Telomere Length Inversely Correlates With Pulse Pressure
and Is Highly Familial
Elisabeth Jeanclos, Nicholas J. Schork, Kirsten O. Kyvik, Masayuki Kimura,
Joan H. Skurnick, Abraham Aviv
Abstract—There is evidence that telomeres, the ends of chromosomes, serve as clocks that pace cellular aging in vitro and
in vivo. In industrialized nations, pulse pressure rises with age, and it might serve as a phenotype of biological aging
of the vasculature. We therefore conducted a twin study to investigate the relation between telomere length in white
blood cells and pulse pressure while simultaneously assessing the role of genetic factors in determining telomere length.
We measured by Southern blot analysis the mean length of the terminal restriction fragments (TRF) in white blood cells
of 49 twin pairs from the Danish Twin Register and assessed the relations of blood pressure parameters with TRF. TRF
length showed an inverse relation with pulse pressure. Both TRF length and pulse pressure were highly familial. We
conclude that telomere length, which is under genetic control, might play a role in mechanisms that regulate pulse
pressure, including vascular aging. (Hypertension. 2000;36:195-200.)
Key Words: blood pressure 䡲 pulse 䡲 age 䡲 twins
T
biological aging of replicating somatic cells in different organ
systems of humans? A related question is: Is the aging of
tissues from persons who are genetically endowed with long
telomeres likely to occur later in life or at a slower pace than
of tissues from persons who inherit short telomeres? Second,
which biological parameters can serve as indicators of aging
in human beings, since for obvious reasons chronological age
(which is determined by calendar time) is a poor criterion for
biological aging?
In light of these considerations, this work had 2 goals. The
first goal was driven by the following concept. Since in
industrialized nations pulse pressure increases with age,13
pulse pressure might serve as a phenotype of cardiovascular
aging. We therefore examined whether pulse pressure correlates with telomere length. The second goal was to examine
whether telomere length is familial.
elomeres, the ends of chromosomes, undergo attrition
(shortening) in their length with each replicative cycle of
cultured somatic cells (reviewed in References 1 through 4).
This process also occurs in vivo because an inverse relation
exists between telomere length in replicating somatic cells
and the age of human beings who have donated these cells
(References 5 through 9; reviewed in References 1 through
4). Thus, the replicative history of somatic cells is a major
determinant of telomere length. Another determinant of
telomere length is heredity, since the high variability in this
parameter among human beings is to a large extent genetically determined.5 Recent experimental data support the
concept that telomeres might serve as “biological clocks,”
pacing not only life span at the cellular level but also aging at
the systemic level. These data show that (1) the prevention of
telomere attrition by the forced expression in cultured somatic cells of the catalytic component of telomerase, the
reverse transcriptase that adds telomere repeats onto the ends
of chromosomes, postpones replicative senescence10,11 and
(2) the “knockout” of telomerase in the mouse amplifies some
characteristics associated with systemic aging in later generations of mice that exhibit substantially shortened telomere
length.12
At least 2 fundamental questions therefore arise with
respect to the clinical implications of telomere biology. First,
can the length of telomeres serve as an in vivo indicator of
Methods
Subjects
DNA samples from white blood cells (WBCs) of 98 healthy twins
(10 monozygotic and 39 dizygotic twin pairs) from the Danish
Twin Register14 were selected to be studied. The ages of the
subjects were 18 to 44 years. The twin pairs were identified in
1994 by responses to a questionnaire. They were selected for
further investigation after they reported negative history for major
metabolic, cardiovascular, and chronic inflammatory and infectious diseases. Clinical evaluation of and blood collection from
Received December 20, 1999; first decision January 11, 2000; revision accepted February 29, 2000.
