Retinal Vascular Changes in Pre-Diabetes and Prehypertension

Reviews/Commentaries/ADA Statements
 R E V I E W             A R T I C L E

Retinal Vascular Changes in Pre-Diabetes
and Prehypertension
New findings and their research and clinical implications
THANH TAN NGUYEN, MBBS1                                                                                      therapy may also reduce retinopathy re-
JIE JIN WANG, MMED, PHD1,2                                                                                   quiring laser treatment (7).
                                                                                                             Hypertensive retinopathy
                                                                                                             Like diabetic retinopathy, classic hyper-
                                                      CLASSIC RETINAL

       he retinal vasculature can be                                                                         tensive retinopathy is well characterized.
       viewed directly and noninvasively,             VASCULAR CHANGES                                       The clinical signs include generalized and
       offering a unique and easily acces-                                                                   focal arteriolar narrowing, arterio-venous
sible “window” to study the health and                Diabetic retinopathy                                   nicking, increased retinal arteriolar light
disease of the human microcirculation                 In individuals with diabetes, the classic              reflex (copper or silver wiring), flame-
in vivo. In the last decade, advances in              primary retinal vascular complication—                 and blot-shaped retinal hemorrhages,
digital retinal photography and imaging               diabetic retinopathy—is well described                 cotton wool spots, and, in severe cases,
techniques have allowed precise charac-               (1). Diabetic retinopathy signs are                    optic disc swelling (8,9). The association
terization of subtle retinal vascular                 broadly divided into nonproliferative and              of these retinal signs with blood pressure
changes in large populations. These ret-              proliferative retinopathy. The prevalence              is consistent and seen in both adults (10 –
inal changes can be broadly divided into              of diabetic retinopathy increases with                 20) and children (21), even in individuals
four groups: 1) classic retinal vascular              duration of diabetes. The Australian Dia-              without clinical hypertension (13–15,
changes in diabetes and hypertension                  betes, Obesity and Lifestyle Study (Aus-               22–24).
(i.e., diabetic and hypertensive retinop-             Diab) showed that the prevalence of
athy), 2) isolated retinopathy signs in               diabetic retinopathy is less than 10% in               ADVANCES IN ASSESSING
individuals with diabetes or hyperten-                those with diabetes duration of less than 5            RETINAL VASCULAR
sion (e.g., microaneurysm, retinal hem-               years but more than 50% in those with 20               CHANGES — Digital retinal photog-
orrhage, or cotton wool spot), 3)                     years or longer diabetes (2). The two ma-              raphy and new imaging technology have
changes in retinal vascular caliber, and              jor risk factors of diabetic retinopathy are           now allowed more precise assessment of
4) changes in retinal vascular architec-              hyperglycemia and hypertension, with                   the subtle changes seen in the retinal
ture (e.g., retinal tortuosity).                      hyperlipidemia as a possible third major               microvasculature (22,25–28). One key
     New studies in large populations now             risk factor. The importance of hypergly-               development has been methods to objec-
show that retinal vascular changes are                cemia has been confirmed in epidemio-                  tively quantify retinal vascular caliber.
common in the general population and                  logical studies (3), as well as two pivotal            Historically, narrowed retinal arteriolar
may precede the subsequent develop-                   clinical trials: the DCCT (Diabetes Con-               caliber, an early hypertensive retinopathy
ment of overt diabetes and hypertension. A            trol and Complications Trial) in patients              sign, has been difficult to measure using
consistent pattern of associations is also            with type 1 diabetes (4) and the UKPDS                 the clinical ophthalmoscope (29). Parr,
                                                                                                             Hubbard, and colleagues (22,30,31) de-
emerging, showing that specific retinal vas-          (UK Prospective Diabetes Study) in pa-
                                                                                                             veloped techniques to measure retinal
cular changes may be related differently to           tients with type 2 diabetes (5). The UK-
                                                                                                             vascular caliber from photographs and
hyperglycemia and blood pressure.                     PDS also showed that blood pressure                    summarized these as the arterio-venous
     In this review, we summarize recent              control reduces the risk of retinopathy in-            ratio (AVR). These techniques are now
studies on the retinal vascular changes               dependent of glycemia levels (6). New                  used in large epidemiological studies
seen in diabetes and hypertension and                 data from the FIELD (Fenofibrate Inter-                (20,22,25,27,28) and have substantial re-
speculate on potential research and clini-            vention and Event Lowering in Diabetes)                producibility.
cal implications.                                     Study now suggest that lipid-lowering                       Recent studies suggest the interpreta-
                                                                                                             tion of the AVR may be overly simplistic.
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●            A smaller AVR was thought to reflect gen-
From the 1Centre for Eye Research Australia, University of Melbourne, Victoria, Australia; the 2Centre for   eralized retinal arteriolar narrowing, since
Vision Research, University of Sydney, New South Wales, Australia; and the 3Singapore Eye Research           venular caliber was assumed to be rela-
Institute, National University of Singapore, Singapore.
   Address correspondence and reprint requests to Tien Yin Wong, MD, PhD, Centre for Eye Research
                                                                                                             tively constant (22). Thus, when a low
Australia, University of Melbourne, 32 Gisborne St., Victoria 3002, Australia. E-mail: twong@                AVR was associated with elevated blood                                                                                              pressure (18,32) and cardiovascular out-
   Received for publication 15 April 2007 and accepted in revised form 20 June 2007.                         comes such as stroke (33,34) and coro-
   Published ahead of print at on 26 June 2007. DOI: 10.2337/dc07-0732.     nary heart disease (35), the associations
   Abbreviations: ARIC, Atherosclerosis Risk In Communities; AVR, arterio-venous ratio.
   A table elsewhere in this issue shows conventional and Système International (SI) units and conversion   were initially thought to reflect general-
factors for many substances.                                                                                 ized arteriolar narrowing. Newer analy-
   © 2007 by the American Diabetes Association.                                                              ses, however, suggest a smaller AVR may

2708                                                                                                    DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007
Nguyen, Wang, and Wong

not only reflect narrower arterioles but
also wider venules (36,37). Furthermore,
arteriolar and venular calibers appear to
reflect different pathophysiological pro-
cesses (37,38).
     Liew et al. (38,39) have suggested the
need to control for venular caliber in sta-
tistical models of arteriolar caliber, and
vice versa, as venular caliber explains
⬃30% of the variability in arteriolar cali-
ber (38)—presumably from shared ge-
netic and ocular factors (22).
     Improvements in imaging software
have also led to quantification of other
architectural changes in the retinal vascu-
lar network (36,37), as well as “batch pro-
cessing” of retinal images (20,22,
25,27,28). There remain technical chal-
lenges. For example, the impact of mag-
nification error (i.e., eyes of different re-
fraction) requires further study (40).

