Nocturnal Systemic Hypotension Increases

the Risk of Glaucoma Progression

Mary E. Charlson, MD,1 Carlos Gustavo de Moraes, MD,2,3 Alissa Link, MPH,1 Martin T. Wells, PhD,4
Gregory Harmon, MD,5 Janey C. Peterson, EdD,1 Robert Ritch, MD,2 Jeffrey M. Liebmann, MD2,3

Objective: The objective of this prospective, longitudinal study of patients with normal-tension glaucoma
(NTG) was to determine whether patients with nocturnal hypotension are at greater risk for visual field (VF) loss
over 12 months than those without nocturnal hypotension.
Design: Prospective, longitudinal study.
Participants: Consecutive patients with NTG with at least 5 prior VF tests were screened for eligibility.
Methods: The baseline evaluation assessed demographic and clinical characteristics, covering systemic
comorbid conditions, including systemic hypertension. All oral and ophthalmologic medications were recorded. A
complete ophthalmological examination was performed at baseline and follow-up. Patients had their blood
pressure (BP) monitored every 30 minutes for 48 hours with an ambulatory recording device at baseline and 6 and
12 months.
Main Outcome Measures: The primary outcome was based on the global rates of VF progression by linear
regression of the mean VF threshold sensitivity over time (decibels/year).
Results: Eighty-five patients with NTG (166 eyes; mean age, 65 years; 67% were women) were included. Of
the 85 patients, 29% had progressed in the 5 VFs collected before study enrollment. The nocturnal mean arterial
pressure (MAP) was compared with the daytime MAP. Multivariate analysis showed that the total time that sleep
MAP was 10 mmHg below the daytime MAP was a significant predictor of subsequent VF progression (P<0.02).
Conclusions: Cumulative nocturnal hypotension predicted VF loss in this cohort. Our data suggest that the
duration and magnitude of decrease in nocturnal blood pressure below the daytime MAP, especially pressures that
are 10 mmHg lower than daytime MAP, predict progression of NTG. Low nocturnal blood pressure, whether
occurring spontaneously or as a result of medications, may lead to worsening of VF defects. Ophthalmology 2014;-
:1e9 ª 2014 by the American Academy of Ophthalmology.

Glaucoma is an optic neuropathy characterized by progres- perfusion in vital organs, if the BP decreases below the
sive loss of retinal ganglion cells and their axons, which leads lower limit, ischemia occurs.
to structural and functional damage to the optic nerve. In this first longitudinal study of NTG using 48-hour BP
Despite being an arbitrary definition, primary open-angle monitoring at 6-month intervals, we tested the hypotheses
glaucoma (POAG) with intraocular pressure (IOP) within that nocturnal systemic hypotension is associated with
the statistically normal range (i.e., <21 mmHg) is commonly progressive VF loss and that the extent and duration of the
defined in practice and research as normal-tension glaucoma nocturnal decrease in mean arterial pressure (MAP) predict
(NTG), when in fact POAG and NTG share similar patho- progression of VF loss. The objective was to determine
genesis. Presently, IOP reduction, the only treatable risk whether patients with nocturnal hypotension are at greater
factor in glaucoma, tends to slow, not halt, disease progres- risk for VF loss over 12 months than those without
sion, suggesting that mechanisms other than IOP may nocturnal hypotension.
contribute to visual field (VF) loss, particularly in the lower
ranges of IOP in patients with glaucoma.1e4 Epidemiologic
studies, cross-sectional studies, and randomized clinical trials Methods
(RCTs) have suggested that systemic blood pressure (BP)
plays a role in pathogenesis of glaucoma. Low BP and lower This longitudinal study was approved by the New York Eye and
diastolic perfusion pressure were among the first risk factors Ear Infirmary and Weill Cornell Medical College Institutional
associated with incident glaucoma in population-based Review Boards for Human Research. The study was performed in
studies.3,5e11 Thus, many studies suggest that nocturnal hy- compliance with the Health Insurance Portability and Account-
potension may identify patients at higher risk for VF pro- ability Act and the tenets set forth in the Declaration of Helsinki.
Written informed consent was obtained from all eligible patients. A
gression.9,12e17 Presumably, it is the difference between
total of 456 patients who were followed at the New York Eye and
arterial and venous pressure that determines blood flow after Ear Infirmary with a diagnosis of NTG were screened for potential
vessels are maximally dilated by autoregulation that mat- eligibility. At baseline examination, the diagnosis of NTG was
ters.18 When systemic BP decreases 20 mmHg below the based on the presence of glaucomatous optic neuropathy, abnormal
patients’ usual BP, vessels dilate to preserve constant 24-2 Swedish Interactive Thresholding Algorithm (SITA) standard