From the Hypertension Research Center (E.J., M.K., A.A.) and the Department of Preventive Medicine and Community Health (J.H.S.), University
of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark; the Department of Epidemiology and Biostatistics (N.J.S.), Case Western
Reserve University, Cleveland, Ohio; the Program for Population Genetics and Department of Biostatistics (N.J.S.), Harvard University School of Public
Health, Boston, Mass; the Jackson Laboratory (N.J.S., J.H.S.), Bar Harbor, Maine; and the Danish Twin Register (K.O.K.), Genetic Epidemiology
Research Unit, Institute of Community Health, Odense University, Denmark.
Correspondence to Abraham Aviv, Room F-464, MSB, Hypertension Research Center, University of Medicine and Dentistry of New Jersey, 185 S
Orange Ave, Newark, NJ 07103-2714. E-mail avivab@umdnj.edu
© 2000 American Heart Association, Inc.
Hypertension is available at http://www.hypertensionaha.org
195
196
Hypertension
August 2000
the co-twins of each twin pair were done on the same day.
Evaluation included a physical examination with measurements
of height, weight, and blood pressure. Blood pressure was
carefully measured by a standard mercury sphygmomanometer,
and the mean of 3 measurements was taken. Measurements were
performed in a sitting position after 5 minutes of complete rest.
Zygosity was established by HLA-typing and by serological
analyses of blood enzyme systems.
Approval to perform this research was obtained from the
Danish Central Scientific Ethics Committee. Consent to send
DNA to the United States was provided by the regional ScientificEthics Committee. Approval to perform the research was also
granted by the Institutional Review Board of the University of
Medicine and Dentistry of New Jersey–New Jersey Medical
School.
Measurement of Terminal Restriction
Fragment Length
DNA samples were coded in Odense University by means of
numbers and the letters A and B, denoting the 2 co-twins of each
twin pair. No other information (ie, zygosity, blood pressure, age,
gender) was revealed by the code. The samples were digested
overnight with restriction enzymes Hinf I (10 U) and RsaI (10 U)
(Boehringer Mannheim). Eighteen DNA samples (⬇5 ␮g each)
from different individuals and 4 DNA ladders (1 kb DNA ladder
plus 1 DNA/Hind III Fragments; GIBCO Life Technologies) were
resolved in a 0.5% agarose gel (20⫻20 cm) at 50 V (GNA-200
Pharmacia Biotech). Duplicates from the same samples were
resolved on different gels. The letter coding (ie, A and B) enabled
the running of DNA samples from each twin pair on the same gel.
After 16 hours, the DNA was depurinated for 30 minutes in 0.25N
HCl, denatured for 30 minutes in 0.5 mol/L NaOH/1.5 mol/L
NaCl, and neutralized for 30 minutes in 0.5 mol/L Tris, pH 8, 1.5
mol/L NaCl. The DNA was transferred for 1 hour to a nylon
membrane, positively charged (Boehringer Mannheim) with the
use of a vacuum blotter (Appligene, ONCOR). The membranes
were then hybridized at 65°C with the telomeric probe (digoxigenin 3⬘-end labeled 5⬘-[CCCTAA]3) overnight in 5⫻SSC, 0.1%
Sarkosyl, 0.02% SDS and 2% Blocking reagent (Boehringer
Mannheim). The membranes were washed at room temperature, 3
times in 2⫻SSC, 0.1% SDS each for 15 minutes and once in
2⫻SSC for 15 minutes. The digoxigenin-labeled probe was
detected by the digoxigenin luminescent detection procedure
(Boehringer Mannheim) and exposed on x-ray film. The mean
terminal restriction fragment (TRF) length was measured as
described before.15 After completion of all TRF measurements in
all samples, the numbers were decoded for data analysis.