                                                Figure 1—Prevalence of retinopathy in nondiabetic populations (A) and nondiabetic and non-
Epidemiology                                    hypertensive populations (B). AusDiab: Australian Diabetes, Obesity and Lifestyle Study (year
There is increasing evidence that typical       began: 1999 –2000, n ⫽ 2,177, aged ⱖ25 years) (2); ARIC (year began: 1987–1990, n ⫽ 10,954,
lesions of diabetic retinopathy (microan-       aged 48 –73 years) (22); BDES: Beaver Dam Eye Study (year began: 1988 –1990, n ⫽ 4,926, aged
eurysms, hemorrhages, and cotton wool           43– 84 years) (15); BMES: Blue Mountain Eye Study (year began: 1992–1994, n ⫽ 3,654, aged
spots), termed isolated retinopathy signs,      49 –97 years) (13); CHS: Cardiovascular Health Study (year began: 1989 –1990, n ⫽ 2,050, aged
are now recognized to be more common            67–97 years) (14), Hoorn (year began: 1989 –1992, n ⫽ 626, aged 50 –74 years) (23), Rotterdam
than previously thought in people with-         (year began: 1990 –1993, n ⫽ 6,191, aged 55–99 years) (44), and Funagata (year began: 2000 –
out diabetes and hypertension (41– 43).         2002, n ⫽ 1,481, aged ⱖ35 years) (24).
Recent studies using retinal photography
to document these signs suggest preva-          Risk factors and pathophysiology                isolated retinopathy in people without di-
lence rates in the general population of        The underlying risk factors and patho-          abetes (50) supports the hypothesis that
5–10% (2,13–15,22–24,44) (Fig. 1A)              physiology of isolated retinopathy signs        inflammatory processes may also be a
and 2.6 – 8.6% among those without dia-         in nondiabetic and normotensive individ-        possible pathway that underlies early sub-
betes or hypertension (13–15,22–24)             uals are poorly understood. Associations        clinical microvascular disease in the pre-
(Fig. 1B). Prospective study data have fur-     of these retinopathy signs with increasing      diabetes or prehypertension state.
ther shown that up to 10% of individuals        age (15,45), elevated blood pressure
aged ⱖ40 years without diabetes may de-         (8,13–15,44,47– 49), and hyperglycemia          Associations with risk of diabetes,
velop these isolated retinopathy signs          (24,44,47,48) have been found. Other            hypertension, and cardiovascular
within 5 years (16,45).                         possible risk factors include hyperlipid-       diseases
     Two studies have reported on the           emia (17,23,48), higher BMI (23,24), and        A clinically relevant question is whether
prevalence of retinopathy in individuals        systemic inflammation (48,50). We can           signs of isolated retinopathy in individu-
with pre-diabetes. In the AusDiab Study,        speculate that isolated retinopathy signs       als without diabetes are markers of the
retinopathy signs were seen in 6.7% of          in normotensive and nondiabetic individ-        future risk of diabetes (i.e., do these pa-
individuals with impaired glucose toler-        uals may represent early microvascular          tients require monitoring for the develop-
ance or impaired fasting glucose (46),          damage from a combination of risk fac-          ment of diabetes). The evidence here is
whereas in the Diabetes Prevention Pro-         tors, including blood pressure and abnor-       not consistent. While previous studies
gram this was seen in 7.9% of individuals       mal glucose metabolism, which may               suggest that detectable retinopathy pre-
with impaired fasting glucose (5.3– 6.9         reflect an underlying process of develop-       cedes the onset of type 2 diabetes by 4 –7
mmol/l) or impaired glucose tolerance,          ing clinical diabetes or hypertension.          years (54), new prospective data from the
who had no history of diabetes (47).                 Animal models and human studies            Blue Mountains (45,55), the Atheroscle-
     These isolated retinopathy signs may       suggest that chronic inflammation and           rosis Risk In Communities (ARIC) (56),
be transient. Population studies show that      glucose-induced arteriolar endothelial          and Beaver Dam (57) studies reported no
between 40 and 70% of these isolated ret-       dysfunction are related to development of       increased risk of diabetes in nondiabetic
inopathy signs seen at baseline are not         classic diabetic retinopathy (51–53). The       individuals with signs of retinopathy.
present 3–5 years later (45,48).                association of inflammation and signs of        However, there are two notable excep-

DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007                                                                                      2709
Retinal vascular changes

Table 1—Associations of retinal vascular changes with diabetes, hypertension, and cardiovascular diseases

Retinal vascular signs                           Associations                                  Populations                            References
Retinopathy                           Impaired fasting glucose                    ARIC                                                  (67)
                                      Obesity                                     Hoorn                                                 (23)
                                      Blood pressure                              ARIC, AusDiab, BDES, BMES, CHS,                (2,13–15,22–24,44)
                                                                                  Funagata, Hoorn, Rotterdam
                                      Incident hypertension                       BDES                                                   (57)
                                      Incident diabetes                           ARIC, BDES                                           (56,57)
                                      Heart disease                               BDES, CHS, ARIC                                    (14,59,114)
                                      Nephropathy                                 ARIC, CHS                                            (60,61)
                                      Cerebrovascular disease                     ARIC, CHS, BDES                                (14,33,34,114–117)
Retinal arteriolar narrowing          Blood pressure                              Funagata, BDES, BMES, Rotterdam,               (10,22,24,32,37,87)
                                                                                  ARIC, CHS
                                      Blood pressure in children                  SCES, SCORM                                             (21)
                                      Measures of atherosclerosis                 Rotterdam                                               (37)
                                      Waist-to-hip ratio                          ARIC                                                    (67)
                                      Incident hypertension                       ARIC, BMES, BDES, Rotterdam                         (58,88–90)
                                      Incident diabetes                           ARIC, BDES                                            (64,65)
                                      Coronary heart disease                      CHS                                                     (98)
Retinal venular dilatation            Impaired fasting glucose                    ARIC, MESA                                            (20,67)
                                      Measures of atherosclerosis                 Rotterdam                                               (37)
                                      Obesity in children                         SCORM                                                   (69)
                                      Waist-to-hip ratio                          ARIC                                                    (67)
                                      Hypertriglyceridemia                        ARIC                                                    (67)
                                      Incident obesity                            BMES                                                   (118)
                                      Incident hypertension                       BMES                                                    (39)
                                      Incident impaired fasting glucose           Rotterdam                                               (66)
                                      Cerebrovascular diseases                    Rotterdam, CHS                                        (96–98)
                                      Carotid artery disease                      Rotterdam                                               (37)
AusDiab, Australian Diabetes, Obesity and Lifestyle Study; BDES, Beaver Dam Eye Study; BMES, Blue Mountain Eye Study; CHS, Cardiovascular Health Study;
MESA, Multi-Ethnic Study of Atherosclerosis; SCES, Sydney Childhood Eye Study; SCORM, Singapore Cohort Study of Risk Factors for Myopia.

