Free
Clinical and Epidemiologic Research  |   January 2013
Phenotypic and Genetic Correlation of Blood Pressure and Body Mass Index with Retinal Vascular Caliber in Children and Adolescents: The Guangzhou Twin Eye Study
Author Affiliations & Notes
  • Yingfeng Zheng
    From the State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China; and the
  • Wenyong Huang
    From the State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China; and the
  • Jian Zhang
    From the State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China; and the
  • Mingguang He
    From the State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China; and the
  • Corresponding author: Mingguang He, Zhongshan Ophthalmic Center, Guangzhou 510060, People's Republic of China; mingguang_he@yahoo.com
Investigative Ophthalmology & Visual Science January 2013, Vol.54, 423-428. doi:10.1167/iovs.12-9543
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Yingfeng Zheng, Wenyong Huang, Jian Zhang, Mingguang He; Phenotypic and Genetic Correlation of Blood Pressure and Body Mass Index with Retinal Vascular Caliber in Children and Adolescents: The Guangzhou Twin Eye Study. Invest. Ophthalmol. Vis. Sci. 2013;54(1):423-428. doi: 10.1167/iovs.12-9543.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: To examine the phenotypic and genetic associations of blood pressure and body mass index (BMI) with retinal vascular caliber.

Methods.: A total of 657 monozygotic and 378 dizygotic twin pairs aged 7 to 19 years were recruited from the Guangzhou Twin Registry. All twins underwent digital retinal photography and measurement of retinal vascular caliber. The genetic correlations between the traits were estimated by applying a multivariate Cholesky model.

Results.: In traditional regression analyses, participants with a higher mean arterial pressure (MAP) and a higher BMI were significantly more likely to have narrower retinal arterioles, whereas participants with a higher BMI level were more likely to have a wider retinal venule (all P < 0.001). In multivariate Cholesky models, only 1% to 2% of the phenotypic variation in retinal arteriole was shared with those in MAP and BMI, although the majority of these phenotypic variations were explained by shared genetic components. The phenotypic variation in retinal venule was not shared with those in MAP and BMI.

Conclusions.: Retinal vascular caliber is significantly but weakly associated with MAP and BMI in children and young adolescents. These phenotypic correlations are mainly attributable to genetic components.