 2014 by the American Academy of Ophthalmology ISSN 0161-6420/14/$ - see front matter 1
Published by Elsevier Inc. http://dx.doi.org/10.1016/j.ophtha.2014.04.016
 Ophthalmology Volume -, Number -, Month 2014

Figure 1. Consolidated standards of reporting trials diagram. BP ¼ blood pressure; IOP ¼ intraocular pressure; NTG ¼ normal-tension glaucoma.

automated perimetry (SAP) examinations (SITA-SAP, Humphrey At follow-up examinations, complete ophthalmic examination
Visual Field Analyzer; Carl Zeiss Meditec, Inc., Dublin, CA) in 1 was performed, including best-corrected visual acuity, slit-lamp
or both eyes, open anterior chamber angles, no identifiable sec- biomicroscopy, gonioscopy, dilated fundoscopy, stereodisc
ondary cause of glaucoma, and all known untreated IOPs
21 photography, automated perimetry, and Goldmann applanation
mmHg by Goldmann applanation tonometry. Glaucomatous optic tonometry. Central corneal thickness was measured using ultra-
neuropathy was defined on the basis of stereophotographic evalu- sonic pachymetry (DGH-550; DGH Technology Inc., Exton, PA).
ation by glaucoma specialists using the following criteria: focal or Visual function was assessed using 24-2 SITA-SAP tests at all
diffuse neuroretinal rim thinning, focal or diffuse retinal nerve fiber follow-up visits. Stereoscopic optic disc photographs were
layer loss, or an inter-eye vertical cup-to-disc ratio asymmetry reviewed by 2 masked experts (C.G.dM., J.M.L.) searching for disc
>0.2 not explained by differences in disc size. The 24-2 VF tests hemorrhages (i.e., a splinter-like or flame-shaped hemorrhage on or
were considered abnormal if the glaucoma hemifield test result was within the retinal nerve fiber layer or neuroretinal rim).21 If the 2
outside normal limits or the pattern standard deviation (PSD) had a investigators disagreed, a third investigator adjudicated.
P value <5%. At least 2 consecutive VF results had to be reliable Systemic BP was monitored every 30 minutes for 48 hours with
on the basis of false-positive rates
25%, false-negative rates, and an ambulatory recording device. Spacelabs ambulatory BP moni-
fixation losses
33%. Patients with significant lens opacity, ocular toring equipment (Spacelabs Healthcare, Issaquah, WA) was
conditions that would affect VF results, fewer than five 24-2 SITA- used.22e24 Patients used a diary to record times of going to bed and
SAP tests, IOPs 21 mmHg, or severe glaucoma VF loss (mean rising, taking medications, and any systemic symptoms. Of those
deviation19 <20 decibels [dB]) were excluded. A total of 216 who consented to participate, 85 patients completed the baseline
patients were eligible for the study. As shown in Figure 1, 97 48-hour BP recording; overall, 67 patients had 3 BP sessions, 5
patients consented to participate. patients had 2 BP sessions, and 13 patients had 1 BP session (Fig
Demographic and clinical history, including the presence of 1). Follow-up evaluations were performed at 6 and 12 months, at
atherosclerotic disease, diabetes, hypertension, migraine, and which time the medical and ophthalmic histories were reviewed for
Raynaud’s disease, were assessed. Comorbidity was assessed using change, 48-hour ambulatory BP measurement was obtained, and
the Charlson index.20 All oral medications were recorded. A the eye examination, including tonometry and perimetry, was
complete ophthalmic history and searching for evidence of prior repeated. The most recent 5 VFs before enrollment and the 3 VFs
disc hemorrhage, laser and incisional surgical interventions, was obtained during the study period were used for outcome analyses.
obtained. Automated pointwise linear regression analysis was performed

2
 Charlson et al 
Nocturnal Blood Pressure and Glaucoma Progression

Table 1. Clinical and Demographic Characteristics of Patients with Progression Before Enrollment versus Patients without Prior
Progression