Data Analysis
To assess the relation of measured factors (eg, gender, blood
pressure, age) with telomere length while controlling for gross
genetic effects on these phenotypes, we used a linear model with
random effect or “variance component” terms (eg, see References
16 through 18). Let y1 and y2 denote telomere length values
collected from a twin pair. Assume that the twin pair trait value
vector, Y⫽[y1,y2], can be modeled with an appropriate bivariate
distribution (eg, bivariate normal) with mean vector, ␮, and
variance-covariance matrix, ⍀, which can be partitioned in the
following way:
(1)
⍀⫽2K ␴ 2a⫹I␴2r
where ␴2a and ␴2r are estimable variance components terms characterizing gross additive genetic effects (ie, aggregate additive effects
of many loci), and random or “error” effects, respectively. The
coefficient terms preceding these variance terms are 2⫻2 coefficient
matrices relating the variance components to the twin pair trait
values. Thus, K is the kinship coefficient matrix with off-diagonal
elements equaling 1.0 for MZ twins and 0.5 for DZ twins, and I is the
identity matrix.
Assume further that ␮ can be modeled as ␮⫽f(X B), where X
is vector of covariates (ie, gender, age) and B is an estimable
regression parameter vector. For gender, men were assigned a
value of 1.0 and women a value of 0.0. We assumed a linear
relation between Y and X. The variance component terms and the
parameter vector B can be estimated by maximum likelihood.
Since we assumed bivariate normality of telomere length among
twins and a linear relation between Y and X, the relevant
log-likelihood equation is:
(2) L(B, ␴ 2a,␴2r Y,X)⫽⫺(1/ 2)log ⍀ ⫺(1/ 2)(Y⫺BX)⬘⍀⫺1(Y⫺BX)
Because more than 1 twin pair was collected, the log-likelihood
equation was determined as the sum of the individual loglikelihood for each twin pair. Thus, our proposed model is
essentially a standard regression model with dependent (paired)
observations and special covariance structure. To test the relation
of a measured factor, for example, pulse pressure, with telomere
length, we tested the significance of the regression coefficient for
that factor by using likelihood ratio tests. To safeguard against
robustness issues, log-transformed variables were also tested.
Note that heritability can be estimated as H⫽␴2a/(␴2a⫹␴2r).
Because our model can accommodate multiple factors in the
analysis, we also performed stepwise regressions that could
determine the set of factors related to a chosen dependent variable
that are statistically optimal and independent in their effects.19 It
must be emphasized that by not directly testing other sources of
“familial aggregation” beyond additive genetic effects (eg, shared
diets, lifestyles, housing), any estimate of heritability from our
analysis probably is biased. This is true for all twin and standard
estimators of heritability that are not exhaustive in terms of the
influences that they model.
Results
Table 1 summarizes general characteristics of the subjects,
ignoring the relatedness of the twins. Reliability of the
measurements of TRF length is described in Figure 1, which
provides a scatterplot of the 2 TRF determinations and a plot
of their difference versus the mean of the determinations.
Table 2 describes the results of the analysis of the relation
between each of the measured factors and TRF length,
systolic blood pressure, diastolic blood pressure, and pulse
pressure in univariate or pairwise settings. We also present
the estimated percentage of variation in each primary variable
that was explained by additive genetic random effects, after
accounting for the effect of the measured factor. There was no
significant correlation between TRF length and age within the
age range of subjects in this group. However, gender showed
a significant relation with TRF length in that women had
longer TRF (also see Table 1). Of the blood pressure
parameters, pulse pressure, which was correlated with age,
showed the strongest relation with TRF length. TRF length
was correlated positively with diastolic blood pressure but
negatively with systolic blood pressure (Table 2), which is
consistent with a negative relation between TRF length and
pulse pressure. In addition, TRF length and pulse pressure
were found to be highly heritable.
The most parsimonious multivariate model (from the use
of a stepwise regression analysis) for TRF length included
only pulse pressure (slope⫽⫺0.01 kb/mm Hg, P⬍0.01; first
row of Table 3). The most parsimonious multivariate model
for pulse pressure included gender, age, and TRF length,
which suggests that the relation between TRF length and
pulse pressure is independent of gender and age.
