Ambulatory Blood Pressure and Diseases of the Eye: Can Low Nocturnal Blood Pressure Be Harmful?

One of the contemporary themes in the field of 24‐hour ambulatory blood pressure monitoring (ABPM) is that a high nocturnal blood pressure (BP) level may be one of the better predictors of cardiovascular risk, and there is increasing acceptance of the general adage that “the lower the BP, the better,” particularly during the night, when the BP is usually at its lowest. The National Institutes of Health will soon be launching a new clinical trial (Systolic Blood Pressure Intervention Trial [SPRINT]) that will investigate the benefits of aggressive reduction of BP to <120/80 mm Hg. These considerations raise the issue of whether there are circumstances in which low BP may actually be harmful. The idea is not new, and it originated with the “J‐curve” hypothesis, ¹ which stated that lowering diastolic BP to <85 mm Hg may increase the risk of recurrent cardiovascular events in patients with coronary artery disease, on the grounds that coronary artery perfusion becomes impaired in persons with diseased arteries when BP is low. This concept has not been widely accepted, however, and a recent Scientific Statement from the American Heart Association has recommended that patients with known coronary artery disease should have their BP lowered to <130/80 mm Hg. ²

THE DISEASES OF THE EYE

One area in which there have been suggestions that low BP may be harmful, which has largely been ignored by the hypertension community, is diseases of the eye. Although we recognize the eye as one of the target organs affected by hypertension, the conventional wisdom is that only very high BP is of concern. However, the ophthalmologic literature includes a substantial number of studies that have used ABPM and concluded that nocturnal hypotension can damage the eye, particularly in anterior ischemic optic neuropathy (AION) and glaucoma. The idea that low nocturnal BP contributes to the progression of visual field loss from glaucoma certainly seems to have taken hold among ophthalmologists, at least in Europe. A Belgian ophthalmologist recently told me that he recommends that such patients have ABPM performed by their family doctors, and if the nocturnal BP is low, that they take a spoonful of salt with Vichy water before they go to bed, on the grounds that this will prevent an excessive decrease in BP during the night.

The idea that glaucoma may be at least in part a vascular disease is not new, and traditionally hypertension and diabetes have been regarded as major risk factors for its development. However, as discussed below, this view may no longer be tenable.

NOCTURNAL HYPOTENSION AND AION

The suggestion that nocturnal hypotension may damage the eye originated with the observation that about 75% of patients with AION first notice a visual defect on waking. ³ Like the cerebral arteries, the retinal arteries are end arteries, such that there may be areas of the retina that are in the “watershed” between 2 arteries. ³ Depending on the anatomy of the retinal circulation, there may be between 2 and 4 watershed zones in the retina. The idea is that if the perfusion pressure is low, such zones may be particularly vulnerable to ischemia. It is thought that there is autoregulation of the retinal circulation, as in the cerebral vessels. Indirect evidence for the loss of autoregulation in the eye comes from reports of permanent blindness when patients with malignant hypertension have their BP lowered suddenly and develop AION. ³

Hayreh and associates ⁴ have studied patients with AION and glaucoma with ABPM and concluded that nocturnal hypotension may contribute to the development of AION; they also observed that patients with AION who were taking antihypertensive drugs showed a significant relationship between nocturnal hypotension and progressive visual field deterioration. ⁵ The data here are not consistent, however. Landau and colleagues ⁶ compared the 24‐hour BP profile in AION patients and healthy controls and found no difference in the dipping pattern, although the morning rise of BP was less pronounced in AION.

EPIDEMIOLOGY OF GLAUCOMA

Most of the literature on the epidemiology of glaucoma in relation to cardiovascular disease states that both hypertension and diabetes are risk factors for the development of glaucoma, together with age, family history, and ethnicity (it is much more common in blacks than whites). An example is a report from the Framingham Heart Study, ⁷ which found that the prevalence was 1.7 times higher in hypertensive than in normotensive participants, and it was more than twice as common in diabetics than in those without diabetes. However, the presence of clinical cardiovascular disease was not related to glaucoma. Other studies have reported increased prevalence in both hypertensive and hypotensive patients. ⁸ Most of these studies defined glaucoma by an increased intraocular pressure ([IOP] >21 mm Hg), measured noninvasively by an applanation tonometer applied to the eyeball. The problem with this criterion is that it fails to allow for normal‐tension glaucoma (NTG), in which the IOP is normal, and that the threshold of 21 mm Hg is quite arbitrary. Thus, prevalence studies have shown that up to half of patients with glaucoma experience damage at normal levels of IOP. The new definition of glaucoma depends on detecting optic disc changes and visual field defects and does not include IOP. ⁹

When glaucoma is defined by the new criteria, epidemiologic studies paint a very different picture. An example is the Barbados Eye Studies, ¹⁰ which examined new cases of glaucoma over a 9‐year period. Despite the high prevalence of both diabetes and hypertension in this largely black population, neither of them were significant predictors of new‐onset glaucoma. There was actually a trend (although not significant) for lower systolic and diastolic BP to be associated with increased risk of glaucoma. As found in the earlier studies, a high IOP still predicted glaucoma. The best BP‐related predictor was a low optic perfusion pressure (defined as the difference between the systemic BP and the IOP), where a 10‐mm Hg lower perfusion pressure was associated with a 34% increase in risk. Another analysis from the same study confirmed that hypertension and diabetes were significant predictors of increased IOP over a 4‐year period. ¹¹ Similar conclusions have been obtained from a report from the Early Manifest Glaucoma Trial (EMGT), ¹² which compared 2 types of treatment in patients with glaucoma and found that a lower systolic perfusion pressure, lower systolic BP, and cardiovascular disease history all predicted progression of the disease during the course of the study. Thus, despite the very different conclusions regarding the role of the arterial BP depending on which definition is used, the case that glaucoma is at least in part a vascular disease is still very strong.

