Modifiable Lifestyle and Environmental Risk Factors Affecting the Retinal Microcirculation



    1. Faculty of Health Sciences and Medicine, Bond University, Robina, Queensland, Australia
    Search for more papers by this author

    1. Centre for Eye Research Australia, University of Melbourne, Royal Victorian Eye and Ear Hospital, Melbourne, Australia
    2. Department of Ophthalmology, Faculty of Medicine, Gadjah Mada University, Yogyakarta, Indonesia
    Search for more papers by this author

Address for correspondence: Kevin R. Serre, Faculty of Health Sciences and Medicine, Bond University, Robina, Qld 4229, Australia.


Please cite this paper as: Serre and Sasongko (2012). Modifiable Lifestyle and Environmental Risk Factors Affecting the Retinal Microcirculation. Microcirculation 19(1), 29–36.


Structural changes within the human retinal vasculature may reflect systemic vascular changes associated with various cardiovascular and metabolic disorders. Recent data suggest that systemic exposure from a range of modifiable lifestyle and environmental risk factors (e.g., diet, physical activity, and smoking) may affect the morphology of the retinal vasculature. Being easily accessible and non-invasively visualized, the retinal microvasculature therefore can be a clinically useful biomarker of reversible sub-clinical physiologic deviation of the systemic circulation as results of such unfavorable exposures. Importantly, quantitative analysis of the retinal microvasculature may be utilized as a prognostic tool, allowing for targeted vascular therapies before the onset of overt cardiovascular and metabolic disorders. This review summarizes the modifiable lifestyle and environmental risk factors that affect retinal microvascular structure and the possible clinical implications of such relationships.

Abbreviations used:

angiotensin-converting enzyme inhibitor


angiotensin-receptor blocker


atherosclerosis risk in communities


arteriole-to-venule ratio


Beaver Dam Eye Study


Blue Mountain Eye Study


body mass index


central retinal artery equivalent


central retinal venular equivalent


fractal dimension


glycemic index


Multi-Ethnic Study of Atherosclerosis




The retinal microcirculation may reflect healthy and pathophysiologic processes affecting systemic circulation [64]. The vascular architecture within the retina, as well as elsewhere in the body, is thought to follow the principles of optimality, which allows the blood distribution to peripheral tissue within the quickest time with the least amount of energy [45,65]. Therefore, deviations from optimal structure of the retinal vasculature (e.g., arteriolar narrowing, venular widening) may represent deviation of the circulation from its optimal state, indicating any pathophysiologic processes.

During the last few decades, the retinal vasculature has received increasing attention. With the advancement of retinal imaging, the retinal vasculature may allow non-invasive visualization to examine and monitor human circulation systems in vivo (Figure 1). For example, computer-based analysis techniques from digital retinal images has allowed accurate and reproducible measurement of several parameters of the retinal vasculature (e.g., caliber, fractal dimension [complexity of vessel network], and branching angle) [6,11,41,61,62]. A number of large-scale epidemiological studies have demonstrated that subtle changes in these parameters carry important information regarding the future risk of systemic vascular diseases [18,25,30,39,40,50,58,60,62].

Figure 1.

 (A) Examples of retinal vascular images with different morphologies, showing, from left to right, decreasing arteriolar caliber and vessel density. (B) Measurement of retinal arteriolar and venular caliber.

Importantly, changes in the retinal vasculature have also been shown to have strong associations with systemic and environmental cardiovascular risk factors in a range of populations (for review see Ref. [51]), even before the clinical manifestation of diseases. These subtle retinal vascular changes have been suggested to mirror preclinical changes in both the cerebral [32] and coronary [53] microcirculations. Although the mechanisms remain questionable, this may indicate that abnormalities in the retinal vasculature incorporate a cumulative effect of systemic damage.

Recently, many of the largest determinants of this sub-optimal retinal microvasculature have been found to be modifiable [40], such as diet and medications. More importantly, information regarding these modifiable exposures, together with measures of the retinal vasculature may eventually provide clinically useful prognostic information regarding systemic disease risk prediction beyond current traditional risk factor assessment. In this review, we summarize recent research examining the modifiable lifestyle and environmental determinants affecting the retinal microvasculature (Table 1) and potential clinical implications of these findings.

