fibrin glue injections and nearly half eventually required surgical repair of the leak.

³⁰ Despite the more aggressive treatment course, outcomes were good in 92% of cases. Five of the 24 children developed rebound high pressure after treatment and needed acetazolamide.

The Burden of Living With It

The time from symptom onset to correct diagnosis can be shockingly long. The meta-analysis cited earlier found a mean of about 32 days, but the range stretched from a single day to nearly 20 years.

³¹ During that time, patients often cycle through misdiagnoses of migraine, tension headache, or psychological stress.

The disability burden while symptomatic is severe. One study calculated a disability weight of 0.432 for active SIH, meaning patients lose roughly 43% of their health-related quality of life. That figure is comparable to or worse than many other chronic conditions the public takes more seriously.

³² A survey-based study found that the overwhelming majority of SIH patients reported depression, anxiety, and stress, with only about 15% able to work full-time during active disease. About 30% were unable to work at all, and nearly 6% had retired early because of the condition.

³³

Treatment does improve things substantially. A meta-analysis of patient-reported outcomes found that perceived health scores on a standard quality-of-life measure roughly doubled after treatment, and headache impact scores dropped by about a third.

³⁴ But even after treatment, many patients reported lingering physical, mental, and social limitations. Recovery from SIH, especially in long-standing cases, is not always the clean resolution patients hope for.

CSF Biomarkers on the Horizon

Diagnosing SIH currently relies on clinical suspicion, MRI, and invasive myelographic studies. Researchers have been looking for CSF biomarkers that might make the process simpler. One promising approach examined proteins in cerebrospinal fluid and found that SIH patients had elevated levels of certain brain-derived molecules. A combination of two of these proteins, lipocalin-type prostaglandin D synthase and brain-type transferrin, could distinguish SIH from non-SIH cases with about 95% sensitivity and 73% specificity.

³⁵ This is still a research tool rather than a routine clinical test, but it points toward a future where a CSF sample alone might flag the diagnosis before someone has to undergo contrast-based imaging of the entire spine.

1
PubMed. The Monro-Kellie hypothesis: applications in CSF volume depletion ↩︎
2
PubMed. A classification system of spontaneous spinal CSF leaks ↩︎
3
JAMA Neurology. Clinical Presentation, Investigation Findings, and Treatment Outcomes of Spontaneous Intracranial Hypotension Syndrome: A Systematic Review and Meta-analysis ↩︎
4
PubMed. Incidence of spontaneous intracranial hypotension in a community: Beverly Hills, California, 2006-2020 ↩︎
5
JAMA. Spontaneous Spinal Cerebrospinal Fluid Leaks and Intracranial Hypotension ↩︎
6
PubMed Central. Defining the typical characteristics of orthostatic headache in patients with spontaneous intracranial hypotension ↩︎
7
Radiology Case Reports. A case of spontaneous intracranial hypotension in a 45-year-old male with headache, behavior changes and altered mental status ↩︎
8
American Journal of Neuroradiology. Spontaneous Intracranial Hypotension: Atypical Radiologic Appearances, Imaging Mimickers, and Clinical Look-Alikes ↩︎
9
PubMed. Time-Dependent Changes in Dural Enhancement Associated With Spontaneous Intracranial Hypotension ↩︎
10
PubMed Central. Factors affecting cerebrospinal fluid opening pressure in patients with spontaneous intracranial hypotension ↩︎
11
PubMed Central. Lateral Decubitus Digital Subtraction Myelography: Tips, Tricks, and Pitfalls ↩︎
12
American Journal of Neuroradiology. Modified Dynamic CT Myelography for Type 1 and 2 CSF Leaks: A Procedural Approach ↩︎
13
PubMed. Lateral decubitus digital subtraction myelography to identify spinal CSF-venous fistulas in spontaneous intracranial hypotension ↩︎
14
PubMed Central. Direct comparison of digital subtraction myelography versus CT myelography in lateral decubitus position: evaluation of diagnostic yield for cerebrospinal fluid-venous fistulas ↩︎
15
PubMed Central. Recent update on epidural blood patch ↩︎
16
Regional Anesthesia & Pain Medicine. Procedural predictors of epidural blood patch efficacy in spontaneous intracranial hypotension ↩︎
17
PubMed Central. Targeted Versus NonTargeted Epidural Blood Patch for Spontaneous Intracranial Hypotension: A Systematic Review and Meta-Analysis ↩︎
18
Headache: The Journal of Head and Face Pain. Spontaneous Intracranial Hypotension Successfully Treated by Epidural Patching With Fibrin Glue ↩︎
19
PubMed Central. Safety and efficacy of transvenous embolization of cerebrospinal fluid-venous fistula in patients with spontaneous intracranial hypotension ↩︎
20
Journal of NeuroInterventional Surgery. Clinical and imaging outcomes of 100 patients with cerebrospinal fluid-venous fistulas treated by transvenous embolization ↩︎
21
PubMed. Treatment and outcome of subdural hematoma in patients with spontaneous intracranial hypotension: a report of 35 cases ↩︎
22
PubMed Central. Risk Factors for Subdural Haematoma in Patients with Spontaneous Intracranial Hypotension ↩︎
23
Brain and Spine. Management of chronic subdural hematoma in spontaneous intracranial hypotension ↩︎
24
PubMed. Treatment and outcome of subdural hematoma in patients with spontaneous intracranial hypotension: a report of 35 cases ↩︎
25
PubMed Central. Early and Delayed Rebound Intracranial Hypertension following Epidural Blood Patch in a Case of Spontaneous Intracranial Hypotension ↩︎
26
PubMed Central. Rebound Intracranial Hypertension ↩︎
27
PubMed. Connective tissue disorders with spontaneous spinal cerebrospinal fluid leaks and intracranial hypotension: a prospective study ↩︎
28
PubMed. Connective tissue disorders in patients with spontaneous intracranial hypotension ↩︎
29
Heliyon. Cerebellar tonsillar descent: A diagnostic dilemma between Chiari malformation type 1 and spinal cerebrospinal fluid leak ↩︎
30
PubMed. Spontaneous intracranial hypotension in childhood and adolescence ↩︎
31
JAMA Neurology. Clinical Presentation, Investigation Findings, and Treatment Outcomes of Spontaneous Intracranial Hypotension Syndrome: A Systematic Review and Meta-analysis ↩︎
32
PubMed. Toward quantifying disability in spontaneous intracranial hypotension: A patient-derived disability weight ↩︎
33
PubMed Central. Assessing the effects of spontaneous intracranial hypotension on quality of life, work ability and disability ↩︎
34
PubMed. Quality of life in patients with spontaneous intracranial hypotension: A systematic review and meta-analysis ↩︎
35
PubMed. Spontaneous intracranial hypotension is diagnosed by a combination of lipocalin-type prostaglandin D synthase and brain-type transferrin in cerebrospinal fluid ↩︎