tions. First, in the Beaver Dam Study,             that isolated microaneurysms, hemor-               sion and prehypertension, are sum-
among individuals aged ⬍65 years at                rhages, and cotton wool spots predict the          marized in Table 1.
baseline, signs of retinopathy were asso-          development of clinical cardiovascular
ciated with an increased 15-year inci-             and cerebrovascular events independent             Associations with diabetes and pre-
dence of diabetes (odds ratio [OR] 3.68            of traditional risk factors (Table 1). Vari-       diabetes
[95% CI 1.23–10.96]) (57). Second, in              ous population-based studies have shown            Consistent associations of retinal venular
the ARIC Study, among participants with            associations of isolated signs of retinopa-        caliber with hyperglycemia, diabetes, and
a family history of diabetes, signs of reti-       thy with the risk of stroke (33), congestive       its complications are now emerging (20).
nopathy were also associated with an in-           heart failure (59), renal dysfunction              Prospective data from three population-
creased 3-year risk of diabetes (2.3 [1.0 –        (60,61), and measures of atherosclerosis           based cohorts have shown that changes in
5.3]) (56). Thus, the literature suggests          (14,62,63). In the ARIC Study, the pres-           retinal vascular caliber may predict the
that isolated signs of retinopathy in indi-        ence of retinopathy was associated with a          development of type 2 diabetes (64,65)
viduals without diabetes are not necessar-         threefold higher risk of congestive heart          and impaired fasting glucose (66). In two
ily markers of future diabetes risk, except        failure in those without previous coro-            early analyses, an association between
possibly in younger individuals and in             nary heart disease (relative risk 2.98 [95%        smaller retinal AVR and incident diabetes
those with a family history of diabetes.           CI 1.50 –5.92]) (59). This later associa-          was found in the ARIC Study (OR 1.71
     Similarly, there is conflicting evi-          tion suggests that microvascular disease           [95% CI 1.13–2.57]; comparing smallest
dence that isolated signs of retinopathy           may be important in the development of             to largest AVR quintile) and the Beaver
are markers of future hypertension risk.           diabetic cardiomyopathy in the absence             Dam Study (1.53 [1.03–2.27]; comparing
Data from the Beaver Dam Study showed              of established coronary artery disease.            smallest to largest AVR quartile) (64,65).
that among nonhypertensive individuals,                                                               Subsequently, the Rotterdam Study dem-
those with signs of retinopathy had a                                                                 onstrated that these associations reflected
higher incidence of hypertension (OR               RETINAL VASCULAR                                   wider retinal venular caliber rather than
1.48 [95% CI 1.05–2.07]) (57), but these           CALIBER — The associations and clin-               narrower arteriolar caliber (1.23 [1.02–
findings are not supported by other stud-          ical significance of early retinal vascular        1.47]; per SD increase in venular caliber)
ies (15,45,58).                                    caliber changes in individuals with diabe-         (66). Reanalysis of the ARIC and Beaver
     There is now substantial evidence             tes and pre-diabetes, as well as hyperten-         Dam studies confirms this finding

2710                                                                                              DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007
Nguyen, Wang, and Wong

(T.Y.W., unpublished data). Thus, it ap-        ber, there is now substantial evidence that    Associations with cardiovascular
pears that wider retinal venular caliber is a   hypertension preferentially affects retinal    diseases
marker of chronic hyperglycemia and the         arteriolar caliber (83) (Table 1). It has      In addition to their associations with dia-
pre-diabetes state and reflects the early       long been known that generalized retinal       betes and hypertension, changes in retinal
microvascular changes that occur in the         arteriolar narrowing is an early character-    vascular caliber have also been linked to a
development of diabetes.                        istic sign of hypertensive retinopathy         range of cardiovascular diseases (Table 1).
     Wider retinal venular caliber has also     (8,84,85). More recent studies using           Wider retinal venular caliber has been as-
been linked to the metabolic syndrome           quantitative measurements of retinal vas-      sociated with carotid artery disease (37),
and its components (20,37,67). In the           cular caliber have now demonstrated a          magnetic resonance imaging– detected la-
Blue Mountains Study, wider retinal             graded association of narrowed retinal ar-     cunar infarcts and white matter lesions
venular caliber was associated with the         terioles with increasing blood pressure in     (96), and clinical stroke events (97,98). In
5-year incidence of obesity among indi-         different populations of various racial/       the Cardiovascular Health Study, wider
viduals of normal weight at baseline (OR        ethnic groups and age-groups (10,20,           retinal venular caliber was predictive of
1.8 [95% CI 1.0 –3.1]; comparing largest        22,24,32,37,86,87).                            incident coronary heart disease (rate ratio
to lowest venular caliber quintile) (68).                                                      3.0 [95% CI 1.6 –5.7]; comparing largest
                                                     Of greater significance are prospec-
Similarly, in children aged 6 – 8 years,                                                       to smallest venular caliber quartile) and
                                                tive findings from four populations that
wider retinal venular caliber was also as-                                                     incident stroke (2.2 [1.1– 4.3]), whereas
sociated with higher BMI (69), suggesting       show retinal arteriolar narrowing is a pre-
                                                                                               narrower arteriolar caliber was predictive
that retinal venular caliber may be influ-      clinical marker of hypertension risk. The      of incident coronary heart disease (2.0
enced by metabolic disorders early in life.     ARIC Study (OR 1.62 [95% CI 1.21–              [1.1–3.7]; comparing smallest to largest
There is also evidence that wider venular       2.18]; comparing smallest to largest AVR       arteriolar caliber quartile) (98). These
caliber is associated with various micro-       quintile) (58), the Beaver Dam Study           findings suggest that both wider venular
vascular complications of diabetes, not         (1.82 [1.39 –2.40]; comparing smallest to      caliber and narrower arteriolar caliber
only diabetic retinopathy (70,71) but also      largest AVR quartile) (88), the Blue           may be markers of early subclinical car-
diabetic nephropathy (72).                      Mountains Study (2.6 [1.7–3.9]; compar-        diovascular disease.
     Despite these observations, the            ing smallest to largest arteriolar caliber
pathophysiological processes underlying         quintile) (89), and the Rotterdam Study        CHANGES IN RETINAL
the association of wider retinal venular        (1.38 [1.23–1.55]; per SD decrease in ar-      VASCULAR ARCHITECTURE —
caliber with hyperglycemia, diabetes, and       teriolar caliber) (90) all reported that       New imaging methods have allowed the
its complications are unclear. It has been      among individuals without hypertension         measurement of other architectural
speculated that retinal venular widening        at baseline, those with narrowed retinal       changes in the retinal microvasculature.
may be the result of increased blood flow       arterioles had a higher risk of hyperten-      Hypertension, for example, has been as-
associated with hyperglycemia (73) and          sion in the subsequent 3–10 years, inde-       sociated with an increase in the retinal
retinal hypoxia (74). Alternatively, it may     pendent of baseline blood pressure levels,     arteriolar length-to-diameter ratio
also reflect inflammatory processes impli-      BMI, and other known hypertension risk         (99,100), increased retinal venular tortu-
cated in the pathogenesis of impaired glu-      factors.                                       osity (99), reduced branching angle at ar-
cose metabolism (75), supported by                   These observations support the hy-        teriole bifurcations (101), and reduced
epidemiological findings of wider retinal       pothesis that peripheral vascular resis-       microvascular density (99,101,102).
venules with elevated systemic inflamma-        tance, reflected by retinal arteriolar             Some of these retinal changes have
tory markers (17,20,37,76). Experiments         narrowing, is an important contributing        also been shown to be associated with in-
have demonstrated that local inflamma-          factor for hypertension development            creased cardiovascular risk. For example,
tory processes lead to wider retinal venu-      (91). Added support to this hypothesis         the Beaver Dam Study demonstrated that
lar calibers. For example, administration       comes from a recent genome-wide link-          suboptimal arteriolar bifurcation and de-
of lipid hydroperoxide in the vitreous of       age analysis from the Beaver Dam Study,        creased arteriolar tortuosity are associated
rats leads to an increase in the retinal                                                       with coronary heart mortality (103).
                                                which demonstrated that associations of
venular diameter (77). Similarly, admin-
                                                retinal arteriolar diameter to multiple ge-
istration of Escherichia coli endotoxin in                                                     IMPLICATIONS FOR
                                                netic loci are linked to regulation of blood
human eyes has been reported to increase                                                       RESEARCH AND CLINICAL
retinal venular diameter (78). Finally, ret-    pressure, endothelial function, and vas-       MANAGEMENT — It is now well
inal venular dilation may be related to en-     culogenesis (92). Thus, retinal arteriolar     recognized that individuals with impaired
dothelial dysfunction, reflecting an            narrowing may be considered a surrogate        glucose metabolism or pre-diabetes have
increased production of nitric oxide (79)       marker of an individual’s genetic predis-      higher mortality from cardiovascular dis-
secondary to higher levels of cytokines         position to hypertension development           ease (104 –106). Similarly, individuals
(80), seen often in association with im-        (93).                                          with high to normal blood pressure or
paired glucose metabolism (81,82) and                Finally, a recent study has shown that    prehypertension (107) are more likely to
diabetes (81).                                  the association between higher blood           develop cardiovascular events (108,109).
                                                pressure and retinal arteriolar narrowing      To permit appropriate preventative strat-
                                                is detectable in healthy children aged 6 – 8   egies, there is therefore great interest in
Associations with hypertension and              years (21), reinforcing the concept that       early detection of individuals with pre-
prehypertension                                 the effects of higher childhood blood          diabetes and prehypertension.
In distinct contrast to the association of      pressure may have an adverse effect on             This review suggests that retinal im-
hyperglycemia with retinal venular cali-        the microcirculation (94,95).                  age analysis offers a novel noninvasive

DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007                                                                                    2711
Retinal vascular changes

measurement of early changes in the vas-       nal vascular changes using new imaging                   on cardiovascular events in 9795 people
culature—not detectable on routine clin-       techniques offers great potential to ad-                 with type 2 diabetes mellitus (the FIELD
ical examination—that may allow the            vance our understanding of the early                     study): randomised controlled trial. Lan-
identification of individuals at risk of di-   pathophysiological pathways of diabetes                  cet 366:1849 –1861, 2005
                                                                                                   8.   Wong TY, Mitchell P: Hypertensive ret-
abetes and hypertension and their subse-       and hypertension development. Recent
                                                                                                        inopathy. N Engl J Med 351:2310 –2317,
quent complications. Retinal vascular          studies support the concept that the reti-               2004
imaging might also permit physicians to        nal vasculature provides a summary mea-             9.   Wong T, Mitchell P: The eye in hyper-
optimize management of individuals with        sure of lifetime exposure to various                     tension. Lancet 369:425– 435, 2007
established diabetes and/or hypertension.      processes involved in the development of          10.    Leung H, Wang JJ, Rochtchina E, Tan
For example, retinal vascular imaging          diabetes and hypertension. Furthermore,                  AG, Wong TY, Klein R, Hubbard LD,
may allow monitoring of chronic varia-         these studies suggest that the effects of                Mitchell P: Relationships between age,
tions in glucose and blood pressure, as        glucose and blood pressure on the retinal                blood pressure, and retinal vessel diam-
well as the presence and severity of sub-      microvasculature are graded and contin-                  eters in an older population. Invest Oph-
clinical microvascular damage. However,        uous, and our current definitions of dia-                thalmol Vis Sci 44:2900 –2904, 2003
a number of issues should be resolved be-      betic and hypertensive retinopathy are            11.    Sharp PS, Chaturvedi N, Wormald R,
                                                                                                        McKeigue PM, Marmot MG, Young SM:
fore retinal vascular imaging can be uti-      arbitrary and do not capture early disease.              Hypertensive retinopathy in Afro-Carib-
lized in clinical practice.                    Future research is clearly needed to assess              beans and Europeans: prevalence and
     First, despite a large body of data on    the ability of retinal vascular imaging to               risk factor relationships. Hypertension
the associations and risk prediction of ret-   provide clinically useful information that               25:1322–1325, 1995
inal vascular caliber measurement in dif-      adds to existing risk prediction models of        12.    Wang JJ, Mitchell P, Leung H, Roch-
ferent population-based studies, there is      diabetes and hypertension.                               tchina E, Wong TY, Klein R: Hyperten-
no accepted standardized classification of                                                              sive retinal vessel wall signs in a general
retinal vascular changes, and a lack of                                                                 older population: the Blue Mountains
age-, sex-, body size–, and blood pres-        References                                               Eye Study. Hypertension 42:534 –541,
sure–specific normative data. New stud-          1. Ciulla TA, Amador AG, Zinman B: Dia-                2003
                                                    betic retinopathy and diabetic macular       13.    Yu T, Mitchell P, Berry G, Li W, Wang JJ:
ies of retinal vascular changes in children,                                                            Retinopathy in older persons without di-
                                                    edema: pathophysiology, screening, and
who are generally free of many systemic             novel therapies. Diabetes Care 26:2653–             abetes and its relationship to hyperten-
conditions, may provide these reference             2664, 2003                                          sion. Arch Ophthalmol 116:83– 89, 1998
data (21,110).                                   2. Tapp RJ, Shaw JE, Harper CA, de Cour-        14.    Wong TY, Klein R, Sharrett AR, Manolio
     Second, for retinal vascular imaging           ten MP, Balkau B, McCarty DJ, Taylor                TA, Hubbard LD, Marino EK, Kuller L,
to be useful for risk stratification, there         HR, Welborn TA, Zimmet PZ: The prev-                Burke G, Tracy RP, Polak JF, Gottdiener
must be a demonstration of independent              alence of and factors associated with di-           JS, Siscovick DS: The prevalence and
predictive value that substantially adds to         abetic retinopathy in the Australian                risk factors of retinal microvascular ab-
traditional methods. This has not been              population. Diabetes Care 26:1731–                  normalities in older persons: the Cardio-
conclusively demonstrated. Different an-            1737, 2003                                          vascular Health Study. Ophthalmology
                                                 3. Klein R, Klein BE, Moss SE, Davis MD,               110:658 – 666, 2003
alytical methods have hampered the com-
                                                    DeMets DL: Glycosylated hemoglobin           15.    Klein R, Klein BE, Moss SE, Wang Q:
parison of results between studies, and             predicts the incidence and progression              Hypertension and retinopathy, arterio-
application-common methods in differ-               of diabetic retinopathy. JAMA 260:                  lar narrowing, and arteriovenous nick-
ent studies will allow data pooling to              2864 –2871, 1988                                    ing in a population. Arch Ophthalmol
generate more valid risk estimates. Addi-        4. The effect of intensive treatment of dia-           112:92–98, 1994
tionally, the role of novel measures of ret-        betes on the development and progres-        16.    Klein R, Klein BE, Moss SE: The relation
inal vascular structure (99,103,111–113)            sion of long-term complications in                  of systemic hypertension to changes in
in predicting diseases remains to be deter-         insulin-dependent diabetes mellitus: the            the retinal vasculature: the Beaver Dam
mined.                                              Diabetes Control and Complications                  Eye Study. Trans Am Ophthalmol Soc 95:
     Third, the predictive value of retinal         Trial Research Group. N Engl J Med 329:             329 –348, 1997
vascular imaging is currently based on as-          977–986, 1993                                17.    Klein R, Sharrett AR, Klein BE, Chamb-
                                                 5. The UK Propsective Diabetes Study                   less LE, Cooper LS, Hubbard LD, Evans
sociations seen in large population-based           Group: Intensive blood-glucose control              G: Are retinal arteriolar abnormalities
samples (114 –117). It is unclear whether           with sulphonylureas or insulin com-                 related to atherosclerosis? The Athero-
the retinal measurements are sufficiently           pared with conventional treatment and               sclerosis Risk in Communities Study.
precise to differentiate risk at an individ-        risk of complications in patients with              Arterioscler Thromb Vasc Biol 20:1644 –
ual level.                                          type 2 diabetes (UKPDS 33). Lancet 352:             1650, 2000
     Finally, it is unknown whether mod-            837– 853, 1998                               18.    Sharrett AR, Hubbard LD, Cooper LS,
ification of risk factors (e.g., increased       6. The UK Propsective Diabetes Study                   Sorlie PD, Brothers RJ, Nieto FJ, Pinsky
physical activity, reduction in weight) or          Group: Tight blood pressure control and             JL, Klein R: Retinal arteriolar diameters
institution of treatment (e.g., diabetes and        risk of macrovascular and microvascular             and elevated blood pressure: the Athero-
antihypertensive medications) may im-               complications in type 2 diabetes:                   sclerosis Risk in Communities Study.
                                                    UKPDS 38. BMJ 317:703–713, 1998                     Am J Epidemiol 150:263–270, 1999
prove retinal vascular measures and              7. Keech A, Simes RJ, Barter P, Best J, Scott   19.    Duncan BB, Wong TY, Tyroler HA,
whether this is associated with lowered             R, Taskinen MR, Forder P, Pillai A, Davis           Davis CE, Fuchs FD: Hypertensive reti-
risks of diabetes, hypertension, and their          T, Glasziou P, Drury P, Kesaniemi YA,               nopathy and incident coronary heart
complications. This remains an important            Sullivan D, Hunt D, Colman P, d’Emden               disease in high risk men. Br J Ophthalmol
area of future research.                            M, Whiting M, Ehnholm C, Laakso M:                  86:1002–1006, 2002
     In conclusion, measurement of reti-            Effects of long-term fenofibrate therapy     20.    Wong TY, Islam FM, Klein R, Klein BE,