Jeanclos et al
TABLE 1.
Variable
Gender, M/F
Age
Height
Telomeres and Pulse Pressure
197
Characteristics of Subjects Participating in the Study
Overall
Mean⫾SD
n
Men,
Mean⫾SD
n
38/60
98
36.98⫾7.95
171.81⫾9.17
Women,
Mean⫾SD
98
37.52⫾6.51
38
36.63⫾77
60
96
179.63⫾7.22
38
166.69⫾6.23
58
n
Weight
72.24⫾18.06
96
83.42⫾20.31
38
64.91⫾11.76
58
BMI
24.47⫾0.37
96
25.17⫾0.41
38
26.72⫾0.29
58
SBP
115.07⫾17.52
96
125.21⫾19.99
38
108.43⫾11.84
58
DBP
64.56⫾9.72
96
65.75⫾9.32
38
63.78⫾9.97
58
PP
50.51⫾17.44
96
59.45⫾19.58
38
44.65⫾13.04
58
TRF
8.73⫾0.79
98
8.51⫾0.88
38
8.87⫾0.71
60
BMI indicates body mass index; SBP and DBP, systolic and diastolic blood pressure, respectively;
PP, pulse pressure (ie, SBP⫺DBP); TRF, terminal restriction fragment. Units: Height (in cm), weight
(in kg), BMI (in kg/m2), SBP and DBP (in mm Hg), and TRF (in kb). n is the number of individuals with
nonmissing values used in the calculations. There were 10 monozygotic twins and 39 dizygotic twins.
Figure 2 offers a graphical depiction of the relation between
TRF length and pulse pressure, ignoring the relatedness of the
twins. The parameters of the linear regression are Pulse Pressure⫽107.57⫺6.54 TRF (r⫽⫺0.30, P⫽0.0032). The parameters of the regression describing the relation between TRF length
and pulse pressure in which average measures for each twin pair
are plotted against each other are as follows: Pulse Pressure⫽111.0⫺6.93 TRF (r⫽⫺0.33, P⫽0.024). We note that for
the multivariate models whose results are described in Table 3,
some variables were not considered as potential predictor variables because of collinearity with other variables. Thus, systolic
blood pressure and diastolic blood pressure were not considered
in analyses involving pulse pressure because pulse pressure is
defined by systolic and diastolic blood pressure values. We also
note that the correlation between mean arterial pressure and TRF
Figure 1. Scatterplot showing relation between first and second
TRF length measurements made on each subject. The 2 measurements were performed 3 months apart. Dashed line represents line of identity. Solid line is regression line taking second
measurement as a function of first measurement (equation:
TRF2⫽0.136⫹0.985 TRF1 (r⫽0.963). Inset shows difference
between first and second measurements plotted against mean
of the 2 measurements. Solid line indicates mean difference
(⫺0.009 kb). Dotted lines represent mean⫾2 SD (SD⫽0.296 kb).
was negligible (r⫽0.07). The consequence is that in multiple
regression models, pulse pressure was a significant predictor of
TRF but mean arterial pressure was not. When pulse pressure
and TRF were adjusted for mean arterial pressure, the partial
correlation of pulse pressure and TRF (Figure 2) was actually
stronger (r⫽⫺0.33). Although height was correlated with pulse
pressure in this cohort (r⫽0.37, P⫽0.0002), height was not
correlated with TRF (r⫽0.11, P⫽0.27). Height thus had little
impact on the explanatory relation between TRF and pulse
pressure; the partial correlations coefficient of pulse pressure and
TRF after adjustment for height was r⫽⫺0.28, P⫽0.0005.
Finally, analyses of the variables after log-transformation
did not change the results appreciably (data not shown).
Discussion
The main finding of this work was that pulse pressure was
inversely correlated with the TRF length in WBCs. This
finding suggests that individuals who are endowed with
relatively longer telomeres manifest a relatively narrow pulse
pressure. In addition, this work showed that the mean length
of telomeres and pulse pressure were highly familial. Our
findings are discussed below.
