Physical Activity and Ocular Perfusion Pressure: The EPIC-Norfolk Eye Study

Abstract

Purpose.: To examine the relationship between physical activity and ocular perfusion pressure (OPP), a consistent risk factor for glaucoma.

Methods.: The relationship between previous physical activity and current OPP in 5650 participants aged 48 to 90 who attended the first (1993–1997) and third (2006–2010) health check as part of the European Prospective Investigation into Cancer (EPIC)-Norfolk study was examined. Usual combined physical activity at work and leisure was assessed using a validated instrument. Individuals were categorized as inactive, moderately inactive, moderately active, or active. Three IOP measurements were obtained (Ocular Response Analyzer [ORA]; Reichert, Inc., Depew, NY). Mean Goldmann correlated IOP (IOPg) from one eye was used in the analysis. Systolic and diastolic blood pressure (BP) were recorded as the mean of two measurements taken with a sphygmomanometer. Associations between physical activity and low (≤40 mm Hg) mean OPP (2/3 mean arterial pressure − IOP) and low (≤50 mm Hg) diastolic OPP (diastolic BP − IOP) were tested using logistic regression, adjusting for age, sex, body mass index, social class, IOP, and BP.

Results.: Active people had a lower risk of mean OPP ≤ 40 mm Hg and diastolic OPP ≤ 50 mm Hg after adjusting for age, sex, social class, and body mass index (odds ratio, 0.75; 95% confidence interval [CI], 0.60–0.93; P < 0.01) and (odds ratio, 0.73, 95% CI, 0.58–0.93; P = 0.01), respectively. The association between physical activity and perfusion pressure was independent of IOP, but largely mediated through diastolic BP.

Conclusions.: Lower levels of physical activity were associated with lower OPP. Further research is needed to investigate the potential benefit of increased physical activity as a safe and simple method of modifying glaucoma risk.

Primary open-angle glaucoma (POAG) is a chronic optic neuropathy, with characteristic changes of the optic nerve head (ONH) and visual field. ¹ The “vascular hypothesis” suggests that abnormal perfusion of the ONH causes ischemia and poor nutrition of the retinal ganglion cell axons. ² Studies have shown an association between vascular factors and open-angle glaucoma (OAG), including systemic hypertension, ³ hypotension, ⁴ vasospasm, ^5,6 and cardiovascular disease. ^7,8 Ocular perfusion pressure (OPP) is the difference between arterial blood pressure (BP) and intraocular pressure (IOP) and is an important determinant of ocular blood flow. ⁹ Cross-sectional and longitudinal studies in different populations have shown that low OPP is a consistent risk factor for higher prevalence, ^{3,10 –14} incidence, ¹⁵ and progression ¹⁶ of OAG. The relationship between BP and OAG is less consistent, with studies reporting positive, ^7,10,11 negative, ¹⁷ and no association, ^15,18,19 indicating that OPP is a more reliable vascular risk marker for OAG. 

At present, IOP is the only modifiable risk factor in the treatment and prevention of POAG. ^20,21 Physical inactivity is an established modifiable risk factor for cardiovascular disease and hypertension. ^{22 –24} There is a consistent association between BP and IOP, ^{25 –28} which suggests that factors which can modify BP, could have a potential impact on IOP and OPP, and consequently the risk of OAG. Physical activity has been shown to reduce systolic BP (sBP) and diastolic BP (dBP) primarily in middle-aged adults. ²⁴ Because lower OPP can result from relatively low BP or relatively high IOP, it is unclear how factors that influence BP could affect ocular perfusion. 

In this study, the relationship between previous physical activity and current OPP in men and women aged 48 to 90 years was examined. 

Methods

The participants in this analysis were part of European Prospective Investigation into Cancer (EPIC)-Norfolk, a prospective population study of men and women aged 39 to 79 years, resident in Norfolk, UK. The original cohort was recruited from registers of participating general practices in Norfolk as part of a 10-country collaborative EPIC study. Detailed descriptions of the selection methodology have been reported previously. ²⁹ The study participants were 99.5% Caucasian and representative of the local population. In 2006, the EPIC-Norfolk cohort began the third questionnaire survey and clinical examination, with an additional full ophthalmic examination. The present study was based on data obtained from 5650 participants aged 48 to 89 who attended both the first (1993–1997) and third (2006–2010) health checks. All participants completed a detailed self-administered health and lifestyle questionnaire and attended a local clinic for a physical examination. The study was approved by the Norfolk Local Research Ethics Committee and adheres to the Declaration of Helsinki. All participants gave signed informed consent. 

