Can Tear Osmolarity Identify Dry Eyes?

Christopher J. Rapuano, MD


January 17, 2013

Evaluation of Tear Osmolarity in Non-Sjögren and Sjögren Syndrome Dry Eye Patients With the TearLab System

Szalai E, Berta A, Szekanecz Z, Szûcs G, Módis L
Cornea. 2012;31:867-871

Study Summary

Szalai and colleagues evaluated 39 eyes of 21 patients with non-Sjögren dry eye syndrome, 39 eyes of 20 patients with Sjögren dry eye syndrome, and 44 eyes of 22 healthy individuals. The goal was to determine whether tear osmolarity measurements were correlated with other classical dry eye tests in these 3 groups of patients. They measured tear osmolarity with the TearLab system (TearLab; San Diego, California) in addition to several other standard objective tests for dry eye disease. The other tests involved measuring conjunctival folds, Schirmer test without anesthesia, tear film break-up time, and corneal fluorescein staining. They also evaluated corneal clarity and the state of the meibomian glands.

They found that tear osmolarity was 297 mOsm/L in non-Sjögren eyes, 303 mOsm/L in Sjögren eyes, and 304 mOsm/L in normal eyes. Tear osmolarity was statistically significantly lower in the non-Sjögren eyes than in the other 2 groups. Increased conjunctival folds (conjunctivochalasis), lower Schirmer value, and shorter tear film break-up times were found in the 2 dry eye groups compared with the normal eyes. No significant correlation was found between tear osmolarity and any of the dry eye tests.


Increased tear osmolarity has long been believed to be directly related to dry eye syndrome. Elevated tear osmolarity was believed to be a result of and also an exacerbating factor in dry eye disease. Tear osmolarity testing has been around for more than a decade. The hope was that tear osmolarity measurement could be used as a litmus test to diagnose dry eye disease. If the result was above a certain number, the patient had dry eye disease. Otherwise, the patient did not.

Earlier systems for measuring tear osmolarity were unable to differentiate dry eye patients from non-dry eye patients. The recent hope was that this relatively new system to measure tear osmolarity would more accurately differentiate dry eye patients from non-dry eye patients. Unfortunately, the results of this study found that tear osmolarity measurements using the TearLab system did not distinguish dry eyes from normal eyes. In fact, the non-Sjögren eyes had a lower tear osmolarity than normal eyes. The reason for this finding is unclear. Perhaps a single measurement does not represent the average tear osmolarity over a day or a week or a month. Perhaps the actual sampling technique alters the results in some patients by causing reflex tearing, thereby diluting the tears and reducing osmolarity. Perhaps these patients didn't have pure dry eye disease (which is true for most patients), but instead had blepharitis or conjunctivochalasis, which may not be associated with high osmolarity. Or perhaps tear osmolarity just isn't as indicative of dry eye disease as we originally thought.

The scientific study of dry eyes has been problematic for many years. Patients naturally focus on the symptoms, whereas doctors focus more on the signs. The main issue is that there is no gold standard test for dry eyes. Furthermore, dry eye symptoms (important to patients) do not often correlate with more objective signs (important for research) of dry eye disease. Consequently, the diagnosis of dry eye disease has traditionally incorporated a combination of signs and symptoms. Signs often include corneal and conjunctival staining with fluorescein, lissamine green, or Rose bengal; low Schirmer score; low tear film break-up time; and low tear meniscus. Other signs that may be evaluated are meibomian gland dysfunction and conjunctivochalasis. These signs are often correlated with dry eye disease, but whether they are a cause or a result of dry eye disease, or just an association, is hotly debated. Symptoms of dry eye disease vary greatly, ranging from dryness, grittiness, irritation, and burning to excess tearing. One validated dry eye questionnaire is the Ocular Surface Disease Index, which is often used in studies. It is puzzling that Szalai and colleagues did not evaluate dry eye symptoms in any manner in this study. It would have been interesting to know whether tear osmolarity was correlated with dry eye symptoms.

Although some studies have shown predictive value of tear osmolarity measurements with signs and symptoms of dry eye disease, many others have not. I agree with the investigators' conclusions that the TearLab osmolarity measurements should not be used alone as an indicator of dry eye disease. Unfortunately, there is no stand-alone test for dry eyes. I put the most weight on symptoms. I also evaluate the lids (for blepharitis), the conjunctiva (for conjunctivochalasis and dryness), and the corneas (for dryness). I then perform ancillary testing. I find tear film break-up time, fluorescein staining of the cornea, lissamine green staining of the conjunctiva, and Schirmer testing (with anesthesia) to be most helpful. Although a single tear osmolarity measurement may not be a litmus test for the diagnosis of dry eye, serial measurements may be useful for evaluating the success of treatment of dry eyes. Further studies will tell.