In industrialized nations, systolic blood pressure continuously rises throughout life (Reference 20; reviewed in Reference 13). Diastolic blood pressure also rises in early life,
but it tends to level off or even decline in older persons.
Hence, pulse pressure manifests progressive widening as a
function of age. Arterial aging, particularly expressed by
stiffness of central elastic arteries, is a major but not the only
factor that determines pulse pressure; other determinants
include left ventricular ejection rate and stroke volume.
Perhaps the most important variable that determines central
arterial stiffness is chronological age.20 –22 However, factors
that might enhance the biological aging of the vasculature,
including essential hypertension,23 non–insulin-dependent diabetes mellitus,24 and a high salt intake,25 have been independently shown to increase arterial stiffness. Collectively,
these observations suggest that aortic pulse pressure might
serve as a phenotype of biological aging of central arteries
198
Hypertension
TABLE 2.
August 2000
Pairwise Relations Between TRF Length, Blood Pressure, and Concomitant Factors
Factor
Cff
TRF
Gen
N-t
TRF
–
–
–
SBP
⫺0.01
Cff
⫺3.51
SBP
Gen
N-t
0.sx
DBP
Gen
N-t
Cff
PP
Gen
N-t
90.7
47
2.73
79.2
48
⫺6.11‡
85.6
47
0.17*
95.9
47
–
–
–
79.5
47
0.83†
79.3
47
0.01*
95.8
48
0.33*
88.8
47
–
–
–
⫺0.66‡
88.8
47
PP
⫺0.01‡
96.4
47
0.72†
85.7
47
⫺0.28‡
85.7
47
–
–
–
Gender
⫺0.40*
95.1
49
87.8
47
1.82
79.7
48
12.6‡
82.9
47
DBP
14.1‡
BMI
0.00
95.7
48
6.80†
88.0
47
4.42*
81.5
48
3.39
82.3
47
Age
⫺0.01
95.4
49
0.74‡
88.8
47
0.01
79.9
48
0.75‡
83.3
47
Cff indicates maximum likelihood estimate of the regression coefficient for the listed factor (*P⬍0.05; †P⬍0.01; ‡P⬍0.005) by likelihood ratio test
based on ␹2 with 1 degree of freedom; gender was coded as 0⫽female, 1⫽male; Gen, percentage of residual variation caused by genetic factors
(see text); and N-t, sample size (number of twin pairs). Dash indicates that the factor was not tested. TRF indicates terminal restriction fragment; SBP,
systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure (ie, PP⫽SBP⫺DBP). Note: After TRF and PP were log-transformed, they
were still highly negatively correlated (⬍0.01).
(for review, see Reference 26) and is a predictor of cardiovascular mortality and morbidity.27–30
We note, however, that the subjects in our study were as
young as 18 years old. Therefore, their brachial pulse pressure was probably higher than that of their aortic or carotid
pulse pressure, given their increased heart rate and amplification of the brachial systolic blood pressure.22,26 In addition,
it is well established that height is a major determinant of the
relation between pulse pressure and heart rate.26,31,32 Although in our cohort, height did not provide an explanation
for the relation between pulse pressure and TRF length,
height (and body mass index) must be evaluated as confounding factors in large-scale examinations of the TRF and pulse
pressure.
Some variations may exist in telomere length among
somatic cells, probably as the result of different proliferative
rates of tissues. Yet, in comparison to other persons, persons
who exhibit either relatively short or long telomeres in one
type of a proliferative somatic cell, respectively, express
relatively short or long telomeres in other somatic cells
(Reference 33; also K. Okuda and A. Aviv, unpublished
data). Thus, the relation between TRF length and pulse
pressure in the brachial artery might hold not only for
telomeres in WBCs but also for telomeres in other replicating
cells, including vascular endothelial cells8 and vascular
smooth muscle cells.34 These cells play a pivotal function in
blood pressure control and vascular aging. In addition, it is
unlikely that height is a factor in heritability of TRF length,
since no relation was observed in our cohort between TRF
length and height.