Table 1.   Summary of recent studies examining the modifiable lifestyle and environmental determinants affecting the retinal microvasculature
Risk factorsAuthorsYear of publicationPopulationMeasured exposureObserved outcome
DietKan et al. [20]200710,659 US adults (ARIC)Increased dietary fiberWider retinal arteriolar and narrower venular caliber
DietKaushik et al. [21]20083654 adults aged 49+ years (BMES)Increased Fish consumptionWider retinal arteriolar and narrower venular caliber
DietKaushik et al. [22]20093654 adults aged 49+ years (BMES)High Glycemic IndexWider retinal venular caliber
DietLim et al. [42]2009823 Singapore School Children 12.8 ± 0.8 yearsHigh Glycemic IndexNo significant association
Physical activityGopinath et al. [15]20112238 six-year-old Sydney studentsIncreased sporting activitiesWider retinal arteriolar caliber
 Increased TV viewing timeNarrower retinal arteriolar caliber
Physical activityAnuradha et al. [4]20115893 US adults from 4 racial/ethnic groups (MESA)Low self-reported physical activityWider retinal venular caliber
 Increased self-reported TV viewing timeWider retinal venular caliber
Physical activityAnuradha et al. [3]20102024 Australian adults (AusDiab)Increased self-reported TV viewing timeWider retinal venular caliber
 Low self-reported physical activityNo significant association
Physical activityTikellis et al. [57]201015,792 US adults 45–64 years (ARIC)Higher self-reported physical activityWider retinal venular caliber
Cardiovascular fitnessHanssen et al. [16]201146 adult recreational runnersHigher cardiovascular fitness (anaerobic threshold)Wider retinal arteriolar caliber and higher AVR
 10-week exercise programRetinal arteriolar dilatation in obese individuals
Cardiovascular fitnessK Serre, MJ Simmons, S Sabapathy, CL Minahan, M Steele, R Kawasaki, JJ Wang, TY Wong, P Mitchell, GC Gassunpublished observations15 women 65–74 years with type 2 diabetes12-week exercise programNo significant changes
Cardiovascular fitnessK Serre, MJ Simmons, S Sabapathy, CL Minahan, M Steele, R Kawasaki, JJ Wang, TY Wong, P Mitchell, GC Gassunpublished observations15 women 65–74 years with type 2 diabetesIncreased time to exhaustion during graded exercise testingIncreased fractal dimension (Df)
ObesityTaylor et al. [54]20071740 Sydney school children (SCES)Increased BMI and weightWider retinal venular caliber
 Increase BMI and waist circumferenceNarrower retinal arteriolar caliber
ObesityCheung et al. [12]2007768 Singapore school children (SCORM)Increased BMI and weightWider retinal venular caliber
ObesityGopinath et al. [14]20103144 Sydney school childrenIncreased BMINarrower retinal arteriolar and wider venular caliber
ObesityLi et al. [37]2011136 Singapore school children (STARS)Increased triceps skinfold and BMIWider retinal venular caliber
Cigarette smokingLiew et al. [40]200815,792 US adults 45–64 years (ARIC)Past cigarette smokingWider retinal venular caliber
Cigarette smokingRosenburg [48]2005Clinical observationPast and current cigarette smokingClinical retinal venular widening
Cigarette smokingJeganathan [19]2005Clinical observationPast and current cigarette smokingClinical retinal venular widening
MedicationTom et al. [55]200919,342 adults 40–79 years undergoing antihypertensive therapy (ASCOT)Calcium channel blocker vs. β-blocker therapyLess retinal arteriole narrowing with calcium channel blocker therapy
MedicationKlein et al. [29]2010147 normotensive adults with type 1 diabetesACEI and ARB therapyNo significant changes
Air pollutionAdar et al. [1]20106814 adults 45–84 years (MESA)Higher long- and short-term PM2.5 exposureNarrower retinal arteriolar caliber

The Influence of Modifiable Lifestyle and Environmental Risk Factors on the Retinal Microvasculature


Dietary fiber intake, regular fish consumption, and low GI diets, such as those high in sugars and simple carbohydrates, are all associated with reduced risk of vascular disease [8,24,31]. Emerging data suggests that the relationship between diet and macrovascular disease may partly be mediated by associated changes in the microcirculation [20–22]. Recent work has shown that diet may have effects on retinal vascular caliber in the general population. For example, data from the ARIC study showed that higher intake of dietary fiber was independently associated with wider retinal arteriolar caliber and narrower venular caliber, indicating a lower risk of cardiovascular diseases [20]. Similarly, findings from the BMES also demonstrated beneficial effects of increasing frequency of fish consumption on retinal microvasculature independent of other cardiovascular risk factors [21].

