Dizziness due to TIA and Stroke 

Timothy C. Hain, MD .• Page last modified: August 22, 2021


Transient ischemic attacks or TIA's are brief episodes of neurological disturbance caused by reduced blood supply to an area of the brain. Strokes are longer lasting neurological disturbance, associated with permanent damage to the brain. This document discusses TIA's and strokes associated with dizziness, also known as Vertebrobasilar TIA's and Strokes. These strokes are from blockages of one or more of the arteries shown on the the picture to the right (Posterior Inferior Cerebellar Artery or PICA, vertebral arteries, anterior inferior cerebellar arteries or AICA, basilar artery, superior cerebellar artery or SCA).

cerebral blood vessals VBI VBI
Normal "posterior circulation" MRA scan of individual with narrowing of right vertebral artery, and dizziness/imbalance. No PICA is seen on the R side. MRI scan of individual with narrowing of both vertebral arteries, just before joining basilar, with spells involving weakness of the leg and dizziness.



A TIA or stroke usually begins abruptly. Reduced blood supply to the back part of the brain can cause dizziness. TIAs are temporary, and strokes have longer lasting symptoms (> 24 hours). In order of decreasing frequency, besides dizziness, other symptoms of vertebrobasilar TIA include visual disturbance, drop attacks, unsteadiness or incoordination, weakness, confusion, headache, hearing loss, numbness, speech disturbance, abnormal noise in the ears, and numbness around the mouth.

Frequency of strokes causing dizziness

Most dizziness is not caused by strokes. Rather most of the time it is probably due to medical conditions such as low blood pressure or low blood sugar. Less commonly, it is from inner ear conditions such as BPPV, vestibular neuritis, or Meniere's disease.

So how much dizziness is caused by vascular disease?. To answer this rigorously, one would need to to develop a sample of dizzy patients, and one would also have to find a way to know for sure that they were having a stroke. We would also have to have a rigorous way of deciding if someone is really "dizzy" (Newman-Toker, 2007). None of these tasks is very feasible, so what we have is pretty rough.

Most of the data about this comes from work by Newman-Toker (e.g. 2007, 2008).

Dizziness accounts for 3.5% to 11% of ED visits (see this page), and stroke makes up 4% of these, so we have then 4% of 11% (i.e. 0.04 * .11, or .0044 =0.44 %) of ED visits being dizziness caused by cererebrovascular sources. Inner ear problems account for about 1/3 of ED dizziness (Newman-Toker et al, 2008), implying that about 1-3% of ED visits are from inner ear dizziness. So out of the dizzy group in the ED, taking the top estimates for both, about 0.44 % is from stroke and 3.6% from inner ear. Roughly an 1:8 ratio.

Saber-Tehrani et al (2014) wrote that "10% to 20% of emergency department dizziness" is made up of acute vestibular syndrome patients (AVS). This would then imply that of the 11% of ED visits for dizziness, about 1-2% of them are for AVS. So it would seem that the estimate of AVS is equal to the estimate for "inner ear dizziness". Probably to high by a factor of 2 or 3, but lets go on.

Saber-Tehrani et al (2014) further wrote that "perhaps 25%" of the AVS syndrome cases are due to stroke. This then implies that 25% of the 1-2% are from stroke, or in other words, then 0.25-0.50 % of ED dizzy cases should have a stroke (which is narrower than cerebrovascular cause). So again, a high estimate -- likely very high.

So it would appear that the bottom line from these authors is that roughly 0.25 to 0.50 % of ED cases should have dizziness due to stroke, or at least a cerebrovascular cause (which could include a TIA). OK, lets just use 0.5% as it is easy to remember, but keep in mind it is probably too high.

So lets step back then and ask -- what percent of all dizziness in all people (not necessarily in the ED) is due to stroke -- well probably considerably less than 0.5 %, as one has to be very dizzy to go to the ED. Still, if we guess that 1/10 people with dizziness go to the ED, then we would have a very rough estimate of 0.05% of all dizziness being associated with a stroke. So about 5/1000.