Introduction
Hypertension and obesity are the major risk factors for cardiovascular disease and associated mortality in middle aged and older people. Emerging evidence suggests that both elevated blood pressure and high body mass index (BMI) can lead to structural damage in microvasculature (e.g., microvascular rarefaction, narrow retinal arterioles), which may predispose individuals to the development of cardiovascular disease. 1 However, the mechanisms linking increased blood pressure and high BMI to microvascular damage are not fully understood. 25  
Advanced retinal photographic technologies and new imaging analyses have now allowed for direct and noninvasive measurement of the microvasculature. 6 A large body of evidence from adult populations shows that retinal vascular caliber is an early marker of microvascular damage associated with long term hypertension, 4 endothelial dysfunction, 7 and other vascular disorders in the peripheral circulation. 8 However, very few data exist regarding the effect of blood pressure and BMI on retinal vascular caliber among children and adolescents. 912  
This study used data from a Chinese twin cohort aged 7 to 19 years. We first assessed the phenotypic associations of blood pressure and BMI with retinal vascular caliber and determined whether there was an interaction between blood pressure and BMI. We then examined the extent to which these phenotypic associations were explained by shared genetic influences. Children and adolescents are ideal subjects in which to examine the effects of blood pressure and BMI on retinal vascular caliber, since young people are generally free of confounding effects caused by other systemic or ocular vascular diseases (e.g., diabetes and diabetic retinopathy) and medication (e.g., antihypertensive medication). 
Methods
Study Population
The study participants were recruited from the Guangzhou Twin Registry, a population-based registry established in Guangzhou, China, and described in detail previously. 13 Briefly, all twins born in Guangzhou between 1987 and 2000 were identified using an official household registry followed by door-to-door verification. 13 Data collection began in 2006. In 2009, 1043 twin pairs aged 9 to 19 years as of July 1, 2009 participated in an annual examination, providing cross-sectional data for blood pressure, height, weight, and retinal vascular caliber. Eight twin pairs with medical (e.g., cerebral palsy) or chronic eye conditions (e.g., cataract, retinopathy of prematurity), missing data, or ungradable retinal photographs were excluded. 
The zygosity of each of the same-sex twin pairs was determined using 16 multiplex STR markers (PowerPlex 16 system; Promega, Madison, WI) 14 at the Forensic Medicine Department of Sun Yat-Sen University. Opposite-sex twin pairs were deemed to be dizygotic without genotyping. 
Written informed consent was obtained either from parents or legal guardians of the twins following a detail explanation of the study. Ethical approval was obtained from the Zhongshan University ethical review board and the ethics committee of the Zhongshan Ophthalmic Center, and the study was conducted in accordance with the Tenets of the World Medical Association's Declaration of Helsinki. 
Retinal Photography and Measurement of Retinal Vascular Caliber
Twins were examined at Zhongshan Ophthalmic Center in 2009. Digital retinal photographs centered on the optic disc were taken using a digital retinal camera (Kowa Non Myd 7 digital fundus imaging system; Kowa, Tokyo, Japan) using standardized settings, after pupil dilation with cyclopentolate 1%. The method used to measure and summarize retinal vascular caliber from digitized retinal photographs has been published previously. 6 In brief, based on computer-assisted software (IVAN; University of Wisconsin, Madison, WI) and a standardized protocol provided by the Retinal Vascular Imaging Centre, University of Melbourne, a trained grader, masked to participant characteristics and blood pressure measurement, measured the caliber of all the retinal arterioles and venules coursing through a specified zone 0.5- to 1-disc diameter away from the optic disc margin. These measurements were summarized as central retinal arteriolar equivalent (CRAE) and venular equivalent (CRVE) in microns, representing the average retinal arteriolar and venular caliber, respectively. 6 Fifty retinal photographs were also remeasured 4 weeks apart. The intraclass correlation coefficients were greater than 0.90 for both retinal arteriolar caliber and venular caliber. 
Blood Pressure Measurement
An automated digital oscillometric sphygmomanometer (Model HEM-907; Omron Healthcare, Matsuzaka, Japan) with appropriate cuff size (bladder length approximately 80% and width at least 40% of the arm circumference, covering the upper arm, but not obscuring the antecubital fossa) was used for measuring blood pressure in the right arm of seated twins, after 5 minutes of quiet rest, by a single trained examiner. Systolic and diastolic blood pressure were determined by the first and fifth Korotkoff sounds, respectively. Blood pressure was measured twice, and the average of two measurements was recorded. Mean arterial blood pressure (MAP) was calculated as 1/3 of the systolic plus 2/3 of the diastolic blood pressure. 
Phenotypic Analysis
Retinal vascular caliber measurements of the right and left eye have been shown to be highly correlated. 6 Thus, the right eye was selected to represent the phenotypic characteristics of subjects in all analyses. We performed standard linear regression with “twins as individuals” to assess the association of blood pressure and BMI with retinal vascular caliber (as a dependent variable) using generalized estimating equation (GEE) modeling. Blood pressure was also categorized as normotension, prehypertension, or hypertension. Prehypertension was defined as a blood pressure greater than or equal to the 90th percentile for systolic or diastolic blood pressure, but below the threshold for hypertension (≥95th percentile), 15 according to age- and sex-specific blood pressure reference values for Chinese persons. 16 BMI was also categorized as normal BMI, overweight, or obesity. Overweight was defined as BMI greater than or equal to the 85th percentile, but below the threshold for obesity (≥95th percentile), 17 according to Chinese age- and sex-specific BMI reference values. 18  
We also estimated twin correlations and cross-trait cross-twin correlations for each of the four phenotypes (MAP, BMI, retinal arteriolar caliber, and retinal venular caliber) separately in monozygotic and dizygotic twin pairs. Twin correlation refers to the correlation of a single phenotype within a twin pair, whereas cross-trait cross-twin correlation refers to the correlation across two phenotypes within a twin pair. A higher cross-trait cross-twin correlation in monozygotic than dizygotic twins is indicative of shared genetic factors for covariance between two phenotypes. 
Genetic Analysis
In classic twin studies, monozygotic twins share 100% of their genetic backgrounds, while dizygotic twins share only 50%. 19 It is also assumed that common environmental effects are 100% shared by members of both monozygotic and dizygotic pairs. 19 Based on these assumptions, the phenotypic variance is separated into additive genetic (A), common environmental (C), and unshared environmental (E) components. The A components represent the sum effect of all alleles, shared 100% within monozygotic twins and 50% within dizygotic twins. The C components represent environmental influences (e.g., diet and socioeconomic status) that are assumed to be shared 100% within both monozygotic and dizygotic twin pairs. The E components represent environmental exposures (e.g., accident, virus infection) or measurement errors unique to each member of a twin pair. Model fitting was performed using the Mx statistical program (http://www.vcu.edu/mx). Reduced models were constructed by removing a specific parameter (A, C, or E), and then comparing the result with the full ACE model. Parameters were removed from the full model if the removal did not result in a significant deterioration of the model fit. The model with the lowest Akaike Information Criteria (AIC) was chosen as the best fitting one. 19  
We then performed a multivariate genetic model fitting by constructing a Cholesky decomposition model with age correction. Age and the first set of factor loadings (i.e., A1, C1, and E1) have impacts on all the 4 phenotypes (MAP, BMI, retinal arteriolar caliber, and retinal venular caliber). The second set of factor loadings (i.e., A2, C2, and E2) has impacts on MAP, retinal arteriolar caliber, and retinal venular caliber, with the third set (i.e., A3, C3, and E3) on retinal arteriolar caliber and venular caliber, and the final set (i.e., A4, C4, and E4) on retinal venular caliber only. We performed genetic modeling analysis by removing each factor loading from the full Cholesky model, leading to a submodel with fewer parameters. The model with the lowest AIC reflects the best balance between goodness of fit and parsimony. We also performed a power calculation by assuming that there are two phenotypes (for the first phenotype: A1 = 0.50, C1 = 0.05, E1 = 0.45; for the second phenotype: A2 = 0.65, C2 = 0.10, E2 = 0.25). Based on the current sample size, we would have a power of 98% to detect a significant genetic correlation and a power of 20% to detect a significant common environment correlation between the two phenotypes. 
Results
Descriptive Statistics
This study included 1035 twin pairs (657 monozygotic and 378 dizygotic pairs), for whom all relevant data were available (Table 1). For all the phenotypes, no significant difference (P > 0.05) in mean or SD was found between the two zygosity groups, fulfilling the basic assumptions for classic twin analyses. No significant differences (P > 0.05) in mean or SD were found between the first born and second born twins, suggesting no birth order effects on these variables. 
Table 1. 
 