On the contrary, high-GI diets have been linked to deleterious anatomic changes in the retinal microvasculature [21,22]. Kaushik et al. [22] suggest that high-GI diets were associated with wider retinal venules and greater stroke mortality in persons 50 years and older. This suggests that postprandial glucose may have deleterious effects on the cerebral microcirculation and may play a significant role in the relationship between diet and stroke mortality. More recent data from 823 schoolchildren aged 12.8 (±0.8) years [42] demonstrated that there was no association between a high-GI diet and retinal arteriolar or venular caliber. This evidence suggests a possible dose-dependent, cumulative effect of diet on the microvasculature over time.

The physiologic influence of diet on the retinal microcirculation is probably complex. Kan et al. [20] found that the effect of fiber intake on retinal microvascular caliber might be confounded by current hypertension and dyslipidemia. This suggests that the beneficial retinal microvascular changes seen with increased fiber intake may not be directly affected by fiber intake itself, but by associated decreases in adverse systemic conditions like hypertension and dyslipidemia. For example, fish consumption is associated with increases in HDL [5]. Increased concentration of HDL has a well-established vaso-protective and anti-atherogenic effect [44] and may alone explain the beneficial retinal microvascular changes associated with higher fish consumption. Findings demonstrating that the microvascular effects of diet were not evident in children free of systemic disease [42] support this theory. Also, regular fish consumption, fiber intake, and low GI diets are all known to reduce systemic inflammation [47,56]. Retinal microvascular changes are known to be affected by inflammatory factors [26], and may be another biologic mechanism through which diet mediates microvascular caliber. Although the mechanisms underlying the above associations may not be completely understood, this data supports the vascular-protective effects of increased dietary fish, fiber, and low GI food consumption.

Physical Activity and Exercise

Sedentary behavior, low levels of physical activity, and low cardiorespiratory fitness are all well-established risk factors for atherosclerosis and CVD [34]. Recent research has also shown that the adverse effects of lack of physical activity and low fitness extends to changes in microvascular structure [3,4,15,16,55]. Sedentary behavior, indicated by time spent watching TV and lower levels of physical activity, assessed via self-report, were found to be associated with retinal venular caliber [3,4,55], suggesting a possible deleterious effect of decreased levels of physical activity and increased sedentary behavior on the microvasculature. In addition, the impact of physical activity on the retinal microvasculature was also observed in a cohort of 6-year children. In the study by Gopinath et al., children who spent more time in outdoor sporting activities had wider mean retinal arteriolar caliber [15], but those who spent more time watching TV had narrower mean retinal arteriolar caliber. More importantly, for each hour of daily television viewing time, similar retinal arteriolar changes are associated with a 10 mmHg increase in systolic blood pressure [15].

Recently, there is also evidence showing the relationship between higher levels of physical fitness and retinal microvascular structure [16]. Higher cardiovascular fitness, as assessed by individual anaerobic threshold, was found to be related to retinal arteriolar dilation and higher retinal AVR [16]. Moreover, 10 weeks of exercise training was also shown to induce arteriolar dilatation in obese individuals and increased AVR in both obese and lean individuals [16]. Conflicting results were found in a study of older women with type 2 diabetes in which no training-induced improvements in retinal vessel caliber were found after 12-weeks of moderate-intensity exercise. In this cohort, however, increased retinal microvascular density, shown by increased Df was associated with increased time to exhaustion during peak exercise testing, a measure of physical fitness.

Observed associations between physical activity and changes in the retinal microvasculature may provide in vivo evidence regarding the effect of physical activity on the systemic circulation. Although the exact pathophysiologic mechanisms behind these relationships is not know, recent research suggests that moderators of vascular tone, specifically NO and ADMA, may play a significant role. ADMA, a nitric oxide synthase inhibitor, has been shown to directly affect retinal arteriole and venule diameter [17]. More recently, Hanssen et al. [16] found that exercise training-induced increases in arteriolar caliber were accompanied by significant decreases in ADMA, suggesting that the NO/ADMA pathway may play a key role in the beneficial changes in microvascular structure associated with regular exercise.