2712                                                                                        DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007
Nguyen, Wang, and Wong

       Cotch MF, Castro C, Sharrett AR, Shahar          Am J Ophthalmol 77:472– 477, 1974            42. Svardsudd K, Wedel H, Aurell E, Tibblin
       E: Retinal vascular caliber, cardiovascu-    31. Parr JC, Spears GF: Mathematic relation-         G: Hypertensive eye ground changes:
       lar risk factors, and inflammation: the          ships between the width of a retinal ar-         prevalence, relation to blood pressure
       multi-ethnic study of atherosclerosis            tery and the widths of its branches. Am J        and prognostic importance: the study of
       (MESA). Invest Ophthalmol Vis Sci 47:            Ophthalmol 77:478 – 483, 1974                    men born in 1913. Acta Med Scand 204:
       2341–2350, 2006                              32. Wong TY, Klein R, Klein BE, Meuer SM,            159 –167, 1978
 21.   Mitchell P, Cheung N, de Haseth K, Tay-          Hubbard LD: Retinal vessel diameters         43. McDonough JR, Garrison GE, Hames
       lor B, Rochtchina E, Islam FMA, Wang             and their associations with age and              CG: Blood pressure and hypertensive
       JJ, Saw SM, Wong TY: Blood pressure              blood pressure. Invest Ophthalmol Vis Sci        disease among negroes and whites: a
       and retinal arteriolar narrowing in chil-        44:4644 – 4650, 2003                             study in Evans County, Georgia. Ann In-
       dren. Hypertension 49:1156 –1162, 2007       33. Wong TY, Klein R, Couper DJ, Cooper              tern Med 61:208 –228, 1964
 22.   Hubbard LD, Brothers RJ, King WN,                LS, Shahar E, Hubbard LD, Wofford MR,        44. Stolk RP, Vingerling JR, de Jong PT,
       Clegg LX, Klein R, Cooper LS, Sharrett           Sharrett AR: Retinal microvascular ab-           Dielemans I, Hofman A, Lamberts SW,
       AR, Davis MD, Cai J: Methods for evalu-          normalities and incident stroke: the Ath-        Pols HA, Grobbee DE: Retinopathy, glu-
       ation of retinal microvascular abnormal-         erosclerosis Risk in Communities Study.          cose, and insulin in an elderly popula-
       ities associated with hypertension/              Lancet 358:1134 –1140, 2001                      tion: the Rotterdam Study. Diabetes 44:
       sclerosis in the Atherosclerosis Risk in     34. Mitchell P, Wang JJ, Wong TY, Smith W,           11–15, 1995
       Communities Study. Ophthalmology                 Klein R, Leeder SR: Retinal microvascu-      45. Cugati S, Cikamatana L, Wang JJ, Kifley
       106:2269 –2280, 1999                             lar signs and risk of stroke and stroke          A, Liew G, Mitchell P: Five-year inci-
 23.   van Leiden HA, Dekker JM, Moll AC,               mortality. Neurology 65:1005–1009,               dence and progression of vascular reti-
       Nijpels G, Heine RJ, Bouter LM, Stehou-          2005                                             nopathy in persons without diabetes: the
       wer CD, Polak BC: Blood pressure, lip-       35. Wong TY, Klein R, Sharrett AR, Duncan            Blue Mountains Eye Study. Eye 20:
       ids, and obesity are associated with             BB, Couper DJ, Tielsch JM, Klein BE,             1239 –1245, 2006
       retinopathy: the Hoorn Study. Diabetes           Hubbard LD: Retinal arteriolar narrow-       46. Wong TY, Barr EL, Tapp RJ, Harper CA,
       Care 25:1320 –1325, 2002                         ing and risk of coronary heart disease in        Taylor HR, Zimmet PZ, Shaw JE: Reti-
 24.   Kawasaki R, Wang JJ, Rochtchina E,               men and women: the Atherosclerosis               nopathy in persons with impaired
       Taylor B, Wong TY, Tominaga M, Kato              Risk in Communities Study. JAMA 287:             glucose metabolism: the Australian Dia-
       T, Daimon M, Oizumi T, Kawata S,                 1153–1159, 2002                                  betes Obesity and Lifestyle (AusDiab)
       Kayama T, Yamashita H, Mitchell P: Car-      36. Patton N, Aslam T, Macgillivray T, Dhil-         study. Am J Ophthalmol 140:1157–1159,
       diovascular risk factors and retinal mi-         lon B, Constable I: Asymmetry of retinal         2005
       crovascular signs in an adult Japanese           arteriolar branch widths at junctions af-    47. The prevalence of retinopathy in im-
       population: the Funagata Study. Oph-             fects ability of formulae to predict trunk       paired glucose tolerance and recent-on-
       thalmology 113:1378 –1384, 2006                  arteriolar widths. Invest Ophthalmol Vis         set diabetes in the Diabetes Prevention
 25.   Couper DJ, Klein R, Hubbard LD, Wong             Sci 47:1329 –1333, 2006                          Program. Diabet Med 24:137–144, 2007
