Dynamic balance between aqueous inflow and outflow facility determines circadian fluctuations in intraocular pressure (IOP). This occurs in healthy individuals and in patients with glaucoma, whether or not treatment has been commenced. The range in IOP variation can be extreme, with the pattern of change also varying between individuals. Patients may demonstrate diurnal or nocturnal IOP peaks.

Patients with glaucoma progression despite IOP apparently within the target range during office/clinic visits remain a significant challenge. Although IOP-independent factors might be important, so might IOP fluctuation. While clinic-based IOP measurements correlate reasonably with mean circadian IOP, they are not predictive of the magnitude of fluctuations or the peak IOP. Patients with highly variable IOP, particularly those with nocturnal peaks, might be missed.

Studies examining IOP fluctuation and disease progression suggest a causal relationship. Home tonometry reveals large fluctuations in IOP in primary open-angle glaucoma patients with an average range of 10 mm Hg from a mean baseline IOP of 16.4 mm Hg. Those with large fluctuations in IOP were much more likely to demonstrate glaucomatous field progression compared to those with small or no fluctuation, independent of baseline IOP. Similarly, post hoc analysis of data from the Advanced Glaucoma Intervention Study (AGIS), to assess IOP control (level and long-term fluctuation) and progression, revealed that eyes with IOP <18 mm Hg at every postoperative visit had relatively stable visual field loss compared with eyes whose IOP fluctuated above 18 mm Hg on more than 50% of visits, in which field progression was definite.

Clinically we cannot measure IOP continuously; we can only sample it. This hampers our attempt to identify and to treat more effectively patients with unstable IOP and increased potential for glaucoma progression. We miss both the extent of IOP fluctuations and its circadian peak. Twenty-four-hour pressure monitoring is impractical. Home IOP monitoring has been trialed but is too erratic. Diurnal tension curves (DTC) have been used experimentally and clinically and do show fluctuation in IOP during daylight hours. However, a DTC misses IOP peaks occurring overnight, thereby underestimating IOP fluctuations.

This is particularly so when sleep laboratory 24-hour IOP measurement demonstrates peak levels routinely during hours of sleep, if measured in the recumbent (“usual night-time”) position. Gradually we are identifying several types of IOP fluctuations, which might be of different significance for individual patients: long-term intervisit IOP fluctuations (as assessed in the AGIS post hoc analysis), short-term circadian fluctuations (as assessed in the Asrani study), and very-short-term fluctuations from second to second or minute to minute (occasioned by eye rubbing, eyelid squeezing, Valsalva effects, postural changes, unassessed to date).

The water drinking test (WDT) might be a practical and reliable means to estimate peak circadian IOP: a volume of water is imbibed over a short period with IOP compared before and for some time afterward. This has been either a fixed volume of 1 liter or variable, such as 10 mL/kg body weight in 5 minutes. By stressing outflow facility, the WDT might reveal elevations in IOP that are ordinarily experienced during a 24-hour period. After the observation in the late 1920s that water ingestion provoked elevated IOP, the WDT was proposed as a diagnostic test for glaucoma. While we now know that its low sensitivity and specificity make it unhelpful diagnostically, recent work points to its potential as a predictor of peak IOP, and thus IOP fluctuations.

The physiology of the WDT is not fully understood. Neither vitreous hydration nor increased aqueous ultra-filtration explains the increases in IOP it provokes. Autonomic nervous stimulation and/or increased episcleral venous pressure remain possible mechanisms. Systemic hypertension occurs following water drinking in autonomic failure, in quadriplegia, after cardiac transplantation, and to a lesser extent in older healthy individuals. Sympathetic stimulation is thought to produce this response and may also modify IOP via yet-to-be-determined mechanisms. Alternatively, consumption of 1 liter of water elevates episcleral venous pressure (EVP) and causes a transient period of negative aqueous outflow. This is likely the result of reduced outflow facility from a decreased pressure gradient across the trabecular meshwork, but other mechanisms have been suggested. Episcleral venous pressure may be implicated in both IOP instability and glaucoma progression as it is elevated in both open-angle and “normal tension” glaucoma patients. The WDT might unmask pathologic EVP effects.

Our group has compared the WDT (10 mL/kg body weight) with diurnal tension curves. All patients had pretesting IOP <18 mm Hg that elevated beyond 18 mm Hg in 12% of patients with the DTC and in 20% with the WDT. Although the extent of IOP fluctuations seen with the DTC correlated poorly with that seen after the WDT, the correlation for peak IOP was very strong. Earlier, in a similar study, Malerbi and associates also compared the DTC with the WDT (1-liter fixed volume). Similar findings highlight the good correlation between peaks seen during the DTC and those provoked by the WDT. That slightly more patients demonstrated peaks after the WDT in both studies points to the potential of this test as a predictor of peak IOP over a 24-hour period.

The response to the WDT appears worse in the presence of progressive disease. Comparison between eyes in patients with asymmetric glaucomatous field loss but similar baseline IOP demonstrates an average WDT-induced peak IOP 0.7 mm Hg higher in the worse eye. Similarly, analysis of glaucoma patients grouped as either nonprogressive or progressive shows an average peak IOP 1.9 mm Hg higher in the latter group despite no difference in age, sex, race, baseline IOP, or visual field indices. Although small, these differences might prove significant over time. The findings support both the IOP peak/fluctuation theory of disease progression and a possible, practical role for the WDT in its assessment.

The utility of the WDT extends beyond unmasking IOP fluctuations. It might also be used to assess the efficacy of treatment. Agents that increase outflow facility (such as prostaglandin and prostamide agonists, alpha-2 agonists, and cholinergic agonists) attenuate the WDT response, whereas this effect is not seen with agents reducing inflow (including beta-blockers and carbonic anhydrase inhibitors). Better still is the response seen following surgery (mean maximum IOP: 11.7 ± 2.6 mm Hg after trabeculectomy versus 17.3 ± 2.7 mm Hg after medical treatment). Laser trabeculoplasty reduced nocturnal peak IOP post treatment, but its effect on the WDT response is yet to be reported.

Further research into the WDT is needed. Areas include elucidation of the underlying physiology of the IOP response, correlation with circadian IOP changes, more precise definitions of “normal” versus “abnormal” WDT responses, effects of argon/selective laser trabeculoplasty on IOP peak, and whether it is safe to use in the setting of renal impairment or congestive cardiac failure. Even without this information, we have learned much in the past decade about the usefulness of the WDT to assess IOP fluctuation and to detect the circadian IOP peak. It is technically simple, within the resources of most clinicians, and available to all seeking to enhance the effectiveness of our treatment of some of our most challenging patients.