Baseline Characteristics of the Twin Individual by Zygosity
Table 1. 
 
Baseline Characteristics of the Twin Individual by Zygosity
All Twin Individuals (N = 2070) Monozygotic Twin (N = 1314) Dizygotic Twin (N = 756)
Age, y 2070 12.7 (3.2) 12.8 (3.1) 12.4 (3.2)
Sex 2070 1072 (51.8) 708 (53.9) 364 (48.2)
BMI, cm/kg2 2070 17.9 (3.1) 17.9 (3.0) 18.0 (3.3)
MAP, mm Hg 2070 75.0 (10.4) 75.0 (10.5) 74.9 (10.2)
CRAE (μm)
 All persons 2070 150.1 (13.4) 149.8 (13.6) 150.5 (12.9)
 Persons with normotension 1772 150.6 (13.4) 150.4 (13.6) 150.9 (13.0)
 Persons with prehypertension 150 148.2 (13.1) 147.5 (13.6) 149.6 (12.3)
 Persons with hypertension 148 146.0 (12.6) 145.6 (13.2) 146.7 (11.3)
 Persons with normal weight 1851 150.3 (13.3) 150.0 (13.5) 150.9 (12.9)
 Persons with overweight 162 148.3 (14.0) 148.5 (14.5) 148.0 (13.5)
 Persons with obesity 57 147.1 (12.4) 145.1 (13.7) 148.6 (11.3)
CRVE (μm)
 All persons 2070 218.4 (19.1) 218.3 (19.3) 218.6 (18.7)
 Persons with normotension 1772 218.4 (19.0) 218.3 (19.2) 218.6 (18.7)
 Persons with prehypertension 150 216.0 (19.6) 215.8 (20.5) 216.3 (17.9)
 Persons with hypertension 148 221.0 (19.7) 221.1 (19.4) 220.8 (20.4)
 Persons with normal weight 1851 218.1 (19.3) 218.1 (19.6) 218.0 (18.7)
 Persons with overweight 162 221.1 (16.8) 220.8 (15.7) 221.4 (18.3)
 Persons with obesity 57 221.8 (19.4) 219.2 (17.9) 223.7 (20.5)
Phenotypic Analysis
In GEE regression models controlling for age and sex, both MAP and BMI were significantly associated with retinal arteriolar caliber as a dependent variable. The significant associations persisted after further adjustment for BMI (in the models for MAP), MAP (in the models for BMI), spherical equivalent, axial length, and retinal venular caliber. Retinal arteriolar caliber decreased by 1.48 μm for each 10-mm Hg increase in MAP, and by 0.30 μm for each kg/m2 increase in BMI (Figure). The presence of prehypertension (β = 2.06, P = 0.01), hypertension (β = 2.17, P = 0.02), overweight (β = 2.24, P = 0.01), or obesity (β = 2.76, P = 0.04) was associated with narrower retinal arteriole. 
Figure. 
 
Linear regression analysis of retinal vascular calibers with MAP and BMI. The linear regression models are adjusted for age, sex, birth order, zygosity, axial length, spherical equivalent, MAP (for the models for BMI), BMI (for the models for MAP), and the fellow retinal vascular caliber.
Figure. 
 