Prior Progression (n[25) % No Prior Progression (n[60) % P Value

Age, yrs 65.313.5 65.910.3 0.84
Female sex 19 76.0 38 63.3 0.26
Angina 0 0.0 3 5.0 0.55
Myocardial infarction 2 8.0 1 1.7 0.21
Cerebrovascular accident 0 0.0 4 6.7 0.32
Hypertension 6 24.0 21 35.0 0.32
Antihypertensive medication
Angiotensin converter enzyme inhibitor 2 33.3 3 14.3 0.30
Angiotensin II receptor blockers 1 16.7 5 23.8 1.00
b-Blocker 4 66.7 5 23.8 0.14
Calcium channel blocker 3 50.0 5 23.8 0.32
Diuretics 1 16.7 4 19.1 1.00
Baseline blood pressure
Nonhypertensives
Systolic 132.915.5 133.517.9 0.92
Diastolic 76.88.5 79.610.6 0.32
Mean arterial pressure 96.59.1 97.712.2 0.71
Hypertensives
Systolic 135.310.3 142.810.9 0.15
Diastolic 82.312.9 84.57.1 0.59
Mean arterial pressure 101.213.7 104.67.5 0.58
Diabetes 1 4.0 3 5.0 1.00
Renal disease 3 12.0 1 1.7 0.074
Migraine 5 20.0 12 20.0 1.00
Raynaud’s disease 5 20.0 10 16.7 0.76
Thyroid condition 6 24.0 16 26.7 0.80
Comorbidity
0 11 44.0 31 51.7 0.70
1 5 20.0 13 21.7
2 6 24.0 11 18.3
3 3 12.0 5 8.4
Age at diagnosis 5312 5512 0.53
Family history of glaucoma 10 40.0 28 46.7 0.65
Glaucoma medication
a-Agonists 7 28.0 12 20.0 0.41
b-Blockers 13 52.0 28 46.7 0.65
Pilocarpine 2 8.0 4 6.7 1
Carbonic anhydrase inhibitors 8 32.0 23 38.3 0.58
Prostaglandins 21 84.0 54 90.0 0.47

Data are n (%) or mean  standard deviation.

using PROGRESSOR software (version 3.3; Medisoft, Ltd., Leeds, linear regression progression criteria have been used and validated
UK) providing slopes (dB/year) of progression both globally and in previous studies.26,27
locally for each point based on threshold sensitivity maps, as well
as its level of significance (P values). Details of the software have Sample Size and Data Analysis
been described elsewhere.25 Global rates of sensitivity change are
provided automatically by PROGRESSOR and represent the We projected 22% progression over 1 year.9 To increase the power
slopes of the average threshold sensitivity of each field of the of the study, we preferentially enrolled patients who had previously
sequence of VF tests analyzed using least-squares linear regres- progressed based on clinical judgement, because they are thought
sion. Therefore, the main outcome measure was the numeric value to be at higher risk for subsequent progression.28,29 With a po-
of the rate of VF change for each eye over the study period. The wer of 80% and an a level of 0.05, we detected a difference of
mean deviation and PSD values of the baseline VF tests were proportions of 0.22 with 72 eyes per group, or 144 eyes overall.
recorded. For a binary classification of progression (i.e., “pro- With a loss to follow-up of 15%, a total sample of 170 eyes was
gressing” vs. “stable” eyes), pointwise linear regression progres- required.
sion criteria were applied to the sequences of 8 SITA 24-2 VF tests. By using the data from the patient’s diary card for each 48-hour
A VF test point was deemed “progressing” if it presented a slope of period, we calculated (1) the 48-hour daytime average MAP while
sensitivity change over time more negative than 1.0 dB loss/year the patient was awake; (2) the 48-hour nocturnal MAP while the
at P<0.01. An eye was deemed “progressing” if the sequence of patient was asleep; and (3) the variability of nocturnal MAP. We
VF tests presented at least 2 adjacent VF test points in the same characterized BP in several ways: (1) the daytime average MAP at
hemifield meeting the aforementioned criterion. These pointwise each follow-up, which we calculated using only the data from

3
 Ophthalmology Volume -, Number -, Month 2014

Table 2. Ophthalmologic Characteristics by Individual Eye for Eyes that Showed Prior Progression versus No Prior Progression (n¼166)

Prior Progression (n[28) % No Prior Progression (n[138) % P Value

Optic
Mean IOP over 3 yrs (SD; mmHg) 12.92.5 13.22.4 0.62
Myopia 8 28.6 44 31.9 0.73
Visual acuity (logMAR) 0.190.42 0.160.27 0.69
Disk size (microns) 1.840.35 2.010.35 0.16
Rim area (microns) 0.910.28 1.020.48 0.25
Spheric equivalent (diopters) 2.43.53 1.877.49 0.69
Vertical cup-to-disc ratio 0.800.11 0.800.13 0.78
CCT (microns) 53543 53934 0.64
MD (dB) 10.535.07 7.486.15 0.02
PSD (dB) 10.753.10 7.984.39 0.00
Prior disc hemorrhage 13 46.4 46 33.3 0.19
Threat to fixation 21 75.0 103 74.6 0.97
Cataracts 17 60.7 62 44.9 0.13
Previous surgery
ALT 4 14.3 10 7.2 0.26
SLT 4 14.3 28 20.3 0.60
Filtering surgery 2 7.1 8 5.8 0.68
Lasik 0 0.0 4 2.9 1.00
Cataract surgery 9 32.1 26 18.8 0.12