Usual physical activity at the baseline health check (from 1993 to 1997) was assessed using two questions relating to activity in the past year. The first question asked about physical activity at work, classified into four categories: sedentary, standing (e.g., hairdresser or guard), physical work (e.g., plumber, nurse), and heavy manual work (e.g., construction worker). The second question asked about hours per week spent in other physical activity for both winter and summer (recreational activity). A simple index allocated each individual into four categories: inactive (sedentary job with no recreational activity), moderately inactive (sedentary job with <0.5 hours recreational activity per day, or standing job with no recreational activity), moderately active (sedentary job with 0.5–1 hour recreational activity, or standing job with <0.5 hours recreational activity per day, or physical job with no recreational activity), and active (sedentary job with >1 hour recreational activity per day, or standing job with >1 hour recreational activity per day, or physical job with at least some recreational activity per day, or heavy manual job). The physical activity scale used has been validated against heart rate monitoring with individual calibration in independent studies. ^30,31 In addition, the index is also inversely related to all cause mortality and incidence of cardiovascular disease in the EPIC-Norfolk cohort. ²³ In the present study, we dichotomized the population into active (as defined above) or less active (inactive, moderately inactive, or moderately active) for the univariate and multivariate analyses. This categorization was selected because initial analysis showed that active people had significantly higher perfusion pressures compared with those in other categories. Furthermore, previous studies of physical activity and glaucoma risk factors focused mainly on people undertaking strenuous exercise or with high levels of fitness, and using active people as the comparison group is more likely to reflect similar mechanisms. ³²  

Social class was recorded using the Registrar General's occupation-based classification system. Social class I are professionals, class II include managerial and technical occupations, class III is subdivided into nonmanual (Class IIInm) and manual skilled (Class IIIm) worker, class IV consists of partly skilled workers, and class V are unskilled manual workers. Usual medication use was recorded from the participant's most recent prescription. All participants with antiglaucoma medications were excluded from the analysis to examine potential associations with untreated IOP and perfusion pressure. 

A health examination was carried out by trained nurses in all health checks, following detailed standard operating procedures. All clinical measurements examined in this study were taken from the third health check from 2006 to 2010. Systolic and diastolic blood pressure was recorded as the mean of two measurements taken from the right arm with the participant seated for 5 minutes, using a blood pressure monitor (Accutorr Plus; Mindray, Huntingdon, UK). Height and weight were measured with participants dressed in light clothing and shoes removed. A stadiometer was used to measure height to the nearest 0.1 m, and a body composition analyzer (Tanita model TBF 300s; Chasmors Ltd., London, UK) was used to measure weight to the nearest 100 g. Body mass index (BMI) was then calculated using the formula: weight (kg)/height (m²). Hip and waist circumference were measured with participants in light clothing. 

Intraocular pressure was measured three times in each eye using a noncontact ocular response analyzer (Ocular Response Analyzer [ORA]; Reichert, Inc., Depew, NY) to generate measures equivalent to Goldmann applanation tonometry, the recognized reference standard in clinical practice. Mean OPP was calculated from 2/3 mean arterial pressure minus IOP, where mean arterial pressure = dBP + 1/3(sBP − dBP). Systolic perfusion pressure (SPP) and diastolic perfusion pressure (DPP) were also calculated by subtracting IOP from sBP and dBP respectively. Only measurements from right eyes were used. 

Study population characteristics were examined using two-sample t-tests for differences in means and for differences in percentages, between men and women. An initial exploratory analysis between physical activity levels and OPP as a continuous variable showed an association with mean OPP and DPP, but not SPP. Therefore, to relate OPP levels to potential OAG risk, mean OPP and DPP were dichotomized using cutoff points that have been identified with increased prevalence of OAG from the literature. The dichotomization resulted in categorization of low mean OPP and DPP as ≤40 and ≤50 mm Hg respectively. ^3,15,33 In addition, risk factor distribution by physical activity and by OPP category was examined. Univariate associations between both physical activity and OPP with potential confounders were explored using linear regression and a χ² test. A stepwise logistic regression model was used to examine the risk of low mean OPP, DPP, and SPP in previously active people compared with other categories combined, adjusting for age, sex, social class, education, BMI, systolic and diastolic blood pressure, and IOP. Indicator variables were used with all categorical variables in the multivariate analysis. The potential effect modification by age, sex, BMI, social class, systolic and diastolic BP, IOP, and BP-lowering and/or vasoactive (including anti-anginal and vasodilators) medication in the relationship between physical activity and mean OPP and DPP was investigated by using interaction terms. All statistical analyses were conducted using statistical software (STATA 10; Statacorp, College Station, TX). 