Linear regression analysis of retinal vascular calibers with MAP and BMI. The linear regression models are adjusted for age, sex, birth order, zygosity, axial length, spherical equivalent, MAP (for the models for BMI), BMI (for the models for MAP), and the fellow retinal vascular caliber.
In the GEE regression models controlling for age and sex, BMI, but not MAP, was associated with retinal venular caliber (P < 0.001). After further adjustment for MAP, spherical equivalent, axial length, and retinal arteriolar caliber, each kilogram per meter squared increase in BMI corresponded to a 0.43-μm increase in retinal venular caliber. After adjusting for age, sex, and other confounding variables, the presence of being overweight was associated with wider retinal venules (β = 2.81, P = 0.02), but the presence of obesity was not statistically significantly associated with retinal venular caliber (β = 3.75, P = 0.08). 
Table 2 shows the phenotypic correlations, twin correlations (the elements in the rectangles on the diagonals), and cross-twin cross-trait correlations (the off-diagonal elements) for age, MAP, BMI, retinal arteriolar caliber, and venular caliber, stratified by sex. The phenotypic correlations among age, MAP, BMI, retinal arteriolar caliber, and venular caliber were significant, ranging from −0.06 to 0.66. Monozygotic twin correlations were significantly greater than dizygotic twin correlations, indicating substantial genetic influences on covariance between MAP, BMI, retinal arteriolar caliber, and venular caliber. For all the four phenotypes, the cross-twin cross-trait correlations in monozygotic twins were greater than in dizygotic twins, consistent with the view that shared genetic factors played an important role in the covariations among the four phenotypes. 
Table 2. 
 
Phenotypic Correlations in MZ and DZ Twins
Table 2. 
 
Phenotypic Correlations in MZ and DZ Twins
All Male Female
AGE BMI MAP CRAE CRVE AGE BMI MAP CRAE CRVE AGE BMI MAP CRAE CRVE
Phenotypic correlations
N = 2070 twin N = 998 male twin N = 1072 female twin
AGE 1.0
BMI 0.66 1.0 0.66 1.0 0.68 1.0
MAP 0.39 0.46 1.0 0.47 0.53 1.0 0.31 0.37 1.0
CRAE –0.18 –0.20 –0.19 1.0 –0.22 –0.22 –0.20 1.0 –0.15 –0.15 –0.17 1.0
CRVE –0.15 –0.08 –0.06 0.57 1.0 –0.16 –0.04 –0.06 0.58 1.0 –0.15 –0.09 –0.04 0.55 1.0
Twin correlations and cross-twin cross-trait correlations
N = 1035 male pairs N = 401 male pairs N = 438 female pairs
AGE BMI1 MAP1 CRAE1 CRVE1 AGE BMI1 MAP1 CRAE1 CRVE1 AGE BMI1 MAP1 CRAE1 CRVE1
MZ twins
N = 657 male pairs N = 303 male pairs N = 354 female pairs
AGE 1.0 0.69 0.44 –0.21 –0.16 1.0 0.69 0.51 –0.26 –0.15 1.0 0.70 0.38 –0.17 –0.17
BMI2 0.69 0.94 0.48 –0.24 –0.10 0.71 0.94 0.57 –0.25 –0.13 0.70 0.93 0.39 –0.20 –0.11
MAP2 0.33 0.42 0.57 –0.18 –0.03* 0.50 0.55 0.66 –0.19 –0.05 0.20 0.28 0.48 –0.16 –0.01
CRAE2 –0.18 –0.20 –0.19 0.63 0.49 –0.25 –0.22 –0.20 0.59 0.46 –0.14 –0.14 –0.16 0.64 0.50
CRVE2 –0.19 –0.12 –0.11 0.47 0.68 –0.22 –0.06 –0.09 0.46 0.71 –0.17 –0.14 –0.12 0.47 0.65
DZ twins
N = 378 male pairs N = 98 male pairs N = 184 female pairs
AGE 1.0 0.63 0.42 –0.17 –0.16 1.0 0.57 0.34 –0.11 –0.08 1.0 0.63 0.42 –0.17 –0.16
BMI2 0.62 0.69 0.37 –0.14 –0.01* 0.56 0.65 0.32 –0.12 –0.08* 0.62 0.72 0.37 –0.12 –0.01*
MAP2 0.35 0.28 0.38 0.01* 0.03* 0.32 0.21 0.29 0.05* 0.04* 0.35 0.28 0.44 0.01* 0.03*
CRAE2 –0.12 –0.09* –0.01* 0.28 0.20 –0.15 –0.12* –0.08* 0.34 0.26 –0.12 –0.09* –0.01* 0.39 0.20
CRVE2 –0.05 –0.02* 0.06* 0.21 0.34 –0.03 –0.07* 0.07* 0.24 0.46 –0.05 –0.02* 0.06* 0.21 0.33
Multivariate Genetic Analysis
The sex-limitation analysis for the full Cholesky model showed that there was no evidence of sex differences in covariance components (difference in X2[1] = 1.52, P = 0.22). Thus, male and female groups were combined in the following analyses. The multivariate model (see Supplementary Material and Supplementary Figure S1) fitting process is shown in Supplementary Table 1 (see Supplementary Material and Supplementary Table S1). Removing the C components from the full ACE model did not result in a significant deterioration in fit (model 2, P = 0.15), while removing the influence of age correction led to a significant reduction in fit (model 3, P < 0.001). Removing the additive genetic covariance between MAP and retinal venular caliber and between BMI and retinal venular caliber, and removing the nonshared environmental covariance of BMI with the other three phenotypes did not significantly worsen the fit. We did not remove random environmental variances unique to MAP, BMI, retinal arteriolar caliber, and venular caliber, as these variances were confounded with measurement error. 
The sources of variance in retinal arteriolar caliber are shown in the upper part of Table 3. The last column provides the amount of variance in retinal arteriolar caliber attributed to age, MAP, BMI, and retinal venular caliber. Each category of variance is further separated into additive genetic and nonshared environmental factors. In the best-fitting multivariate model, 58.5% of the variance in retinal arteriolar caliber was determined by genetic factors, 38.5% by nonshared environmental factors, and 3% by age influences. Of the 58.5% genetic variance, 1.0% could be explained by genes shared with MAP, and 1.6% by genes shared with BMI. Despite the low phenotypic correlation, 83.3% (h2/[h2 + E2] = 1%/[1% + 0.2%]) of the phenotypic correlation between MAP and retinal arteriolar caliber was attributable to shared genetic factors, and the remaining 16.7% was attributed to non-shared environmental factors. The phenotypic correlation between BMI and retinal arteriolar caliber was entirely attributable to shared genetic factors. The sources of variance in retinal venular caliber are shown in the lower part of Table 3. Approximately 65.1% of the variance in retinal venular caliber could be explained by genetic influences, of which 35.4% was shared with genes for retinal arteriolar caliber. MAP and BMI had no genetic sharing with retinal venular caliber. 
Table 3. 
 