ALT ¼ argon-laser trabeculoplasty; CCT ¼ central corneal thickness; dB ¼ decibels; IOP ¼ intraocular pressure; logMAR ¼ logarithm of the minimum
angle of resolution; MD ¼ mean deviation; PSD ¼ pattern standard deviation; SD ¼ standard deviation; SLT ¼ selective-laser trabeculoplasty.
Progression was defined on the basis of pointwise linear regression criteria, as described in the text.
Statistical significance at 5% level is shown in boldface.

when the patient is awake, according to his/her diary card; (2) without progression. Most patients were female, the average age
measures of nighttime mean MAP; and (3) variability of BP, was 65 years, and 32% had systemic hypertension, of whom
particularly its departure from its usual daytime MAP. The first 77% were taking oral medications to reduce systemic BP. Most
measure of variability was simply the time (averaged across the 48 patients were receiving prostaglandin analog eye drops, and half
hours) that the nighttime BP was less than the daytime average. of patients were receiving b-adrenergic antagonist drops. There
The second measure of variability is similar: the time spent at least were no significant differences in any of the characteristics listed
some level c below the daytime average, where c¼10 and 20. The in Table 1 for those who did and did not have progression
third measure of variability was the area under curve, representing during the 36 months before the first 48 hour BP monitoring.
total nighttime time*magnitude spent below the daytime average; Table 2 displays the baseline ophthalmologic characteristics
the fourth was the area below the defined constant c. according to whether the specific eye (right or left eye) had prior
The main analysis used the rate of VF change for each eye over progression. The mean follow-up IOP was 13 mmHg. Some 36%
the study period. A seemingly unrelated regression equations of eyes had a threat to fixation, whereas 74% of eyes had a prior
approach was used to assess the relationship of predictors to disc hemorrhage. The mean corneal thickness was 535 mm for
outcome.30 The multivariate approach is appropriate because the both groups. As expected change in mean deviation and PSD
outcome equation for each eye has eye-specific (right eye/left was significantly greater in patients with progression in the prior
eye) explanatory variables. In addition, because the 2 eye outcomes 36 months. Otherwise, there were no other significant differences
are correlated (BreuschePagan test of independence; P ¼ 0.04), in eyes according to recent progression.
the statistical efficiency of the estimators is gained by estimating
the 2 equations simultaneously. The regression equations control Outcomes
for age, sex, prior progression, IOP, presence of cataract, central
corneal thickness, use of topical pilocarpine, a-agonists, b- Overall, 26 of the 85 patients (29%) progressed after 1 year of
blockers, oral statins, and hypertension medications. Hypothesis follow-up on the basis of the pointwise linear regression progres-
tests use a Wald test to assess the significance of effects in the right sion criteria. Of the patients who had progression before study
and left eye regression equations simultaneously rather than entry, 52% (13 of 25) continued to progress after 1 year of follow-
separate tests using each eye’s equation. To avoid the undesired up versus 21% of those without prior progression. Thus, prior
effect of collinearity between predictors (“time” and “area”), 2 progression was a significant predictor of subsequent progression
distinct models were built using each of these predictors separately, (P<0.01).
but also adjusting for the same confounders.
Forty-Eight-Hour Nocturnal Blood Pressures
Results Figure 2 shows the MAP when awake versus the MAP at night at
baseline for a patient who had progression over 12 months. The
Baseline Characteristics total time that patients spent during sleep at MAPs below their
daytime MAP and the total area below their daytime MAP were
Overall, 29% of patients had progression of VF loss during the 36 both evaluated. At any given time point, total time and total area
months before enrollment. Table 1 contrasts the 29% (25/85) of during sleep below daytime MAP were highly correlated
patients with VF progression in the prior 36 months versus those (Pearson’s r ¼ 0.78 at baseline, r ¼ 0.73 at 6 months, and r ¼

4
 Charlson et al 
Nocturnal Blood Pressure and Glaucoma Progression

Figure 2. Example of how mean arterial pressures (MAPs) while awake versus during sleep were calculated. This example depicts a patient who progressed
on the basis of our visual field (VF) criteria (see text). The time range (x-axis) corresponds to the time period awake or asleep during a 48-hour of blood
pressure (BP) monitoring. The top graph depicts the MAP variation while awake (dotted line) and the corresponding average (vertical dashed line). In the
bottom graph, the dashed line depicts the average MAP (calculated in the previous plot) while the patient was awake and the solid line depicts the MAP
variation during sleep.