To some extent, one can predict risk of stroke. Risk factors that are well known include:

**Numbers in () are derived from Whisnant et al, 1996

Risk from elevated blood pressure is steep and clear. For example, in the UK TIA trial, risk for recurrent stroke increased by 28% for every incremental increase of 10 mm Hg in systolic blood pressure between 130 and 160 (Farrell et al, 1991).

Although LDL-cholesterol increases risk, HDL cholesterol seems to reduce risk of stroke (Sacco et al, 2001). If your HDL cholesterol is > 35, then subtract one risk factor (a negative risk factor). Mitral valve prolapse is not a significant risk factor, overall (0.8 risk). TIA is a very strong risk factor for stroke (5.6 x risk).  In general, relative risk for most of the above factors decreases with age (Whisnant et al, 1996), lending support for a unaggressive approach to risk factors in individuals of advanced age.

Risk from cholesterol can be further stratified into three groups, based on LDL (total cholesterol - HDL)-(triglycerides).

LDL Risk Factors Risk Level for vascular disease
< 130 None Low
130-159 Less than 2 Moderate
>130 more than 2 High

Controllable risk factors include being overweight, having high (> 140/90) or low blood pressure, heart disease, diabetes and smoking. Atrial fibrillation is a particularly important risk factor -- stroke occurs in 4.5% of untreated patients with atrial fibrillation per year. While uncommon, chiropractic neck manipulations can cause compression or tears of the vertebral arteries (Vibert et al, 1993; Smith et al, 2003), and for this reason, maneuvers involving neck "cracking" should be specifically avoided in individuals with vertigo. Whiplash injuries can also damage the vertebral arteries as the arteries traverse the vertebrae of the neck.

It is presently suggested that LDL be less than 100.

Dissection of vertebral artery (cutoff of left sided vessel). Image courtesy of Ruth Ramsey, M.D. On the right hand side is another image of a vertebral dissection in a different patient. The vertebral artery on the left side is smaller and irregularly filled.


The diagnosis of TIA or stroke is usually made by a neurologist. Diagnosis is based upon having a compatible group of symptoms, exclusion of other reasonable causes such as disease of the inner ear, and identification of a cause of reduced blood flow.

Even when the examining doctor is very experienced, it is not always possible to be sure that a patient with dizziness does NOT have a stroke (see case example). Except in a few situations (i.e. BPPV), the clinical signs of vertigo are specific enough to exclude stroke. Because scanning every dizzy patient would be immensely costly, it is practically necessary to accept that with standard medical care, some patients may not be immediately diagnosed (if at all).

Testing to establish this diagnosis is individual to each patient, but usually will include blood tests for anemia and diseases of the circulation, an MR or CT angiogram test to visualize the blood vessels in the head and neck, and a hearing test and ENG test to exclude ear disease. Other common tests include the CT scan, EEG, EKG and Holter or event monitor. Vertebral artery Doppler may be helpful in some(Sakaguchi et al. 2003).

Recent studies suggest that atrial fibrillation occurs frequently in persons with TIA's, and that ambulatory cardiac monitoring may be indicated more often than thought in the past (Tayal et al, 2008; Cotter et al, 2013). This is particularly relevant as atrial-fibrillation is often treated with stronger blood thinners than other stroke-related conditions (see discussion of Coumadin treatment below). We think that the less invasive systems -- such as simply ambulatory event monitoring, are usually more sensible than the "implantable" systems, which seem to us to be both expensive as well as overly invasive. There are now "over the counter" heart monitoring devices that may make things easier.


As noted above, TIA is a strong risk factor for stoke. The cumulative risk of stroke in persons having TIA's is about 18% in untreated patients, and about 10% in treated patients. The risk is highest in the first month (4-8%), and 12-13% in the first year (Toole, 1991). About 11% of persons with TIA in the emergency department will have a stroke within 90 days. There is also a significant risk of having a heart attack.




The brainstem graphic is courtesy of Northwestern University