Not only pulse pressure but also the TRF changes with
age.5–9 It appears, however, that different phases in the rate of
telomere attrition exist throughout life.35 The initial phase (ie,
birth to 5 years) is marked by a relatively high rate of
telomere attrition. The subsequent phase that includes adolescence and young adulthood is marked by an apparent
stabilization of telomere length. Thereafter, telomere attrition
resumes at a slower rate than during the first 5 years of life.
The majority of subjects we studied were within the age range
in which the rate of telomere attrition slows down or levels
off altogether, accounting for the lack of correlation between
the telomere length and age in this group.
In this study, we found that the TRF length in WBCs
was highly familial, for example, confirming observations
by Slagboom et al5 showing heritability of TRF length in
lymphocytes. There is evidence that the TRF length differs
among subpopulations of WBCs (eg, References 7 and 36),
but as indicated earlier, the differences in the TRF length
TABLE 3. Most Parsimonious Multiple Regression Models for TRF Length and
Blood Pressure
N-t
Gender
BMI
PP
Age
TRF
Gen.
Ran
TRF
47
NS
NS
⫺0.01†
NS
–
84.0
16.0
SBP
47
13.83†
NS
–
0.72†
NS
85.2
14.8
DBP
47
NS
4.52*
–
NS
2.78*
79.3
20.7
PP
47
10.82†
NS
–
0.70†
⫺4.24*
81.1
18.9
Each row corresponds to the model fitting results for the variable listed in the first column. Gender
was coded as 0⫽female, 1⫽male; TRF, terminal restriction fragment; SBP, systolic blood pressure;
DBP, diastolic blood pressure; and PP, pulse pressure (SBP⫺DBP). Entries are the maximum
likelihood estimates of regression coefficients for the factors’ contribution to the model for which the
variable listed in the first column was taken as the dependent variable, except for the “Gen” and
“Ran” columns, which give the estimated fraction of residual variation explained by genetic and
random factors, respectively. NS indicates not significant and therefore the factor was not included
in the final model (ie, P⬎0.05), all listed coefficients were significant (*P⬍0.05); †P⬍0.01). Dash
indicates that the factor was not tested in the model. N-t is the number of twin pairs with nonmissing
values of all relevant variables.
Jeanclos et al
Telomeres and Pulse Pressure
199
References
Figure 2. Relation between pulse pressure and TRF length.
Regression line represents best-fit linear regression of pulse
pressure on TRF.
within subpopulations of somatic cells are far smaller than
differences in the TRF length among persons of the same
age. For instance, in the same donor, differences in
telomere length between naı̈ve and memory T lymphocytes
could at most reach 2 kb,7 whereas differences in WBCs or
lymphocytes among donors of the same age could be as
high as 5 kb.5,37 Thus, the respective findings by Slagboom
et al5 and us in lymphocytes and WBCs indicate that high
heritability of TRF length is likely to be expressed in all
cell types. This conclusion was also reached by Martens
et al.33
There are substantial data about heritabilities of systolic
and diastolic blood pressures (reviewed in Reference 38)
but little information about heritability of pulse pressure.39
Heritabilities of systolic and diastolic blood pressures in
this study were higher than in previous reports.38 This may
be due to the fact that our model did not accommodate
other unmeasured factors, such as shared diets, living
conditions, and so forth, which could contribute to similarity in twin values and be erroneously attributed to
genetic effects.
We propose that to gain a better appreciation of the link
between telomere biology and vascular aging in human
beings, large-scale investigations should be undertaken to
explore further the relation between telomere length and
pulse pressure at a wide age range.