NOCTURNAL HYPOTENSION AND GLAUCOMA

In the past 10 years or so, there have been a number of studies that have investigated the value of ABPM in patients with glaucoma and have concluded that abnormalities in the diurnal rhythm of BP contribute to its development. These have generally been of 2 sorts: cross‐sectional studies of patients with and without glaucoma and prospective studies of the progression of visual field defects.

At least 4 studies that compared patients with glaucoma with healthy control participants actually found no significant difference in dipping between the 2 groups. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ A fifth study of 51 patients with NTG and 28 controls found that the only BP difference between the NTG patients and the controls was the nighttime diastolic BP, which was slightly higher in the NTG patients (69 mm Hg vs 64 mm Hg). ⁸ However, there was also a slightly higher variability of nocturnal BP (P=.03 for systolic and .02 for diastolic) in the NTG patients. One study found that patients with NTG had a greater number of low systolic BP readings (<90 mm Hg) during the night in comparison with patients with high‐tension (ie, high‐IOP) glaucoma. ¹⁷ But one of the other studies found that nocturnal diastolic BP variability was reduced in NTG patients compared with controls (P<.05). ¹⁶

The studies investigating whether nocturnal hypotension is associated with visual loss from glaucoma also gave very mixed results. Two reported that progression of visual field loss was associated with increased dipping. ¹⁴ , ¹⁸ Two others, however, found no clear relation between dipping and progression but showed that progression tended to be greater in nondippers and extreme dippers (although the numbers of patients were very small). ¹⁹ , ²⁰ And 2 studies concluded that dipping was less in patients whose visual fields got worse. ⁵ , ¹⁵

Despite these very mixed findings, the individual papers were surprisingly optimistic about the role of nocturnal BP in glaucoma, including statements such as “Increased fluctuation of BP may lead to ocular perfusion pressure fluctuation and may cause ischemic episodes of the optic nerve head” ⁸ and “The nocturnal dip of systolic pressure should be considered as a predictive factor of disease progression in NTG and primary open angle glaucoma.” ²⁰

OTHER CLINICAL CONDITIONS IN WHICH NOCTURNAL HYPOTENSION MAY INCREASE RISK

The idea that an excessive nocturnal decrease in BP may be harmful is not limited to diseases of the eye and may apply to cerebral and myocardial ischemia. The idea of the J‐curve or U‐curve relationship between cardiovascular events and BP was first mooted for myocardial ischemia by Cruikshank and colleagues ¹ in 1987 in a small, uncontrolled study. Although the concept was dismissed by the recent American Heart Association statement about target BP in patients with known coronary heart disease, there is recent evidence that supports the idea: in the International Verapamil‐Trandolapril Study (INVEST), ²¹ Messerli and coworkers reported that there was a J‐shaped relationship between BP and all‐cause mortality, which was most marked for diastolic BP. A few studies have shown that ischemic episodes during electrocardiographic monitoring may occur during the night. ²² Pierdomenico and associates ²³ classified patients with coronary artery disease according to their dipping status (nondippers, dippers, and extreme dippers) and performed Holter recording for 24 hours. Patients were monitored before and after treatment with nitrates, atenolol, or verapamil, which reduced the frequency of ischemic episodes to the same extent in all 3 groups during the day but had very different effects at night. The frequency of episodes was reduced in the nondippers, but it was increased in the extreme dippers, presumably because of excessive BP reduction.

It is generally assumed that the adage of “the lower the BP, the better” applies to stroke reduction, and the results of the Perindopril Protection Against Recurrent Stroke Study (PROGRESS) ²⁴ are often quoted in support of this. In this study, it was observed that antihypertensive treatment reduced recurrent strokes even in persons who were normotensive to begin with. The increased risk of stroke in extreme dippers remains controversial and has been reported in one Japanese study, ²⁵ but not in another. ²⁶ In addition, it may be the exaggerated morning surge of BP rather than the low nocturnal pressure that best predicts risk. Nonetheless, there is some other evidence that there may be a U‐curve relationship between ischemic stroke and BP. A review of 1004 cases of cerebral infarction and 1097 cases of cerebral hemorrhage ²⁷ analyzed the distributions of BP on admission to hospital and found that for cerebral infarction the mortality was highest at the 2 extremes of BP. With cerebral hemorrhage, there was more of a J‐shaped relationship: mortality did not increase at the lowest BP levels, but it was not any lower than at moderate levels of pressure. While these results are disturbing, it should be remembered that the BP at the time of hospitalization may not reflect the patient's true pressure.

CONCLUSIONS

Despite the intense interest in the role of nocturnal hypotension in the development of AION and glaucoma, the evidence obtained with ABPM studies is very inconsistent. Nevertheless, there is reason to think that the concept may be plausible. The introduction of the new definition of glaucoma that does not include IOP has led to the conclusion that systemic hypertension is not a major risk factor, but a low ocular perfusion pressure is. It will take larger and better‐controlled studies to determine whether a low nocturnal BP may be harmful in some patients, and the upcoming Systolic Blood Pressure Intervention Trial (SPRINT) study might be able to provide some answers.