0.77 at 12 months). The total time that patients spent during sleep predictor of global progression (P ¼ 0.02). The area that represents
below their daytime BP was correlated across baseline and 6 and the time multiplied by the magnitude of nocturnal BP that was >10
12 months (intraclass correlation coefficient ¼ 0.37). The total mmHg below daytime MAP also predicted progression (P ¼ 0.03).
area during sleep below their daytime BP was also moderately Progression also was associated with prior progression (P ¼ 0.01)
correlated at the 3 time points (intraclass correlation and corneal thickness (P ¼ 0.02) and was weakly associated with
coefficient ¼ 0.53). age (P ¼ 0.1). Progression was not associated with previous
As an example, Figure 3 shows the MAP during sleep at cataract surgery (P ¼ 0.78) or the use of ophthalmic b-blockers
baseline for patients who had VF progression, and Figure 4 (P ¼ 0.53), a-2 adrenergic agonists (P ¼ 0.50), or other medica-
shows patients who did not have VF progression. The analysis tions such as statins (P ¼ 0.22). There were no significant differ-
considers progression over 12 months in relation to cumulative ences in progression between systemic hypertensive and
nocturnal BP over three 48-hour periods of monitoring: at base- normotensive individuals (P ¼ 0.59). Over the follow-up year,
line and 6 and 12 months. multivariate analysis showed that IOP in either eye was not a
predictor of subsequent VF progression (P ¼ 0.37).

Predictors of Visual Field Outcomes

Alternate Definitions of Hypotension
To identify predictors of progression, eye-specific multivariate
analyses controlling for age, sex, and other variables, as well as Time. We examined the associations of the progression outcome
eye-specific (right eye/left eye) variables, were used to assess the with the other definitions of time, specifically the time a patient spent
various predictors of progression. Taken together, the three 48- during sleep when his or her MAP was below the daytime MAP and
hour measures of BP contrasting daytime mean BP with BP dur- the time during sleep more than 20 mmHg under the daytime MAP.
ing sleep shows that the total time that the MAP during sleep is Figure 5 shows the distribution of patients according to the time
more than 10 mmHg below the daytime mean is a significant spent under the daytime MAP. Many patients had time under the

5
 Ophthalmology Volume -, Number -, Month 2014

Figure 3. Each graph depicts the mean arterial pressures (MAPs) during
sleep of progressing patients. The time range (x-axis) corresponds to the
time asleep during the baseline 48-hour of blood pressure monitoring. The
dashed line depicts the average MAP while the patient was awake, and the
continuous line depicts the MAP variation during sleep.
Figure 4. Each graph depicts the mean arterial pressure (MAPs) during
sleep of nonprogressing patients. The time range (x-axis) corresponds to the
daytime MAP, but only a few had time <20 mmHg. A multivariate time asleep during the baseline 48-hour of blood pressure monitoring. The
analysis showed that the total time that a patient’s MAP during sleep dashed line depicts the average MAP while the patient was awake, and the
was >20 mmHg below daytime MAP was associated with global continuous line depicts the MAP variation during sleep.
progression (P ¼ 0.01), but total time under daytime MAP was
not (P ¼ 0.35). Thus, the total time <10 mmHg predicted
progression, as did time <20 mmHg; however, not many patients defined decreases in nocturnal MAP consonant with autor-
had long times <20 mmHg. egulatory theory. The range in which autoregulation can
Area. We examined the associations of the progression preserve constant perfusion is approximately 20 mmHg
outcome with the other areas measures, that is, area during sleep surrounding the patients’ usual MAP.31 Once the BP
under the daytime MAP and area during sleep >20 mmHg under
the daytime MAP. Figure 6 shows the distribution of patients
decreases below the lower limit, ischemia occurs.
according to the total area under the daytime MAP. A Ambulatory BP monitoring has long established that BP
multivariate analysis showed that the area under daytime MAP decreases during the night in the majority of individuals,
(P ¼ 0.03) and the area >10 mmHg below daytime MAP (P ¼ but the issue is how much and for how long.32 Nocturnal
0.03) did predict progression, whereas the area >20 mmHg hypotension poses ischemic risk for other regional
below daytime MAP (P ¼ 0.11) was a weaker predictor. Thus, circulations, specifically stroke and myocardial infarction.
cumulative nocturnal hypotension predicts VF loss. The concept of autoregulation provided the framework for
our earlier RCT of patients undergoing coronary artery
bypass that showed maintaining higher MAPs during
Discussion cardiopulmonary bypass significantly reduced major
ischemic cardiac and neurologic morbidity and mortality.17
In a population with treated NTG followed longitudinally Epidemiologic studies, cross-sectional studies, and RCTs
with 48-hour BP monitoring, this study demonstrated that have suggested that systemic BP plays a role in the patho-
the duration and magnitude of nocturnal systemic hypo- genesis of glaucoma. Low BP and lower diastolic perfusion
tension, that is, MAP during sleep below the daytime MAP, pressure were among the first risk factors associated with
are associated with progressive VF loss. These data have incident glaucoma in population-based studies.3,5e11
important implications for our understanding of glaucoma Numerous cross-sectional studies, most with a single 24-
pathophysiology and future treatment strategies. Our study hour BP recording, have evaluated VF loss in relation to

6
 Charlson et al 
Nocturnal Blood Pressure and Glaucoma Progression

Figure 5. The percentage distribution of patients according to the time spent below the daytime mean arterial pressure (MAP).