The effect of obesity on the retinal microcirculation has been well established. Arteriolar caliber narrowing, venular caliber widening and lower AVR have been found to be associated with obesity in both children and adult populations [18,27,28,57,59,60], suggesting that obesity may cause deleterious microvascular changes before clinical signs and symptoms of vascular disease are present. In children, greater BMI was associated with wider retinal venular caliber and narrower arterioles, weight and body surface area were associated with wider retinal venules only, and larger waist circumference was associated with narrower retinal arterioles [52]. In the SCORM [12], greater BMI and weight were associated with wider retinal venular caliber. Consistent with this evidence, more recent studies also demonstrated that BMI and triceps skinfold [14,37] were found to be associated with wider retinal venular caliber and narrower retinal arteriolar caliber in healthy, pre-adolescent children, supporting an early adverse effect of obesity on microvascular structure.

Although the mechanisms underlying the association between obesity and retinal vessel diameter are unclear, several possible explanations exist. Systemic inflammation is thought to contribute to the vascular complications associated with obesity [7]. Systemic inflammation is also associated with changes in retinal venular caliber [26], and therefore may be the mechanism through which obesity affects retinal microvascular structure. Obesity is also related to increased total blood volume [46], and retinal venular dilatation may be a regulatory response to maintain blood flow. These relationships between obesity and retinal microvascular changes may help explain the association between childhood obesity and complications such as hypertension, diabetes, and cardiovascular morbidity and mortality that occur later in life [13].

Cigarette Smoking

The Rotterdam Study [18], BDES [26], MESA [60], Wisconsin Epidemiologic Study of Diabetic Retinopathy [28], and BMES [23] have all demonstrated a consistent association between wider retinal vessel caliber and cigarette smoking, suggesting that adverse macrovascular outcomes associated with smoking may be partly mediated by deleterious changes in microvascular health. More recently, the ARIC study has demonstrated a temporal association between past smoking and wider retinal venules, independent of current smoking status [40], indicating that smoking may provoke long-term structural changes in microcirculation.

In a study examining the relative importance of current and past systemic determinants of retinal venular caliber, it was found that current smoking surpassed mean arterial blood pressure, higher white blood cell count, BMI, and LDL cholesterol levels as the largest determinant of wider retinal venular caliber [40]. The use of retinal venular caliber as a marker of damage from prolonged smoking has been strengthened by recent observations in which ophthalmologists have reported noting retinal venular widening in patients with a history of smoking [19,48].

Endothelial dysfunction and chronic inflammation have been shown to be associated with both retinal vessel caliber [26,60] and smoking [2,33,43], and may partially explain the observed associations between the two. Furthermore, longitudinal studies are required if the cumulative consequences of lifetime exposure to smoking, as well as a timeline for improvement after cessation, are to be determined. More importantly, additional research into the pathophysiology underlying these associations is clearly needed.


The BDES [63] and BMES [36,38] have examined the impact of specific medication use on retinal microvascular structure. These studies found associations between topical β-blockers and retinal arteriolar and venular narrowing [36], and hormone replacement therapy and lower AVR [38]. In the study by Thom et al. [54], it was shown that hypertensive patients receiving calcium channel blocker amlodipine besylate had narrower arterioles than those receiving the β-blocker atenolol. Although these data suggest that antihypertensive treatment may prove useful in decreasing retinal arteriole narrowing due to hypertension, the effects of BP lowering itself was not accounted for. Generalized arteriole narrowing has been shown to be associated with past elevated blood pressure levels [35] and any relationship between antihypertensive medication and retinal microvascular structure may be due to associated decreases in blood pressure.