Results

The demographic characteristics of the 5650 men and women in this study are shown in Table 1. Men were, on average, older than women. They also had higher BMI, systolic, and diastolic blood pressure. Intraocular pressure was similar in both men and women with a mean of 16.0 mm Hg for the complete cohort. There were higher proportions of men who were classified as professional and managerial, as well as skilled manual workers. Overall, men were more active than women. The mean OPP, SPP, and DPP for this cohort was 49.1 mm Hg (95% confidence interval [CI], 48.90–49.3), 120.0 mm Hg (95% CI, 119.5–120.4), and 62.6 mm Hg (95% CI, 62.3–62.8), respectively. 

Table 1.

 

Sample Characteristics

Variable	Men (n = 2512)	Women (n = 3138)	P *
Age	68.8 ± 0.2	67.4 ± 0.1	<0.01
Body mass index	27.2 ± 0.1	26.6 ± 0.1	<0.01
Systolic blood pressure, mm Hg	136.5 ± 0.3	135.4 ± 0.3	<0.01
Diastolic blood pressure, mm Hg	79.9 ± 0.2	77.3 ± 0.2	<0.01
Intraocular pressure, mm Hg	15.9 ± 0.1	16.0 ± 0.1	0.08
Social class			<0.01
Professional	9.4 (237)	8.4 (262)
Manager	42.4 (1065)	39.4 (1235)
Skilled nonmanual	12.5 (313)	19.4 (610)
Skilled manual	23.2 (583)	18.8 (590)
Semiskilled	10.7 (269)	11.4 (358)
Unskilled	1.8 (45)	2.6 (83)
Previous physical activity levels			<0.01
Less active	73.3 (1841)	79.7 (2502)
Active	26.7 (671)	20.3 (636)

 Data are presented as mean ± SE or % (n).

*   P-value from two-sample t-test for continuous variables and χ² test for categorical variables.

The distribution of risk factor variables by previous physical activity levels are shown in Table 2. People who were active were younger and had higher dBP, and higher mean OPP and DPP. The more active individuals were also more likely to have manual employment and less likely to be prescribed antihypertensive or vasoactive medication. There was no association between physical activity and the variables BMI, sBP, IOP, and SPP. As SPP levels did not differ between different levels of physical activity, the remainder of the analysis was focused on mean OPP and DPP only. The association between risk factor variables and mean OPP and DPP categories is shown in Table 3. People with low mean OPP were older, more likely to be women, with lower BMI, and sBP, and higher dBP. Similar associations were observed with DPP, except people with low DPP had lower dBP. Both low mean OPP and DPP was associated with higher mean IOP of approximately 3 mm Hg, consistent with a higher risk of OAG. There was no association between either low mean OPP or DPP with social class. A greater proportion of people with low DPP were prescribed antihypertensive or vasoactive medication. People taking these medicines were also more likely to have higher sBP, but lower dBP because systolic hypertension is more common in older people. ³⁴  

Table 2.

 

Distribution of Variables by Physical Activity

	Baseline Physical Activity Levels
	Less Active (n = 4343)	Active (n = 1307)	P *
Age	68.3 ± 0.1	67.0 ± 0.2	<0.01
Body mass index	26.9 ± 0.1	26.7 ± 0.1	0.1
Systolic blood pressure, mm Hg	135.9 ± 0.3	136.0 ± 0.4	0.8
Diastolic blood pressure, mm Hg	78.2 ± 0.1	79.4 ± 0.3	<0.01
Intraocular pressure, mm Hg	16.0 ± 0.1	15.9 ± 0.1	0.43
Mean ocular perfusion pressure, mm Hg	49.0 ± 0.1	49.6 ± 0.2	<0.01
Systolic ocular perfusion pressure, mmHg	120.0 ± 0.2	120.0 ± 0.4	0.75
Diastolic ocular perfusion pressure, mmHg	62.3 ± 0.1	63.6 ± 0.3	<0.01
Social class†
Professional	9.6 (417)	6.3 (82)	<0.01
Manager	41.7 (1811)	37.4 (489)
Skilled nonmanual	17.8 (774)	11.4 (149)
Skilled manual	18.6 (809)	27.9 (364)
Semiskilled	10.3 (447)	13.8 (180)
Unskilled	2.0 (85)	3.3 (43)
Current Medication‡	38.0 (1648)	32.0 (418)	<0.01