       TY, Sorlie PD, Cooper LS, Brothers RJ,       37. Ikram MK, de Jong FJ, Vingerling JR,         48. Wong TY, Klein R, Islam FM, Cotch MF,
       Nieto FJ: Reliability of retinal photogra-       Witteman JC, Hofman A, Breteler MM,              Couper DJ, Klein BE, Hubbard L, Shar-
       phy in the assessment of retinal micro-          de Jong PT: Are retinal arteriolar or            rett AR: Three-year incidence and cumu-
       vascular characteristics: the Athero-            venular diameters associated with mark-          lative prevalence of retinopathy: the
       sclerosis Risk in Communities Study.             ers for cardiovascular disorders? The            Atherosclerosis Risk in Communities
       Am J Ophthalmol 133:78 – 88, 2002                Rotterdam Study. Invest Ophthalmol Vis           Study. Am J Ophthalmol 143:970 –976,
 26.   Sherry LM, Wang JJ, Rochtchina E,                Sci 45:2129 –2134, 2004                          2007
       Wong T, Klein R, Hubbard L, Mitchell P:      38. Liew G, Sharrett AR, Kronmal R, Klein R,     49. Wong TY, Klein R, Duncan BB, Nieto FJ,
       Reliability of computer-assisted retinal         Wong TY, Mitchell P, Kifley A, Wang JJ:          Klein BE, Couper DJ, Hubbard LD, Shar-
       vessel measurement in a population.              Measurement of retinal vascular caliber:         rett AR: Racial differences in the preva-
       Clin Experiment Ophthalmol 30:179 –              issues and alternatives to using the arte-       lence of hypertensive retinopathy.
       182, 2002                                        riole to venule ratio. Invest Ophthalmol         Hypertension 41:1086 –1091, 2003
 27.   Li H, Hsu W, Lee ML, Wong TY: Auto-              Vis Sci 48:52–57, 2007                       50. van Hecke MV, Dekker JM, Nijpels G,
       matic grading of retinal vessel caliber.     39. Liew G, Wong TY, Mitchell P, Wang JJ:            Moll AC, Heine RJ, Bouter LM, Polak BC,
       IEEE Trans Biomed Eng 52:1352–1355,              Are narrower or wider retinal venules as-        Stehouwer CD: Inflammation and endo-
       2005                                             sociated with incident hypertension?             thelial dysfunction are associated with
 28.   Wong TY, Knudtson MD, Klein R, Klein             Hypertension 48:e10, 2006                        retinopathy: the Hoorn Study. Diabeto-
       BE, Meuer SM, Hubbard LD: Computer-          40. Patton N, Aslam TM, MacGillivray T,              logia 48:1300 –1306, 2005
       assisted measurement of retinal vessel           Deary IJ, Dhillon B, Eikelboom RH, Yo-       51. Miyamoto K, Khosrof S, Bursell SE, Ro-
       diameters in the Beaver Dam Eye Study:           gesan K, Constable IJ: Retinal image             han R, Murata T, Clermont AC, Aiello
       methodology, correlation between eyes,           analysis: concepts, applications and po-         LP, Ogura Y, Adamis AP: Prevention of
       and effect of refractive errors. Ophthal-        tential. Prog Retin Eye Res 25:99 –127,          leukostasis and vascular leakage in
       mology 111:1183–1190, 2004                       2006                                             streptozotocin-induced diabetic reti-
 29.   Wagener HP, Clay GE, Gipner JF: Clas-        41. Leibowitz HM, Krueger DE, Maunder                nopathy via intercellular adhesion mol-
       sification of retinal lesions in the pres-       LR, Milton RC, Kini MM, Kahn HA,                 ecule-1 inhibition. Proc Natl Acad Sci U S
       ence of vascular hypertension: report            Nickerson RJ, Pool J, Colton TL, Ganley          A 96:10836 –10841, 1999
       submitted to the American Ophthalmo-             JP, Loewenstein JI, Dawber TR: The Fra-      52. Izuora KE, Chase HP, Jackson WE, Coll
       logical Society by the committee on              mingham Eye Study monograph: an                  JR, Osberg IM, Gottlieb PA, Rewers MJ,
       Classification of Hypertensive Disease of        ophthalmological and epidemiological             Garg SK: Inflammatory markers and di-
       the Retina. Trans Am Ophthalmol Soc 45:          study of cataract, glaucoma, diabetic ret-       abetic retinopathy in type 1 diabetes. Di-
       57–73, 1947                                      inopathy, macular degeneration, and vi-          abetes Care 28:714 –715, 2005
 30.   Parr JC, Spears GF: General caliber of the       sual acuity in a general population of       53. Joussen AM, Poulaki V, Le ML, Koizumi
       retinal arteries expressed as the equiva-        2631 adults, 1973–1975. Surv Ophthal-            K, Esser C, Janicki H, Schraermeyer U,
       lent width of the central retinal artery.        mol 24:335– 610, 1980                            Kociok N, Fauser S, Kirchhof B, Kern TS,

DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007                                                                                            2713
Retinal vascular changes

       Adamis AP: A central role for inflamma-             of diabetes mellitus. Arch Intern Med               Y: Lipid hydroperoxide stimulates leu-
       tion in the pathogenesis of diabetic reti-          165:1060 –1065, 2005                                kocyte-endothelium interaction in the
       nopathy. FASEB J 18:1450 –1452, 2004          66.   Ikram MK, Janssen JA, Roos AM, Ri-                  retinal microcirculation. Exp Eye Res 75:
 54.   Harris MI, Klein R, Welborn TA,                     etveld I, Witteman JC, Breteler MM,                 69 –75, 2002
       Knuiman MW: Onset of NIDDM occurs                   Hofman A, van Duijn CM, de Jong PT:           78.   Kolodjaschna J, Berisha F, Lung S,
       at least 4 –7 yr before clinical diagnosis.         Retinal vessel diameters and risk of im-            Schaller G, Polska E, Jilma B, Wolzt M,
       Diabetes Care 15:815– 819, 1992                     paired fasting glucose or diabetes: the             Schmetterer L: LPS-induced microvas-
 55.   Cugati S, Mitchell P, Wang JJ: Do reti-             Rotterdam Study. Diabetes 55:506 –510,              cular leukocytosis can be assessed by
       nopathy signs in non-diabetic individu-             2006                                                blue-field entoptic phenomenon. Am J
       als predict the subsequent risk of            67.   Wong TY, Duncan BB, Golden SH, Klein                Physiol Heart Circ Physiol 287:H691–
       diabetes? Br J Ophthalmol 90:928 –929,              R, Couper DJ, Klein BE, Hubbard LD,                 H694, 2004
       2006                                                Sharrett AR, Schmidt MI: Associations         79.   Wilkinson-Berka JL: Vasoactive factors
 56.   Wong TY, Mohamed Q, Klein R, Couper                 between the metabolic syndrome and                  and diabetic retinopathy: vascular endo-
       DJ: Do retinopathy signs in non-diabetic            retinal microvascular signs: the Athero-            thelial growth factor, cycoloxygenase-2
       individuals predict the subsequent risk             sclerosis Risk In Communities study. In-            and nitric oxide. Curr Pharm Des
       of diabetes? Br J Ophthalmol 90:301–                vest Ophthalmol Vis Sci 45:2949 –2954,              10:3331–3348, 2004
       303, 2006                                           2004                                          80.   Chester AH, Borland JA, Buttery LD,
 57.   Klein R, Klein BEK, Mos SE, Wong TY:          68.   Wang JJ, Taylor B, Wong TY, Chua B,                 Mitchell JA, Cunningham DA, Hafizi S,
       The relationship of retinopathy in per-             Rochtchina E, Klein R, Mitchell P: Reti-            Hoare GS, Springall DR, Polak JM, Ya-
       sons without diabetes to the 15-year in-            nal vessel diameters and obesity: a pop-            coub MH: Induction of nitric oxide syn-
       cidence of diabetes and hypertension:               ulation-based study in older persons.               thase in human vascular smooth muscle:
       Beaver Dam Eye Study. Trans Am Oph-                 Obesity (Silver Spring) 14:206 –214,                interactions between proinflammatory
       thalmol Soc 104:98 –107, 2006                       2006                                                cytokines. Cardiovasc Res 38:814 – 821,
 58.   Wong TY, Klein R, Sharrett AR, Duncan         69.   Cheung N, Saw SM, Islam FM, Rogers                  1998
       BB, Couper DJ, Klein BE, Hubbard LD,                SL, Shankar A, de Haseth K, Mitchell P,       81.   de Rekeneire N, Peila R, Ding J, Colbert
       Nieto FJ: Retinal arteriolar diameter and           Wong TY: BMI and retinal vascular cali-             LH, Visser M, Shorr RI, Kritchevsky SB,
       risk for hypertension. Ann Intern Med               ber in children. Obesity (Silver Spring)            Kuller LH, Strotmeyer ES, Schwartz AV,
       140:248 –255, 2004                                  15:209 –215, 2007                                   Vellas B, Harris TB: Diabetes, hypergly-
 59.   Wong TY, Rosamond W, Chang PP,                70.   Klein R, Klein BE, Moss SE, Wong TY,                cemia, and inflammation in older indi-
       Couper DJ, Sharrett AR, Hubbard LD,                 Sharrett AR: Retinal vascular caliber in            viduals: the health, aging and body
       Folsom AR, Klein R: Retinopathy and                 persons with type 2 diabetes: the Wis-              composition study. Diabetes Care 29:
       risk of congestive heart failure. JAMA              consin Epidemiological Study of Dia-                1902–1908, 2006
       293:63– 69, 2005                                    betic Retinopathy: XX. Ophthalmology          82.   Ling PR, Mueller C, Smith RJ, Bistrian
 60.   Wong TY, Coresh J, Klein R, Muntner P,              113:1488 –1498, 2006                                BR: Hyperglycemia induced by glucose
       Couper DJ, Sharrett AR, Klein BE, Heiss       71.   Klein R, Klein BE, Moss SE, Wong TY,                infusion causes hepatic oxidative stress
       G, Hubbard LD, Duncan BB: Retinal mi-               Hubbard L, Cruickshanks KJ, Palta M:                and systemic inflammation, but not
       crovascular abnormalities and renal dys-            Retinal vascular abnormalities in per-              STAT3 or MAP kinase activation in liver
       function: the atherosclerosis risk in               sons with type 1 diabetes: the Wisconsin            in rats. Metabolism 52:868 – 874, 2003
       communities study. J Am Soc Nephrol 15:             Epidemiologic Study of Diabetic Reti-         83.   Nguyen TT, Wong TY: Retinal vascular
       2469 –2476, 2004                                    nopathy: XVIII. Ophthalmology 110:                  manifestations of metabolic disorders.
 61.   Edwards MS, Wilson DB, Craven TE,                   2118 –2125, 2003                                    Trends Endocrinol Metab 17:262–268,
       Stafford J, Fried LF, Wong TY, Klein R,       72.   Wong TY, Shankar A, Klein R, Klein BE:              2006
       Burke GL, Hansen KJ: Associations be-               Retinal vessel diameters and the inci-        84.   Tso MO, Jampol LM: Pathophysiology of
       tween retinal microvascular abnormali-              dence of gross proteinuria and renal in-            hypertensive retinopathy. Ophthalmol-
       ties and declining renal function in the            sufficiency in people with type 1                   ogy 89:1132–1145, 1982
       elderly population: the Cardiovascular              diabetes. Diabetes 53:179 –184, 2004          85.   Walsh JB: Hypertensive retinopathy: de-
       Health Study. Am J Kidney Dis 46:214 –        73.   Grunwald JE, Riva CE, Baine J, Brucker              scription, classification, and prognosis.
       224, 2005                                           AJ: Total retinal volumetric blood flow             Ophthalmology 89:1127–1131, 1982
 62.   Rema M, Mohan V, Deepa R, Ravikumar                 rate in diabetic patients with poor glyce-    86.   Leung H, Wang JJ, Rochtchina E, Wong
       R: Association of carotid intima-media              mic control. Invest Ophthalmol Vis Sci 33:          TY, Klein R, Mitchell P: Impact of cur-
       thickness and arterial stiffness with dia-          356 –363, 1992                                      rent and past blood pressure on retinal
       betic retinopathy: the Chennai Urban          74.   Saldivar E, Cabrales P, Tsai AG, Intagli-           arteriolar diameter in an older popula-
       Rural Epidemiology Study (CURES-2).                 etta M: Microcirculatory changes during             tion. J Hypertens 22:1543–1549, 2004
       Diabetes Care 27:1962–1967, 2004                    chronic adaptation to hypoxia. Am J           87.   Wong TY, Hubbard LD, Klein R, Marino
 63.   Wong TY, Klein R, Islam FM, Cotch MF,               Physiol Heart Circ Physiol 285:H2064 –              EK, Kronmal R, Sharrett AR, Siscovick
       Folsom AR, Klein BE, Sharrett AR, Shea              H2071, 2003                                         DS, Burke G, Tielsch JM: Retinal micro-
       S: Diabetic retinopathy in a multi-ethnic     75.   Caballero AE: Metabolic and vascular                vascular abnormalities and blood pres-
       cohort in the United States. Am J Oph-              abnormalities in subjects at risk for type          sure in older people: the Cardiovascular
       thalmol 141:446 – 455, 2006                         2 diabetes: the early start of a dangerous          Health Study. Br J Ophthalmol 86:1007–
 64.   Wong TY, Klein R, Sharrett AR, Schmidt              situation. Arch Med Res 36:241–249,                 1013, 2002
       MI, Pankow JS, Couper DJ, Klein BE,                 2005                                          88.   Wong TY, Shankar A, Klein R, Klein BE,
       Hubbard LD, Duncan BB: Retinal arte-          76.   Klein R, Klein BE, Knudtson MD, Wong                Hubbard LD: Prospective cohort study
       riolar narrowing and risk of diabetes               TY, Tsai MY: Are inflammatory factors               of retinal vessel diameters and risk of hy-
       mellitus in middle-aged persons. JAMA               related to retinal vessel caliber? The Bea-         pertension. BMJ 329:79, 2004
       287:2528 –2533, 2002                                ver Dam Eye Study. Arch Ophthalmol            89.   Smith W, Wang JJ, Wong TY, Roch-
 65.   Wong TY, Shankar A, Klein R, Klein BE,              124:87–94, 2006                                     tchina E, Klein R, Leeder SR, Mitchell P:
       Hubbard LD: Retinal arteriolar narrow-        77.   Tamai K, Matsubara A, Tomida K, Mat-                Retinal arteriolar narrowing is associated
       ing, hypertension, and subsequent risk              suda Y, Morita H, Armstrong D, Ogura                with 5-year incident severe hyperten-