Sources of Variance in CRAE and CRVE and Their Correlations Based on the Best-Fitting Multivariate Genetic Model
Table 3. 
 
Sources of Variance in CRAE and CRVE and Their Correlations Based on the Best-Fitting Multivariate Genetic Model
h2 (%)* e2 (%)* Age2 (%)* Total (%)*
Sources of variance for CRAE
 Age 3.0 3.0
 BMI 1.6 1.6
 MAP 1.0 0.2 1.2
 CRVE 35.4 2.8 38.2
 CRAE specific 20.5 35.5 56.0
 CRAE total 58.5 38.5 3.0 100
Sources of variance for CRVE
 Age 2.0 2.0
 BMI
 MAP 0.0 0.0
 CRAE 35.4 2.8 38.2
 CRVE specific 29.7 30.1 59.8
 CRVE total 65.1 32.9 2.0 100
We also performed supplementary analyses by adjusting for ocular magnification effect on retinal vascular caliber measurement. These analyses were performed on the basis of the Bengtsson formula: uncorrected retinal vascular caliber × (1 − 0.017 × spherical equivalent refraction) = corrected retinal vascular caliber. Adjustment for refractive error did not attenuate the phenotypic association revealed previously, and it only slightly attenuated the heritability estimates: the heritability estimate was 51% (95% confidence interval [CI]: 46%–56%) for refraction-adjusted retinal arteriolar caliber and 57% (95% CI: 52%–62%) for refraction-adjusted retinal venular caliber. This slight attenuation may be due to the adjustment of genetic effect for refractive errors between siblings, or simply due to the measurement errors in refraction. The refraction adjustment did not alter the phenotypic and genetic correlations with retinal arteriolar caliber (data not shown). 
Discussion
This twin study documented three major findings. First, we confirmed a high heritability estimate for MAP, BMI, and retinal vascular caliber, consistent with previous twin studies. 20,21 These high heritability estimations appear to suggest that genetic variation accounts for a relatively high proportion of phenotypic variation in these traits, a finding being subject to the equal environment assumption in analysis of twin study. 
Second, there were significant (although very low) phenotypic correlations of MAP, BMI, and retinal microvasculature in children and adolescents. The significant phenotypic association between MAP and retinal arteriolar caliber is consistent with those from two early population-based studies, one in Australia (among 12 year olds) and the other in Singapore (among children aged 7–9 years), that children with higher blood pressure are at risk of having narrower retinal arteriole. 912 Consistently, we also identified a significant association between narrower retinal arteriole and the presence of prehypertension, suggesting that hypertensive microvascular changes may occur in young people whose blood pressure levels were previously considered within the normal range. 15 Our study also confirms the finding among Australian children that higher BMI is associated with both a narrower retinal arteriole and a wider retinal venule, 912 supporting the concept that children and adolescents with high BMI may already have a higher risk of end-organ damage. 
The third major finding is that the phenotypic correlation between MAP and retinal arteriolar caliber was predominantly explained by shared genetic components, and so was the correlation between BMI and retinal arteriolar caliber (Table 3). However, these phenotypic correlations were very weak. For example, only 1.2% of the total phenotypic variance for retinal arteriolar caliber was shared with MAP. This suggests that retinal arteriolar caliber may not be a major factor explaining the phenotypic variation in MAP and BMI. Given the weak phenotypic covariance between blood pressure and retinal arteriolar caliber, identifying their shared genes would be challenging and require very large sample sizes, as well as advanced genetic technology. In the Beaver Dam Eye Study (n = 1762 persons), several linkage regions (1p36, 7q21, 11q14, and 17q11) for retinal vessel diameter overlapped with the regions for essential hypertension, 22 but these regions were not subsequently replicated by others. 21,23 In the Cardiovascular Health Study (n = 1842 persons), the polymorphisms of three candidate hypertension genes (ADD1, ADRB2, and GNB3) were not associated with retinal vascular caliber. 24  
The strengths of our study include its simultaneous measurement of retinal vascular caliber, blood pressure, and BMI in a large cohort of young twins and the unique use of genetic Cholesky model that examined the genetic correlation between vascular risk factors and retinal vascular caliber. Several limitations should be highlighted as well. First, due to the cross-sectional nature of the study, we were unable to examine the causal relationship between the vascular risk factors and retinal vascular caliber. Second, while retinal vascular caliber as a continuous quantitative trait has been used as a marker for systemic vascular diseases by some investigators, although being statistically significant, its clinical significance remains uncertain and needs further investigation. 25 Third, our young twin subjects were generally free from hypertension and other cardiovascular risk factors. Structural changes in the retinal vasculature may be more likely to be adaptive in the young but maladaptive in older people. 9 In addition, some genetic effects that govern blood pressure and retinal vasculatures may emerge during adulthood. 26 Thus, our results may not be applicable to adults. Finally, although the distribution of baseline characteristics (blood pressure) and retinal vascular calibers in our young twins were very similar to those reported in population-based studies, 27 the generalizability of results from twin study to the general population should be cautiously applied as it may only apply to the distribution of phenotypes in this particular cohort. 