BP using many different criteria.13,33,34 These cross- cited measured only brachial BP and not actual ocular
sectional studies have suggested that patients with VF pro- perfusion pressure.
gression had more nocturnal hypotension.12e14,35 In those Glaucoma is a progressive optic neuropathy with a va-
with VF progression, greater nocturnal hypotension was riety of phenotypes, including NTG and high-tension glau-
found in both normotensive and hypertensive patients.13,14 coma. Although they are both the same clinical entity
Randomized clinical trials also supported that systemic BP (POAG), patients with NTG have features that suggest IOP-
may be an important determinant of VF progression.26,36,37 independent mechanisms may play a significant role in its
In the Low-pressure Glaucoma Study, the use of oral anti- progression.38e41 For instance, in NTG, optic disc pheno-
hypertensive medication and lower MAP both increased the types are more often described as senile-sclerotic or focal-
risk VF progression.26 Of note, our study and the others ischemic.42,43

Figure 6. The percentage distribution of patients according to the total area below the daytime mean arterial pressure (MAP).

7
 Ophthalmology Volume -, Number -, Month 2014

To date, IOP reduction is the only proven treatment for Treatment Study: a randomized trial determines that topical
all types of open-angle glaucoma, including NTG. The ocular hypotensive medication delays or prevents the onset of
Early Manifest Glaucoma Trial, which compared treatment primary open-angle glaucoma. Arch Ophthalmol 2002;120:
with observation alone among patients with open-angle 701–13. discussion 829e30.
glaucoma (high-tension glaucoma or NTG) showed that 5. Francois J, Neetens A. The deterioration of the visual field in
glaucoma and the blood pressure. Doc Ophthalmol 1970;28:
lowering IOP decreases the risk of VF progression.44 70–132.
Focused only on NTG, the Collaborative Normal-tension 6. Drance SM. Some factors in the production of low tension
Glaucoma Study compared treated with untreated patients; glaucoma. Br J Ophthalmol 1972;56:229–42.
the intent-to-treat analysis of the Collaborative Normal- 7. Meyer JH, Brandi-Dohrn J, Funk J. Twenty four hour blood
tension Glaucoma Study, which includes all patients from pressure monitoring in normal tension glaucoma. Br J Oph-
randomization, revealed that IOP lowering did not slow thalmol 1996;80:864–7.
rates of progression.1 Nonetheless, the pragmatic analysis 8. Bonomi L, Marchini G, Marraffa M, et al. Vascular risk factors
showed that IOP lowering was beneficial, as also proven for primary open angle glaucoma: the Egna-Neumarkt Study.
in other clinical trials, but seemingly at the expense of Ophthalmology 2000;107:1287–93.
treatment-induced cataract.36 In the Low-pressure Glau- 9. Graham SL, Drance SM, Wijsman K, et al. Ambulatory blood
pressure monitoring in glaucoma. The nocturnal dip.
coma Study, an RCT in which both arms were treated with Ophthalmology 1995;102:61–9.
IOP-lowering medications, none of the IOP parameters were 10. Tielsch JM, Katz J, Sommer A, et al. Hypertension, perfusion
associated with progression.26 It is important to realize that pressure, and primary open-angle glaucoma. A population-
our study was not designed or powered to disprove the based assessment. Arch Ophthalmol 1995;113:216–21.
importance of lowering the IOP in NTG. In fact, one 11. Leske MC, Wu SY, Hennis A, et al; BESs Study Group. Risk
should be reminded that IOP lowering has been reported factors for incident open-angle glaucoma: the Barbados Eye
to improve ocular perfusion in NTG.19 Studies. Ophthalmology 2008;115:85–93.
Current guidelines for treatment of hypertensive patients 12. Kaiser HJ, Flammer J, Graf T, Stumpfig D. Systemic blood