  N = 5650. Data are presented as mean ± SE or % (n).

*   P-value from simple linear regression for linear trends in continuous variables and χ² test for categorical variables.

†  There are some missing values.

‡  Antihypertensive or vasoactive medication.

Table 3.

 

Association between Risk Factor Variables and Ocular Perfusion Pressure

	Mean OPP			Diastolic OPP
	≤40 mm Hg (n = 594)	>40 mm Hg (n = 5056)	P *	≤50 mm Hg (n = 499)	>50 mm Hg (n = 5151)	P *
Age	68.1 ± 0.1	67.3 ± 0.3	0.02	69.8 ± 0.4	67.8 ± 0.1	<0.01
Body mass index	25.7 ± 0.2	27.0 ± 0.1	<0.01	26.3 ± 0.2	26.9 ± 0.1	0.02
Systolic blood pressure, mm Hg	114.6 ± 0.6	138.4 ± 0.2	<0.01	120.7 ± 0.8	137.4 ± 0.2	<0.01
Diastolic blood pressure, mm Hg	79.9 ± 0.1	66.3 ± 0.3	<0.01	64.1 ± 0.3	79.9 ± 0.1	<0.01
Intraocular pressure, mm Hg	18.6 ± 0.2	15.7 ± 0.1	<0.01	18.4 ± 0.2	15.7 ± 0.1	<0.01
Sex			<0.01			<0.01
Men	8.6 (215)	12.1 (379)		7.3 (183)	10.1 (316)
Women	91.4 (2297)	87.9 (2759)		92.7 (2329)	89.9 (2822)
Social class†			0.2			0.07
Professional	9.4 (56)	8.8 (443)		8.8 (44)	8.8 (455)
Manager	43.3 (257)	40.4 (2043)		40.3 (201)	40.8 (2099)
Skilled nonmanual	17.0 (101)	16.3 (822)		18.8 (94)	16.1 (829)
Skilled manual	16.7 (99)	21.2 (1074)		16.6 (83)	21.2 (1090)
Semiskilled	70 (11.8)	11.0 (557)		13.6 (68)	10.9 (559)
Unskilled	1.9 (11)	2.3 (117)		1.8 (9)	2.3 (119)
Current Medication‡	31.7 (188)	37.1 (1878)	<0.01	42.1 (210)	36.0 (1856)	<0.01

  N = 5885. Data are presented as mean ± SE or % (n).

*   P-value from two-sample t-test for continuous variables and χ² test for categorical variables.

†  There were some missing values.

‡  Antihypertensive or vasoactive medication.

In a multivariate logistic regression model active people at baseline had 25% lower risk of low mean OPP after adjusting for age, sex, social class, and BMI compared with all other categories (odds ratio [OR], 0.75; 95% CI, 0.60–0.93; P < 0.01; Table 4). The effect of physical activity on OPP was independent of IOP, but largely dependent on dBP. Similarly, the risk of low (≤50 mm Hg) DPP was 27% lower (OR, 0.73; 95% CI, 0.58–0.93; P = 0.01) in active people. As with OPP, the effect of physical activity on DPP was independent of age, sex, social class, BMI, IOP, and sBP but was largely mediated through dBP. 

Table 4.