2714                                                                                                DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007
Nguyen, Wang, and Wong

       sion: the Blue Mountains Eye Study. Hy-             SA, Hughes AD: Arteriolar length-diam-               the JNC 7 report. JAMA 289:2560 –
       pertension 44:442– 447, 2004                        eter (L:D) ratio: a geometric parameter of           2572, 2003
 90.   Ikram MK, Witteman JC, Vingerling JR,               the retinal vasculature diagnostic of hy-     109.   Hsia J, Margolis KL, Eaton CB, Wenger
       Breteler MM, Hofman A, de Jong PT:                  pertension. J Hum Hypertens 10:417–                  NK, Allison M, Wu L, LaCroix AZ, Black
       Retinal vessel diameters and risk of hy-            418,1996                                             HR: Prehypertension and cardiovascular
       pertension: the Rotterdam Study. Hyper-      101.   Stanton AV, Wasan B, Cerutti A, Ford S,              disease risk in the Women’s Health Ini-
       tension 47:189 –194, 2006                           Marsh R, Sever PP, Thom SA, Hughes                   tiative. Circulation 115:855– 860, 2007
 91.   Mulvany MJ: Are vascular abnormalities              AD: Vascular network changes in the           110.   Cheung N, Islam FM, Saw SM, Shankar
       a primary cause or secondary conse-                 retina with age and hypertension. J Hy-              A, de Haseth K, Mitchell P, Wong TY:
       quence of hypertension? Hypertension                pertens 13:1724 –1728, 1995                          Distribution and associations of retinal
       18:I52–I57, 1991                             102.   Kutschbach P, Wolf S, Sieveking M, Ittel             vascular caliber with ethnicity, gender,
 92.   Xing C, Klein BE, Klein R, Jun G, Lee KE,           TH, Schulte K, Reim M: Retinal capillary             and birth parameters in young children.
       Iyengar SK: Genome-wide linkage study               density in patients with arterial hyper-             Invest Ophthalmol Vis Sci 48:1018 –1024,
       of retinal vessel diameters in the Beaver           tension: 2-year follow-up. Graefes Arch              2007
       Dam Eye Study. Hypertension 47:797–                 Clin Exp Ophthalmol 236:410 – 414,            111.   Gould DB, Phalan FC, van Mil SE, Sund-
       802, 2006                                           1998                                                 berg JP, Vahedi K, Massin P, Bousser
 93.   Wang JJ, Wong TY: Genetic determi-           103.   Witt N, Wong TY, Hughes AD, Chatur-                  MG, Heutink P, Miner JH, Tournier-
       nants of retinal vascular caliber: addi-            vedi N, Klein BE, Evans R, McNamara                  Lasserve E, John SW: Role of COL4A1 in
       tional insights into hypertension patho-            M, Thom SA, Klein R: Abnormalities of                small-vessel disease and hemorrhagic
       genesis. Hypertension 47:644 – 645,2006             retinal microvascular structure and risk             stroke. N Engl J Med 354:1489 –1496,
 94.   Ingelfinger JR: Pediatric antecedents of            of mortality from ischemic heart disease             2006
       adult cardiovascular disease–awareness              and stroke. Hypertension 47:975–981,          112.   Gekeler F, Shinoda K, Junger M, Bartz-
       and intervention. N Engl J Med 350:                 2006                                                 Schmidt KU, Gelisken F: Familial retinal
       2123–2126, 2004                              104.   Sorkin JD, Muller DC, Fleg JL, Andres R:             arterial tortuosity associated with tortu-
 95.   Barker DJ, Osmond C, Golding J, Kuh D,              The relation of fasting and 2-h postchal-            osity in nail bed capillaries. Arch Oph-
       Wadsworth ME: Growth in utero, blood                lenge plasma glucose concentrations to               thalmol 124:1492–1494, 2006
       pressure in childhood and adult life, and           mortality: data from the Baltimore Lon-       113.   Muiesan ML, Grassi G: Assessment of
       mortality from cardiovascular disease.              gitudinal Study of Aging with a critical             retinal vascular changes in hyperten-
       BMJ 298:564 –567, 1989                              review of the literature. Diabetes Care 28:          sion: new perspectives. J Hypertens 24:
 96.   Ikram MK, De Jong FJ, Van Dijk EJ, Prins            2626 –2632, 2005                                     813– 814, 2006
       ND, Hofman A, Breteler MM, De Jong           105.   Brunner EJ, Shipley MJ, Witte DR, Fuller      114.   Wong TY, Klein R, Nieto FJ, Klein BE,
       PT: Retinal vessel diameters and cerebral           JH, Marmot MG: Relation between                      Sharrett AR, Meuer SM, Hubbard LD,
       small vessel disease: the Rotterdam Scan            blood glucose and coronary mortality                 Tielsch JM: Retinal microvascular ab-
       Study. Brain 129:182–188, 2006                      over 33 years in the Whitehall Study.                normalities and 10-year cardiovascular
 97.   Ikram MK, de Jong FJ, Bos MJ, Vinger-               Diabetes Care 29:26 –31, 2006                        mortality: a population-based case-con-
       ling JR, Hofman A, Koudstaal PJ, de Jong     106.   Glucose tolerance and mortality: com-                trol study. Ophthalmology 110:933–940,
       PT, Breteler MM: Retinal vessel diame-              parison of WHO and American Diabetes                 2003
       ters and risk of stroke: the Rotterdam              Association diagnostic criteria: the DE-      115.   Wong TY, Klein R, Sharrett AR, Couper
       Study. Neurology 66:1339 –1343, 2006                CODE study group: European Diabetes                  DJ, Klein BE, Liao DP, Hubbard LD,
 98.   Wong TY, Kamineni A, Klein R, Sharrett              Epidemiology Group: Diabetes Epide-                  Mosley TH: Cerebral white matter le-
       AR, Klein BE, Siscovick DS, Cushman                 miology: Collaborative analysis Of Diag-             sions, retinopathy, and incident clinical
       M, Duncan BB: Quantitative retinal                  nostic criteria in Europe. Lancet 354:               stroke. JAMA 288:67–74, 2002
       venular caliber and risk of cardiovascu-            617– 621, 1999                                116.   Wong TY, Mosley TH Jr, Klein R, Klein
       lar disease in older persons: the cardio-    107.   Gu D, Wildman RP, Wu X, Reynolds K,                  BE, Sharrett AR, Couper DJ, Hubbard
       vascular health study. Arch Intern Med              Huang J, Chen CS, He J: Incidence and                LD: Retinal microvascular changes and
       166:2388 –2394, 2006                                predictors of hypertension over 8 years              MRI signs of cerebral atrophy in healthy,
 99.   Hughes AD, Martinez-Perez E, Jabbar                 among Chinese men and women. J Hy-                   middle-aged people. Neurology 61:806 –
       AS, Hassan A, Witt NW, Mistry PD,                   pertens 25:517–523, 2007                             811, 2003
       Chapman N, Stanton AV, Beevers G,            108.   Chobanian AV, Bakris GL, Black HR,            117.   Wong TY, Klein R, Sharrett AR, Nieto FJ,
       Pedrinelli R, Parker KH, Thom SA:                   Cushman WC, Green LA, Izzo JL Jr,                    Boland LL, Couper DJ, Mosley TH, Klein
       Quantification of topological changes in            Jones DW, Materson BJ, Oparil S,                     BE, Hubbard LD, Szklo M: Retinal mi-
       retinal vascular architecture in essential          Wright JT Jr, Roccella EJ: The Seventh               crovascular abnormalities and cognitive
       and malignant hypertension. J Hypertens             Report of the Joint National Committee               impairment in middle-aged persons: the
       24:889 – 894, 2006                                  on Prevention, Detection, Evaluation,                Atherosclerosis Risk in Communities
100.   King LA, Stanton AV, Sever PS, Thom                 and Treatment of High Blood Pressure:                Study. Stroke 33:1487–1492, 2002

DIABETES CARE, VOLUME 30, NUMBER 10, OCTOBER 2007                                                                                                   2715
You can also read