Normal Pressure Hydrocephalus
On this page
Direct answer
A wet, wobbly and wacky elderly patient — urinary incontinence, a magnetic gait and slowing cognition — with ventricles enlarged out of proportion to cortical atrophy defines possible normal pressure hydrocephalus, one of the genuinely treatable dementias. The gait disorder usually arrives first and responds best: short shuffling steps, broad base, feet glued to the floor, arms still swinging, impaired dual-task walking. Imaging shows an Evans index above 0.3 with tight high convexity sulci and enlarged Sylvian cisterns — the DESH pattern — and CSF pressure and constituents are normal. The diagnostic pivot is the CSF tap test: remove 30-50 mL by lumbar puncture and measure gait before and after; a demonstrable improvement predicts shunt responsiveness, with prolonged external lumbar drainage used when the tap test is equivocal but suspicion stays high. Ventriculoperitoneal shunting with a programmable valve is the treatment, and the realistic hierarchy of benefit is gait first, cognition next, continence last.
What you must remember
- Triad order matters: gait apraxia is typically first, most prominent and most shunt-responsive; cognition slows as subcortical-type deficit (bradyphrenia, executive failure) rather than amnesia; incontinence is late.
- Imaging criteria: Evans index (maximum frontal horn width divided by inner skull width) above 0.3; DESH — dilated Sylvian cisterns with tight high convexity sulci; absent or mild hippocampal atrophy argues away from Alzheimer disease.
- Tap test protocol: standardised gait assessment (10-metre walk time, step count, turn), lumbar puncture removing 30-50 mL, repeat assessment within 24-72 hours; a 10-20% gait improvement is the usual positive threshold per guidelines.
- Secondary NPH: up to a minority follow subarachnoid haemorrhage, meningitis, head injury or posterior fossa surgery — these patients respond better than the idiopathic group.
- Comorbidity caveat: vascular parkinsonism and Alzheimer pathology frequently coexist with NPH and cap the achievable improvement; white matter hyperintensity burden tempers prognosis.
- Shunt choices and complications: programmable valve preferred in the elderly; principal risks are subdural haematoma (over-drainage), infection, blockage and low-pressure headache; valve adjustment is non-invasive.
- Alternative interventions: endoscopic third ventriculostomy has a limited role; serial lumbar punctures are a bridge, not therapy; lumbar-peritoneal shunt is an option in selected patients.
From suspicion to shunt: the pathway walked
A 72-year-old retired teacher has spent 18 months slowing: her family noticed shuffling and freezing at doorways first, then repeated urinary accidents, then forgotten conversations. Examination shows a broad-based magnetic gait with preserved arm swing, no rigidity or tremor, brisk but symmetric reflexes, and frontal-type cognitive slowing with intact delayed recall relative to her processing speed. MRI: Evans index 0.36, tight convexity sulci, dilated Sylvian cisterns, hippocampi preserved. The pathway runs: baseline timed 10-metre walk (28 seconds, 24 steps); lumbar puncture with opening pressure normal at 14 cm of water, 40 mL removed; next morning the same walk takes 19 seconds. Positive tap test, idiopathic NPH, and after counselling about the one-in-ten risk of subdural collection, a programmable ventriculoperitoneal shunt is inserted. At three months she walks unaided and is continent by day; her memory has brightened only at the edges — the expected hierarchy. Had the tap test been negative with strong clinical conviction, external lumbar drainage over 2-3 days would decide; had her MRI shown hippocampal atrophy and a low-pressure-flow void pattern of small vessel disease, the family conversation would have been about realistic expectations instead.
Where candidates slip
Two questions dominate the viva: "Is it NPH or vascular parkinsonism?" — the latter shows lower-body-predominant bradykinesia with a history of strokes and does not have DESH — and "What does the tap test actually predict?" — it predicts gait response to shunting, moderately; a negative tap test does not exclude benefit, which is why extended drainage exists as a second gate. In India, the practical reality is that programmable valves and revision surgery are out-of-pocket costs for most families, so patient selection is not merely academic: tapping the patient most likely to benefit is an economic decision as much as a clinical one, and exam cases increasingly hint at that with "family can afford one intervention" phrasing.
Frequently asked questions
What is the Evans index and its diagnostic threshold?
The maximum width of the frontal horns divided by the inner skull width at the same level; a value above 0.3 indicates ventriculomegaly disproportionate to atrophy.
How is a CSF tap test performed and interpreted?
Remove 30-50 mL of CSF after baseline gait timing and repeat the assessment within 24-72 hours; meaningful gait improvement predicts shunt responsiveness.
Which component of the NPH triad responds best to shunting?
The gait disorder — magnetic apraxia improves first and most reliably, followed by cognition, with continence least dependable.
What is DESH on MRI?
Disproportionately enlarged subarachnoid space hydrocephalus — enlarged Sylvian cisterns with tight high-convexity sulci supporting the hydrocephalus pattern.
What is the most feared shunt complication in the elderly NPH patient?
Subdural haematoma from over-drainage in a brain with stretched bridging veins — risk reduced by programmable valves set to higher opening pressures.