To examine the effects of ACEI and ARB therapy on retinal vessel diameter, Klein et al. [29] examined a cohort of normotensive individuals with type 1 diabetes receiving antihypertensive treatment. No significant effect of ACEIs or ARBs on retinal vessel caliber was found in this population. This suggests that the beneficial effects of antihypertensive treatment on the retinal microcirculation may be limited to those individuals whose retinal arterioles would probably be narrowed at baseline, and any relationship may be mediated by associated reductions in blood pressure. Further studies, including healthy control subjects, are required to determine if medications have an effect on retinal microcirculation despite controlling for improvements in systemic diseases. If certain medications are found to have direct beneficial effects on retinal microvascular structure, targeted therapeutic interventions may be used to manage preclinical signs of systemic disease.


Epidemiological studies have demonstrated significant increases in cardiovascular morbidity and mortality with increased long- and short-term exposure to air pollution [10]. Impaired microvascular function has been suggested to play a role in these associations. Recently, data from the population-based MESA has shown that retinal arteriolar caliber were narrower and venular caliber were wider among persons living in areas with increased long- and short-term exposure to PM2.5 [1]. Three gram/m3 increases in PM2.5 concentration was associated with arteriole narrowing equivalent to those seen with an age increase of seven years, a more traditional cardiovascular risk factor. These results suggest that important vascular changes occur with small increases in long- and short-term air pollution exposures.

Wider venular diameter with chronic air pollution exposure are consistent with similar investigations into the effects of smoking on retinal microvascular structure [18,19,23,26,28,40,48,60], and may be mediated in part by similar, inflammation-related mechanisms. Long-term exposure to air pollution is known to promote inflammation and endothelial dysfunction [9,49], and may lead to disruptions of microvascular autoregulatory function and venular widening within the retina. Practically, these findings are important in that subclinical microvascular changes (arteriolar narrowing and venular widening) were reported in individuals exposed to PM2.5 levels well below established regulatory thresholds [1]. These data may provide information necessary to establish safer and more accurate regulatory air quality standards.

Conclusions and Clinical Implications

Recent work with regard to selected modifiable risk factors and the retinal microcirculation have added to expanding evidence relating modifiable lifestyle and environmental risk factors to adverse cardiovascular outcomes. It appears that exposure to modifiable risk factors may affect systemic physiology, which is reflected in retinal microvascular structure. As an easily accessible site in which the human microcirculation can be visualized non-invasively and quantified, the potential use of retinal imaging as a biomarker indicating reversible pathophysiologic processes within the systemic circulation is promising.

Nevertheless, evidence showing that quantitative assessment of retinal microvasculature may provide prognostic information beyond current traditional risk factors is very limited. Currently, there have been no established reference levels for age, gender, or disease status, which therefore still limits the utility of retinal imaging as a tool to monitor cardiovascular and metabolic risk in asymptomatic patients or those who have other traditional, positive risk factors.

More longitudinal studies are also needed to determine if changes in retinal microvascular structure can revert to normal with various interventions. Retinal imaging may provide clinicians with a personalized and specific marker to measure the effects of specific interventions on disease progression. Further research into the possible prognostic value of retinal imaging in subjects without clinically evident systemic diseases, such as healthy children, may provide direct associations between retinal vascular structure and risk factors. Long-term systemic disease risk stratification early in life may provide clinicians with information necessary to target microvascular risk factors in therapeutic interventions, even before overt signs of systemic diseases become evident.


Advancing our understanding of the pathophysiology behind changes in retinal microvascular structure in diseased states may aid in the development of novel prediction and intervention strategies for a range of systemic conditions. Although retinal imaging shows a great deal of promise as a potentially powerful clinical tool, further epidemiologic research is needed if it is to become widely used in disease-risk stratification.


Author Biographies

inline image

Kevin Serre is PhD researcher in the Health Sciences and Medicine Faculty at Bond University in Australia. BSc(H) 2004 in Molecular Biology, Queen’s University and Masters of Sports Science 2006, Bond University. His research focuses on the responses in vascular function to exercise in women aged 65-74 years with type 2 diabetes. Kevin is currently the Strength and Conditioning Specialist for the Canadian Military’s Special Operations Regiment.

inline image

Muhammad Bayu Sasongko, MD is a research fellow at the Centre for Eye Research Australia, University of Melbourne, Australia. His research interest includes retinal vascular image analysis and its clinical relevance to systemic vascular diseases and general ophthalmic epidemiology. He is currently undertaking research exploring novel markers obtained from various retinal vascular imaging techniques for diabetic complications and other systemic vascular diseases.