urge aggressive targets for reducing systemic BP, especially pressure in glaucoma patients. Graefes Arch Clin Exp Oph-
in those with diabetes or renal disease.45 If such aggressive thalmol 1993;231:677–80.
13. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL.
BP goals are accompanied by nocturnal hypotension in Nocturnal arterial hypotension and its role in optic nerve head
patients with glaucoma, new, less-aggressive targets might and ocular ischemic disorders. Am J Ophthalmol 1994;117:
have to be set to prevent further VF loss. Despite the un- 603–24.
questionable benefits of treating systemic hypertension, our 14. Hayreh SS, Podhajsky P, Zimmerman MB. Role of nocturnal
data suggest that lowering BP too aggressivelyd arterial hypotension in optic nerve head ischemic disorders.
particularly at nightdmay be detrimental to peripheral tis- Ophthalmologica 1999;213:76–96.
sues such as the optic nerve, a watershed zone.46 15. Charlson ME, MacKenzie CR, Gold JP, et al. The preoperative
In conclusion, the magnitude and duration of nocturnal and intraoperative hemodynamic predictors of postoperative
hypotension identify patients with NTG who have VF myocardial infarction or ischemia in patients undergoing
progression. Ambulatory monitoring of systemic BP should noncardiac surgery. Ann Surg 1989;210:637–48.
16. Charlson ME, MacKenzie CR, Gold JP, et al. Intra-
become part of routine assessment of patients with NTG, operative blood pressure. What patterns identify patients at
particularly among those who continue to progress despite risk for postoperative complications? Ann Surg 1990;212:
IOP lowering. Nocturnal BP should be considered a modi- 567–80.
fiable risk factor in NTG. Randomized trials will be required 17. Gold JP, Charlson ME, Williams-Russo P, et al. Improvement
to assess the efficacy of different interventions designed to of outcomes after coronary artery bypass. A randomized trial
avoid nocturnal hypotension to prevent VF loss in patients comparing intraoperative high versus low mean arterial pres-
with NTG, as well as to test the effect of more aggressive sure. J Thorac Cardiovasc Surg 1995;110:1302–14.
IOP-lowering therapy in these cases. 18. Costa VP, Harris A, Anderson D, et al. Ocular perfusion
pressure in glaucoma. Acta Ophthalmol 2013 Nov 18 [Epub
ahead of print].
References 19. Ramdas WD, Wolfs RC, Hofman A, et al. Ocular perfusion
pressure and the incidence of glaucoma: real effect or artifact?
The Rotterdam Study. Invest Ophthalmol Vis Sci 2011;52:
1. Collaborative Normal-Tension Glaucoma Study Group. The 6875–81.
effectiveness of intraocular pressure reduction in the treatment 20. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method
of normal-tension glaucoma. Am J Ophthalmol 1998;126: of classifying prognostic comorbidity in longitudinal studies:
498–505. development and validation. J Chronic Dis 1987;40:373–83.
2. Musch DC, Gillespie BW, Niziol LM, et al; CIGTS Study 21. Jonas JB, Iester M. Disc hemorrhage and glaucoma [letter].
Group. Intraocular pressure control and long-term visual field Ophthalmology 1995;102:365–6.
loss in the Collaborative Initial Glaucoma Treatment Study. 22. Groppelli A, Omboni S, Parati G, Mancia G. Evaluation of
Ophthalmology 2000;118:1766–73. noninvasive blood pressure monitoring devices Spacelabs
3. Leske MC, Heijl A, Hyman L, et al; EMGT Group. Predictors 90202 and 90207 versus resting and ambulatory 24-hour intra-
of long-term progression in the Early Manifest Glaucoma arterial blood pressure. Hypertension 1992;20:227–32.
Trial. Ophthalmology 2007;114:1965–72. 23. O’Brien E, Atkins N, Mee F, O’Malley K. Comparative ac-
4. Kass MA, Heuer DK, Higginbotham EJ, et al; Ocular Hy- curacy of six ambulatory devices according to blood pressure
pertension Treatment Study Group. The Ocular Hypertension levels. J Hypertens 1993;11:673–5.