 

Multivariate Adjusted Relative Risk of Low Ocular Perfusion Pressure for People Who Were Active Compared to Not Active at Baseline

Age, Sex, and Social Class Adjusted		Age, Sex, Social Class, and BMI Adjusted		Age, Sex, Social Class, BMI, and IOP Adjusted		Age, Sex, Social Class, BMI, and Systolic BP Adjusted		Age, Sex, Social Class, BMI, Systolic, and Diastolic BP Adjusted
OR (95% CI)	P *	OR (95% CI)	P *	OR (95% CI)	P *	OR (95% CI)	P *	OR (95% CI)	P *
Risk of Low Mean OPP <40 mm Hg in Active People at Baseline
0.76 (0.61–0.95)	0.02	0.7 (0.60–0.93)	<0.01	0.73 (0.58–0.92)	<0.01	0.76 (0.59–0.99)	0.04	0.82 (0.62–1.09)	0.2
Risk of Low DPP <50 mm Hg in Active People at Baseline
0.74 (0.58–0.94)	0.01	0.73 (0.58–0.93)	0.01	0.73 (0.57–0.94)	0.01	0.75 (0.58–0.97)	0.03	0.94 (0.68–1.31)	0.7

*   P-value from Wald test for association.

†  Compared with all other levels of physical activity at baseline.

There was no evidence of an interaction between physical activity and the examined risk factors, except for medications in the model for DPP (likelihood ratio test for presence of interaction = 0.02). The relative risk of low DPP stratified by current medication is shown in Table 5. A higher level of previous physical activity was associated with low DPP, but only in people who were currently prescribed antihypertensive or vasoactive medication. Inclusion of medications in the multivariate model for OPP did not significantly alter the observed estimates shown in Table 4. 

Table 5.

 

Relative Risk of Low Diastolic Ocular Perfusion Pressure for People Who Were Active Compared with Not Active at Baseline Stratified by Current Medication

Age, Sex, and Social Class Adjusted		Age, Sex, Social Class, and BMI Adjusted		Age, Sex, Social Class, BMI, and IOP Adjusted		Age, Sex, Social Class, BMI, and Systolic BP Adjusted		Age, Sex, Social Class, BMI, Systolic, and Diastolic BP Adjusted
OR (95% CI)	P *	OR (95% CI)	P *	OR (95% CI)	P *	OR (95% CI)	P *	OR (95% CI)	P *
Risk of Low DPP <50 mm Hg in Active People at Baseline† with Current Medication ‡
0.52 (0.34–0.80)	<0.01	0.52 (0.34–0.81)	<0.01	0.49 (0.31–0.77)	<0.01	0.52 (0.33–0.83)	<0.01	0.58 (0.32–1.05)	0.07
Risk of Low DPP <50 mm Hg in Active People at Baseline† with No Current Medication ‡
0.90 (0.67–1.20)	0.46	0.90 (0.67–1.21)	0.48	0.91 (0.67–1.23)	0.54	0.93 (0.68–1.27)	0.64	1.19 (0.80–1.77)	0.40

*   P-value from Wald test for association.

†  Compared with all other levels of physical activity at baseline.

‡  Antihypertensive or vasoactive medication.

Discussion

In the present large population study of 5650 middle-aged and older Caucasian men and women, a high level of habitual physical activity at baseline was associated with a reduced risk of low OPP and DPP 10 years later. Using meaningful cutoff points of mean OPP and DPP that are associated with increased risk of OAG, ^3,12 it was shown that active people had a 25% lower risk of low mean OPP and DPP compared with people with lower levels of habitual physical activity. The findings of the present study indicate a potential protective effect of physical activity on the risk for developing OAG. The association between physical activity and OAG risk was independent of age, sex, social class, BMI, IOP, and systolic BP, but largely mediated through dBP. 

This is the first time, to our knowledge, that the relationship between physical activity and OPP has been investigated. However, there have been a large number of studies that have examined the effect of physical activity on BP and IOP, the two components of OPP. A large number of randomized controlled trials have shown that physical activity reduces resting sBP, with an average reduction of 3.4 mm Hg. ²⁴ This in turn translates into a protective effect on the risk of coronary heart disease, ²² for which high BP is a major risk factor. The effect of physical activity on IOP has recently been reviewed by Risner et al. ³² Most studies have reported a reduction in IOP immediately or shortly after exercise, furthermore, people who have higher levels of fitness have also been shown to have lower IOP. ³⁵ Although these studies were primarily based on small numbers of people undertaking higher levels of exercise, there is a strong suggestion that physical activity might have a protective effect on the risk of OAG through IOP. In the present study, active people had slightly lower IOP at follow-up, although this was not statistically significant. 