In summary, blood pressure and BMI already influence retinal microvasculature in children and adolescents; narrow retinal arteriole already occurs in prehypertensive and overweight children and adolescents. Although the majority of shared phenotypic variance between retinal vascular caliber and MAP and BMI is attributed to genetic components, the shared phenotypic variance itself is very low. Thus, retinal vascular caliber may not be a major factor explaining the phenotypic variation in MAP and BMI, and identifying shared genes between these traits may be very difficult. Future studies may examine whether adult populations have similar findings. 
Supplementary Materials
References
Levy BI Ambrosio G Pries AR Struijker-Boudier HA. Microcirculation in hypertension: a new target for treatment? Circulation . 2001; 104: 736–741. [CrossRef]
Silventoinen K Magnusson PK Neovius M Does obesity modify the effect of blood pressure on the risk of cardiovascular disease? A population-based cohort study of more than one million Swedish men. Circulation . 2008; 118: 1637–1642. [CrossRef] [PubMed]
Kannel WB Zhang T Garrison RJ. Is obesity-related hypertension less of a cardiovascular risk? The Framingham Study. Am Heart J . 1990; 120: 1195–1201. [CrossRef] [PubMed]
Wang JJ Rochtchina E Liew G The long-term relation among retinal arteriolar narrowing, blood pressure, and incident severe hypertension. Am. J. Epidemiol . 2008; 168: 80–88. [CrossRef] [PubMed]
Wang JJ Wong TY. Genetic determinants of retinal vascular caliber: additional insights into hypertension pathogenesis. Hypertension . 2006; 47: 644–645. [CrossRef] [PubMed]
Wong TY Knudtson MD Klein R Klein BE Meuer SM Hubbard LD. Computer-assisted measurement of retinal vessel diameters in the Beaver Dam Eye Study: methodology, correlation between eyes, and effect of refractive errors. Ophthalmology . 2004; 111: 1183–1190. [CrossRef] [PubMed]
Nguyen TT Islam FM Farouque HM Retinal vascular caliber and brachial flow-mediated dilation. The multi-ethnic study of atherosclerosis. Stroke . 2010; 41: 1343–1348. [CrossRef] [PubMed]
Ikram MK Janssen JA Roos AM Retinal vessel diameters and risk of impaired fasting glucose or diabetes: the Rotterdam study. Diabetes . 2006; 55: 506–510. [CrossRef] [PubMed]
Mitchell P Cheung N de Haseth K Blood pressure and retinal arteriolar narrowing in children. Hypertension . 2007; 49: 1156–1162. [CrossRef] [PubMed]
Gopinath B Baur LA Wang JJ Blood pressure is associated with retinal vessel signs in preadolescent children. J Hypertens . 2010; 28: 1406–1412. [CrossRef] [PubMed]
Taylor B Rochtchina E Wang JJ Body mass index and its effects on retinal vessel diameter in 6-year-old children. Int J Obes (Lond) . 2007; 31: 1527–1533. [CrossRef] [PubMed]
Cheung N Saw SM Islam FM BMI and retinal vascular caliber in children. Obesity (Silver Spring) . 2007; 15: 209–215. [CrossRef] [PubMed]
He M Ge J Zheng Y Huang W Zeng J. The Guangzhou Twin Project. Twin Res Hum Genet . 2006; 9: 753–757. [CrossRef] [PubMed]
Tomsey CS Kurtz M Kist F Hockensmith M Call P. Comparison of PowerPlex 16, PowerPlex1.1/2.1, and ABI AmpfISTR Profiler Plus/COfiler for forensic use. Croat Med J . 2001; 42: 239–243. [PubMed]
Chobanian AV Bakris GL Black HR Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension . 2003; 42: 1206–1252. [CrossRef] [PubMed]
Mi J Wang T Meng L Development of blood pressure reference standards for Chinese children and adolescents [in Chinese]. CJEBP . 2010; 5: 5–14.
Must A Dallal GE Dietz WH. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am J Clin Nutr . 1991; 53: 839–846. [PubMed]
Group of China Obesity Task Force. Body mass index reference norm for screening overweight and obesity in Chinese children and adolescents [in Chinese]. Zhonghua Liu Xing Bing Xue Za Zhi . 2004; 25: 97–102. [PubMed]
Neale MC Cardon LA. Methodology for genetic studies of twins and families . Dordrecht: Kluwer Academic Publisher; 1992.
Taarnhøj NC Larsen M Sander B Heritability of retinal vessel diameters and blood pressure: a twin study. Invest Ophthalmol Vis Sci . 2006; 47: 3539–3544. [CrossRef] [PubMed]
Sun C Zhu G Wong TY Quantitative genetic analysis of the retinal vascular caliber: the Australian Twins Eye Study. Hypertension . 2009; 54: 788–795. [CrossRef] [PubMed]
Xing C Klein BE Klein R Jun G Lee KE Iyengar SK. Genome-wide linkage study of retinal vessel diameters in the Beaver Dam Eye Study. Hypertension . 2006; 47: 797–802. [CrossRef] [PubMed]
Cheng CY Reich D Wong TY Admixture mapping scans identify a locus affecting retinal vascular caliber in hypertensive African Americans: the Atherosclerosis Risk in Communities (ARIC) Study. PLoS Genet . 2010; 6: e1000908. [CrossRef] [PubMed]
Sun C Wang JJ Islam FM Hypertension genes and retinal vascular calibre: the Cardiovascular Health Study. J Hum Hypertens . 2009; 23: 578–584. [CrossRef] [PubMed]
McGeechan K Liew G Macaskill P Risk prediction of coronary heart disease based on retinal vascular caliber (from the Atherosclerosis Risk In Communities [ARIC] Study). Am J Cardiol . 2008; 102: 58–63. [CrossRef] [PubMed]
Kupper N Ge D Treiber FA Snieder H. Emergence of novel genetic effects on blood pressure and hemodynamics in adolescence: the Georgia Cardiovascular Twin Study. Hypertension . 2006; 47: 948–954. [CrossRef] [PubMed]
Cheung N Islam FM Saw SM Distribution and associations of retinal vascular caliber with ethnicity, gender, and birth parameters in young children. Invest Ophthalmol Vis Sci . 2007; 48: 1018–1024. [CrossRef] [PubMed]
Footnotes
 Supported by grants from the National Natural Science Foundation of China (30772393, 81100686) and Fundamental Research Funds of State Key Laboratory of Ophthalmology.
Footnotes
3  These authors contributed equally to the work presented here and should therefore be regarded as equivalent authors.
Footnotes
 Disclosure: Y. Zheng, None; W. Huang, None; J. Zhang, None; M. He, None
Figure. 
 