8
 Charlson et al 
Nocturnal Blood Pressure and Glaucoma Progression

24. O’Brien E, O’Malley K, Mee F, et al. Ambulatory blood patients with normal-tension glaucoma and patients with
pressure measurement in the diagnosis and management of therapeutically reduced intraocular pressures. Am J Oph-
hypertension. J Hum Hypertens 1991;5(Suppl):23–30. thalmol 1998;126:487–97.
25. Fitzke FW, Hitchings RA, Poinoosawmy D, et al. Analysis of 37. Greenfield DS, Liebmann JM, Ritch R, Krupin T; Low-Pres-
visual field progression in glaucoma. Br J Ophthalmol sure Glaucoma Study Group. Visual field and intraocular
1996;80:40–8. pressure asymmetry in the Low-pressure Glaucoma Treatment
26. De Moraes CG, Liebmann JM, Greenfield DS, et al; Low- Study. Ophthalmology 2007;114:460–5.
pressure Glaucoma Treatment Study Group. Risk factors for 38. King D, Drance SM, Douglas G, et al. Comparison of visual
visual field progression in the Low-pressure Glaucoma field defects in normal-tension glaucoma and high-tension
Treatment Study. Am J Ophthalmol 2012;154:702–11. glaucoma. Am J Ophthalmol 1986;101:204–7.
27. De Moraes CG, Liebmann CA, Susanna R, et al. Examination 39. Chauhan BC, Drance SM, Douglas GR, Johnson CA. Visual
of the performance of pointwise linear regression progression field damage in normal-tension and high-tension glaucoma.
criteria to detect glaucomatous visual field change [report Am J Ophthalmol 1989;108:636–42.
online]. Clin Experiment Ophthalmol 2012;40:e190–6. 40. Araie M, Yamagami J, Suziki Y. Visual field defects in
28. Nouri-Mahdavi K, Hoffman D, Gaasterland D, Caprioli J. normal-tension and high-tension glaucoma. Ophthalmology
Prediction of visual field progression in glaucoma. Invest 1993;100:1808–14.
Ophthalmol Vis Sci 2004;45:4346–51. 41. Caprioli J, Spaeth GL. Comparison of visual field defects in
29. Bengtsson B, Patella VM, Heijl A. Prediction of glaucomatous the low-tension glaucomas with those in the high-tension
visual field loss by extrapolation of linear trends. Arch Oph- glaucomas. Am J Ophthalmol 1984;97:730–7.
thalmol 2009;127:1610–5. 42. Yamazaki Y, Hayamizu F, Miyamoto S, et al. Optic disc
30. Zellner A. An efficient method of estimating seemingly un- findings in normal tension glaucoma. Jpn J Ophthalmol
related regressions and tests for aggregation bias. J Am Stat 1997;41:260–7.
Assoc 1962;57:348–68. 43. Ahrlich KG, De Moraes CG, Teng CC, et al. Visual field
31. Skinhoj E, Strandgaard S. Pathogenesis of hypertensive en- progression differences between normal-tension and exfolia-
cephalopathy. Lancet 1973;1:461–2. tive high-tension glaucoma. Invest Ophthalmol Vis Sci
32. Mancia G, Parati G, Omboni S, et al. Ambulatory blood 2010;51:1458–63.
pressure monitoring. Clin Exp Hypertens 1999;21:703–15. 44. Heijl A, Leske MC, Bengtsson B, et al; Early Manifest
33. Graham SL, Drance SM. Nocturnal hypotension: role in glau- Glaucoma Trial Group. Reduction of intraocular pressure and
coma progression. Surv Ophthalmol 1999;43(Suppl):S10–6. glaucoma progression: results from the Early Manifest Glau-
34. Tokunaga T, Kashiwagi K, Tsumura T, et al. Association be- coma Trial. Arch Ophthalmol 2002;120:1268–79.
tween nocturnal blood pressure reduction and progression of 45. Lenfant C, Chobanian AV, Jones DW, Roccella EJ. Seventh
visual field defect in patients with primary open-angle glaucoma report of the Joint National Committee on the Prevention,
or normal-tension glaucoma. Jpn J Ophthalmol 2004;48:380–5. Detection, Evaluation, and Treatment of High Blood Pressure
35. Collignon N, Dewe W, Guillaume S, Collignon-Brach J. (JNC 7): resetting the hypertension sails. Hypertension
Ambulatory blood pressure monitoring in glaucoma patients. 2003;41:1178–9.
The nocturnal systolic dip and its relationship with disease 46. Kario K, Shimada K. Risers and extreme-dippers of nocturnal
progression. Int Ophthalmol 1998;22:19–25. blood pressure in hypertension: antihypertensive strategy for
36. Collaborative Normal-Tension Glaucoma Study Group. nocturnal blood pressure. Clin Exp Hypertens 2004;26:
Comparison of glaucomatous progression between untreated 177–89.

Footnotes and Financial Disclosures

Originally received: December 31, 2013. Financial Disclosure(s):
Final revision: February 16, 2014. The author has no proprietary or commercial interest in any materials
Accepted: April 17, 2014. discussed in this article.
Available online: ---. Manuscript no. 2014-5. Abbreviations and Acronyms:
1
Center for Integrative Medicine, Weill Cornell Medical College, New BP ¼ blood pressure; IOP ¼ intraocular pressure; MAP ¼ mean arterial
York, New York. pressure; NTG ¼ normal-tension glaucoma; POAG ¼ primary open-angle
2
Einhorn Clinical Research Center, New York Eye and Ear Infirmary at glaucoma; PSD ¼ pattern standard deviation; RCT ¼ randomized clinical
Mount Sinai School of Medicine, New York, New York. trial; SAP ¼ standard automated perimetry; SITA ¼ Swedish Interactive
3
Department of Ophthalmology, New York University Medical Center, Thresholding Algorithm; VF ¼ visual field.
New York, New York. Correspondence:
4
Department of Statistical Science, Cornell University, Ithaca, New York. Mary E. Charlson, MD, Weill Cornell Medical College, Division of Clin-
5
Department of Ophthalmology, Weill Cornell Medical College, New ical Epidemiology and Evaluative Sciences Research, 1300 York Ave, Box
York, New York. #46, New York, NY 10065. E-mail: mecharl@med.cornell.edu.