The evidence for OPP as a risk factor for OAG is strong and more consistent than that for BP. Cross-sectional studies from the United States, ^3,11,13 Italy, ¹⁰ The Netherlands, ¹⁴ Barbados, ¹⁵ Sweden, ¹⁹ and Singapore ¹² from populations of White, Black, Hispanic, and Malay origin all demonstrate an association between low OPP and prevalence of OAG. In the Baltimore Eye Study, there was a clear threshold effect between DPP at 50 mm Hg and OAG, with no association at a higher DPP with OAG. ³ Similar threshold effects have been observed in other studies. ^10,12,13 In the 9-year follow-up of the Barbados Eye Study cohort, Leske et al. ¹⁵ showed that participants with low mean OPP (<40 mm Hg) at baseline had nearly three times the risk of incident OAG compared with those with higher baseline mean OPP. Participants with low DPP (<53 mm Hg) also had a higher risk of OAG. Similar associations were also observed in the 4-year follow-up Barbados Eye Study. ³³ Based on the findings of the Baltimore and Barbados studies, OPP <40 mm Hg and DPP <50 mm Hg levels were used as cutoffs to identify levels of low perfusion pressure in the present study. Although the Baltimore and Barbados studies were performed in different populations using different methodologies, use of these OPP and DPP levels provided some indication for potential risk of OAG. 

Low perfusion pressure can be due to relatively low BP or relatively high IOP, but there is no conclusive evidence to show that either variable alone is solely responsible for low ocular perfusion in OAG. In the present study, active people had lower IOP, higher dBP, and significantly higher DPP. It was found that the association of physical activity was independent of IOP, but largely attenuated with the addition of dBP in the multivariate model, suggesting that the effect was mediated by the higher dBP in active people. Diastolic BP decreases in older people, and pulse amplitude widens, ³⁶ reflecting a decrease in arterial compliance. ³⁷ Higher levels of physical activity are associated with lower risk of arterial stiffness. ³⁸ Therefore, it was of interest that for the present study population, active people were found to have narrower pulse amplitude (data not presented). This suggests that the beneficial effects of physical activity on ocular perfusion may be due, in part, to differences in arterial compliance. Furthermore, wider pulse amplitude has been shown to be associated with a higher prevalence of OAG. ¹⁴  

In the stratified analysis, a higher level of physical activity was associated with low DPP, but only in those who were currently prescribed antihypertensive or vasoactive medication. One possible explanation is that use of these medicines could be a proxy indicator for poor cardiovascular health, and the protective effect of physical activity on tissue perfusion was magnified due to greater relative differences between people who were active and had minor indications for medication compared with less active people who may have had multiple pathologies. Whereas in those without BP-lowering or vasoactive treatment, there was less difference in perfusion between active and inactive groups as both groups have relatively good cardiovascular health. 

Ocular perfusion pressure is an important determinant of ocular blood flow, ^9,39 which in turn is maintained by autoregulatory mechanisms that allow adaptation to metabolic demands. Vascular dysregulation and unstable ocular blood flow has been linked to OAG and OAG progression. ⁴⁰ In particular, primary vascular dysregulation (PVD) has a major effect on autoregulation and is a risk factor for normal tension glaucoma (NTG). ⁴⁰ PVD is more likely to occur in women, similarly, there are more women than men among NTG patients, and untreated women have a more rapid visual field progression compared with untreated men. ⁴¹ Pasquale and co-workers ⁴² also showed a protective effect from higher BMI in women with incident NTG (IOP ≤ 21 mm Hg at diagnosis). People with PVD tend to have normal or low BMI, ⁴³ and the reported protective effect of higher BMI may be associated with PVD. In the present study, women and people with lower BMI had low perfusion pressure and this is consistent with a vascular dysregulation mechanism. 

The threshold effect of DPP and OAG observed in epidemiologic studies support findings from laboratory studies that demonstrate a threshold effect in the vascular autoregulatory response to IOP increases. ^44,45 Therefore, with higher IOP and dysregulation in older people, a higher dBP may protect the ONH from poor perfusion and unstable blood flow. Physical activity has a protective effect on other disorders associated with dysregulation in older people, such as frailty syndrome, ⁴⁶ and similar physiologic pathways may be evident in OAG. 