Linear regression analysis of retinal vascular calibers with MAP and BMI. The linear regression models are adjusted for age, sex, birth order, zygosity, axial length, spherical equivalent, MAP (for the models for BMI), BMI (for the models for MAP), and the fellow retinal vascular caliber.
Figure. 
 
Linear regression analysis of retinal vascular calibers with MAP and BMI. The linear regression models are adjusted for age, sex, birth order, zygosity, axial length, spherical equivalent, MAP (for the models for BMI), BMI (for the models for MAP), and the fellow retinal vascular caliber.
Table 1. 
 
Baseline Characteristics of the Twin Individual by Zygosity
Table 1. 
 
Baseline Characteristics of the Twin Individual by Zygosity
All Twin Individuals (N = 2070) Monozygotic Twin (N = 1314) Dizygotic Twin (N = 756)
Age, y 2070 12.7 (3.2) 12.8 (3.1) 12.4 (3.2)
Sex 2070 1072 (51.8) 708 (53.9) 364 (48.2)
BMI, cm/kg2 2070 17.9 (3.1) 17.9 (3.0) 18.0 (3.3)
MAP, mm Hg 2070 75.0 (10.4) 75.0 (10.5) 74.9 (10.2)
CRAE (μm)
 All persons 2070 150.1 (13.4) 149.8 (13.6) 150.5 (12.9)
 Persons with normotension 1772 150.6 (13.4) 150.4 (13.6) 150.9 (13.0)
 Persons with prehypertension 150 148.2 (13.1) 147.5 (13.6) 149.6 (12.3)
 Persons with hypertension 148 146.0 (12.6) 145.6 (13.2) 146.7 (11.3)
 Persons with normal weight 1851 150.3 (13.3) 150.0 (13.5) 150.9 (12.9)
 Persons with overweight 162 148.3 (14.0) 148.5 (14.5) 148.0 (13.5)
 Persons with obesity 57 147.1 (12.4) 145.1 (13.7) 148.6 (11.3)
CRVE (μm)
 All persons 2070 218.4 (19.1) 218.3 (19.3) 218.6 (18.7)
 Persons with normotension 1772 218.4 (19.0) 218.3 (19.2) 218.6 (18.7)
 Persons with prehypertension 150 216.0 (19.6) 215.8 (20.5) 216.3 (17.9)
 Persons with hypertension 148 221.0 (19.7) 221.1 (19.4) 220.8 (20.4)
 Persons with normal weight 1851 218.1 (19.3) 218.1 (19.6) 218.0 (18.7)
 Persons with overweight 162 221.1 (16.8) 220.8 (15.7) 221.4 (18.3)
 Persons with obesity 57 221.8 (19.4) 219.2 (17.9) 223.7 (20.5)
Table 2. 
 
Phenotypic Correlations in MZ and DZ Twins
Table 2. 
 
Phenotypic Correlations in MZ and DZ Twins
All Male Female
AGE BMI MAP CRAE CRVE AGE BMI MAP CRAE CRVE AGE BMI MAP CRAE CRVE
Phenotypic correlations
N = 2070 twin N = 998 male twin N = 1072 female twin
AGE 1.0
BMI 0.66 1.0 0.66 1.0 0.68 1.0
MAP 0.39 0.46 1.0 0.47 0.53 1.0 0.31 0.37 1.0
CRAE –0.18 –0.20 –0.19 1.0 –0.22 –0.22 –0.20 1.0 –0.15 –0.15 –0.17 1.0
CRVE –0.15 –0.08 –0.06 0.57 1.0 –0.16 –0.04 –0.06 0.58 1.0 –0.15 –0.09 –0.04 0.55 1.0
Twin correlations and cross-twin cross-trait correlations
N = 1035 male pairs N = 401 male pairs N = 438 female pairs
AGE BMI1 MAP1 CRAE1 CRVE1 AGE BMI1 MAP1 CRAE1 CRVE1 AGE BMI1 MAP1 CRAE1 CRVE1
MZ twins
N = 657 male pairs N = 303 male pairs N = 354 female pairs
AGE 1.0 0.69 0.44 –0.21 –0.16 1.0 0.69 0.51 –0.26 –0.15 1.0 0.70 0.38 –0.17 –0.17
BMI2 0.69 0.94 0.48 –0.24 –0.10 0.71 0.94 0.57 –0.25 –0.13 0.70 0.93 0.39 –0.20 –0.11
MAP2 0.33 0.42 0.57 –0.18 –0.03* 0.50 0.55 0.66 –0.19 –0.05 0.20 0.28 0.48 –0.16 –0.01
CRAE2 –0.18 –0.20 –0.19 0.63 0.49 –0.25 –0.22 –0.20 0.59 0.46 –0.14 –0.14 –0.16 0.64 0.50
CRVE2 –0.19 –0.12 –0.11 0.47 0.68 –0.22 –0.06 –0.09 0.46 0.71 –0.17 –0.14 –0.12 0.47 0.65
DZ twins
N = 378 male pairs N = 98 male pairs N = 184 female pairs
AGE 1.0 0.63 0.42 –0.17 –0.16 1.0 0.57 0.34 –0.11 –0.08 1.0 0.63 0.42 –0.17 –0.16
BMI2 0.62 0.69 0.37 –0.14 –0.01* 0.56 0.65 0.32 –0.12 –0.08* 0.62 0.72 0.37 –0.12 –0.01*
MAP2 0.35 0.28 0.38 0.01* 0.03* 0.32 0.21 0.29 0.05* 0.04* 0.35 0.28 0.44 0.01* 0.03*
CRAE2 –0.12 –0.09* –0.01* 0.28 0.20 –0.15 –0.12* –0.08* 0.34 0.26 –0.12 –0.09* –0.01* 0.39 0.20
CRVE2 –0.05 –0.02* 0.06* 0.21 0.34 –0.03 –0.07* 0.07* 0.24 0.46 –0.05 –0.02* 0.06* 0.21 0.33
Table 3. 
 
Sources of Variance in CRAE and CRVE and Their Correlations Based on the Best-Fitting Multivariate Genetic Model
Table 3. 
 
Sources of Variance in CRAE and CRVE and Their Correlations Based on the Best-Fitting Multivariate Genetic Model
h2 (%)* e2 (%)* Age2 (%)* Total (%)*
Sources of variance for CRAE
 Age 3.0 3.0
 BMI 1.6 1.6
 MAP 1.0 0.2 1.2
 CRVE 35.4 2.8 38.2
 CRAE specific 20.5 35.5 56.0
 CRAE total 58.5 38.5 3.0 100
Sources of variance for CRVE
 Age 2.0 2.0
 BMI
 MAP 0.0 0.0
 CRAE 35.4 2.8 38.2
 CRVE specific 29.7 30.1 59.8
 CRVE total 65.1 32.9 2.0 100
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×