The observed differences in perfusion pressure between active and less active people are small; active people had 1.2 mm Hg higher mean DPP than those who were less active. Although this may appear clinically insignificant on an individual level, small shifts in population mean values can have a significant impact on disease frequency. In randomized controlled trials, a 5 to 6 mm Hg reduction in dBP is associated with a reduced risk of cardiovascular disease, ⁴⁷ similarly, the Early Manifest Glaucoma Treatment study showed a 10% to 13% increased risk in glaucoma progression per mm Hg higher IOP. ¹⁶ Furthermore, in the Advanced Glaucoma Intervention Study, the difference in mean IOP between patients with progressive disease and those with stable disease was 2.4 mm Hg. ⁴⁸  

The present study has several limitations. There are potential measurement errors in the assessment of physical activity and OPP. The physical activity scale has been validated with heart rate, ³¹ and previous studies using this questionnaire have shown increased risk of cardiovascular mortality with lower levels of physical activity, ⁴⁹ suggesting that the instrument is reliable and accurate. A noncontact tonometer was used in the present study, which is less accurate than contact tonometers such as Goldmann applanation tonometry. ⁵⁰ Furthermore, BP was taken at one point in time, and may not be an accurate assessment of the individual's diurnal BP. The potential for nonsystematic measurement error in physical activity and/or perfusion pressure would attenuate any association detected. Therefore finding a relationship between physical activity and ocular perfusion pressure can give us insights into possible lifestyle and biological determinants which are not well understood. Measurements were also taken during daytime hours, and the potential effect of nocturnal dips in BP and ocular perfusion cannot be assessed in this study. Glaucoma patients have a greater reduction in nocturnal DPP compared with controls, ⁵¹ therefore participants at risk of glaucoma in this study may have even lower nocturnal DPP than that observed in our daytime measurements. Secondly, OPP was estimated using brachial BP measurements, which is higher than ocular arterial pressure due to the hydrostatic column effect. However, clinical and epidemiologic studies that demonstrate an association between OPP and OAG have all used this surrogate measure, suggesting that although the calculation may not accurately reflect true OPP, it consistently reflected the risk of OAG. Use of the uncorrected formula to calculate DPP also allowed us to use cutoffs from other studies to indicate glaucoma risk. Additional analysis using the corrected formula (2/3 dBP − IOP) indicated similar associations with previous physical activity. However, the correlation between these surrogate measures of OPP and actual perfusion of the optic nerve head remains to be determined, and direct measurement of OPP will improve accuracy of risk estimation. Thirdly, physical activity was assessed at only one point in time, and changes in lifestyle were not taken in to account. Nonetheless, there is little evidence for the impact of other lifestyle factors on OAG or IOP, and any residual confounding factor is unlikely to alter the observed results. 

A recent perspective from the Blood Flow in Glaucoma Discussion Group concluded that it is unlikely that glaucoma treatments based on altering optic nerve perfusion will soon be available. ⁵² The unlikely introduction of such a novel therapy in the near future is primarily due to adverse cardiovascular implications of treatments that could alter ocular perfusion by modification of BP, particularly in older people. In the present study, it has been shown that higher levels of physical activity appeared to have a long-term beneficial impact on OPP, and potentially reduce the frequency of low OPP in older people and reduce risk of OAG. The statistical association detected here does not necessarily indicate a direct causal relationship, due to underlying complexities in the vascular mechanisms. Furthermore, the relationship between habitual physical activity and OAG has yet to be determined. 

Conclusions

People with a previous active lifestyle have a lower risk of having low DPP and/or OPP, an association that appears to be partly mediated by dBP. Physical activity may offer a safe and simple method for reducing the risk of developing OAG. However, the role of lifestyle modification in the prevention of OAG requires further investigation. 

Footnotes

 Supported by Research into Ageing (London), Medical Research Council (London), National Institute of Health Research (London), National Institute for Health Research (UK) Biomedical Research Centre at Moorfields Eye Hospital, the UCL Institute of Ophthalmology (DFG-H, PJF), and The Richard Desmond Charitable Trust (via Fight for Sight) (PJF).

Footnotes

 Disclosure: J.L.Y. Yip, None; D.C. Broadway, None; R. Luben, None; D.F. Garway-Heath, None; S. Hayat, None; N. Dalzell, None; P.S. Lee, None; A. Bhaniani, None; N.J. Wareham, None; K.-T. Khaw, None; P.J. Foster, None