Acknowledgments
Elisabeth Jeanclos is a postdoctoral fellow of the American Heart
Association, Northeastern Consortium, Heritage Affiliate; Nicholas
J. Schork is supported in part by National Institute of Health grants
RR03655-11 from NCRR and HL-94011 and HL-54998 from the
National Heart, Lung, and Blood Institute; Kirsten O. Kyvik was
supported by the Danish Diabetes Association, Novo Nordic Foundation, and Sygekassernes Helsefond; Masayuki Kimura was partially supported by a grant-in-aid from the American Heart Association; and Abraham Aviv was partially supported by National
Institutes of Health grant HL-47906. We thank Dr Anthanase
Benetos for reading the manuscript and for his valuable comments.
1. Greider CW. Telomeres and senescence: the history, the experiment, the
future. Curr Biol 1998; 8:R178 –R181.
2. Blackburn EH. Telomeres: no end in sight. Cell. 1994;77:621– 623.
3. Broccoli D, Cooke H. Aging, healing, and the metabolism of telomeres.
Am J Hum Genet. 1993;52:657– 660.
4. Harley CB. Telomere loss: mitotic clock or genetic time bomb? Mutat
Res. 1991;256:271–282.
5. Slagboom PE, Droog S, Boomsma DI. Genetic determination of telomere
size in humans: a twin study of 3 age groups. Am J Hum Genet. 1994;
55:876 – 882.
6. Vaziri H, Schachter F, Uchida I, Wei L, Zhu X, Effros R, Choen D,
Harley CB. Loss of telomeric DNA during aging of normal and trisomy
21 human lymphocytes. Am J Hum Genet. 1993;52:661– 667.
7. Weng NP, Levine BL, June CH, Hodes RJ. Human naive and memory T
lymphocytes differ in telomeric length and replicative potential. Proc Natl
Acad Sci U S A. 1995;92:11091–11094.
8. Chang E, Harley CB. Telomere length and replicative aging in human
vascular tissues. Proc Natl Acad Sci U S A. 1995;92:11190 –11194.
9. Effros RB. Replicative senescence in the immune system: Impact of the
Hayflick limit on T-cell function in the elderly. Am J Hum Genet.
1998;62:1003–1007.
10. Bodnar AG, Quellete M, Frolkis M, Holt SE, Chiu CP, Morin GB, Harley
CB, Shay JW, Lichtsteiner S, Wright HE. Extension of life-span by
introduction of telomerase into normal human cells. Science. 1998;279:
349 –352.
11. Morales CP, Holt SE, Quellete M, Kaur KJ, Yan Y, Wilson KS, White
MA, Wright WE, Shay JW. Absence of cancer-associated changes in
human fibroblasts immortalized with telomerase. Nat Genet. 1999;21:
115–118.
12. Rudolph KL, Chang S, Lee HW, Blasco M, Gottlieb GJ, Greider C,
DePinho RA. Longevity, stress response, and cancer in aging telomerasedeficient mice. Cell. 1999;96:701–712.
13. Whelton PK, He J, Klag MJ. Blood pressure in Westernized population.
In: Swales JD, ed. Textbook of Hypertension. London, UK: Blackwell
Scientific Publications; 1994:11–21.
14. Kyvik KO, Green A, Beck-Nielsen H. The New Danish Twin Register:
establishment and analysis of twinning rates. Int J Epidemiol. 1995;24:
589 –596.
15. Harley CB, Futcher AB, Greider CW. Telomeres shorten during aging of
human fibroblasts. Nature. 1990;345:458 – 460.
16. Schork NJ. The design and use of variance component models in the
analysis of human quantitative pedigree data. Biometrical J. 1993;4:
387– 405.
17. Schork NJ. Extended multipoint identity-by-descent analysis of human
quantitative traits: efficiency, power, and modeling considerations. Am J
Hum Genet. 1993;53:1306 –1319.
18. Searle SR, Casella G, McCulloch CE. Variance Components. New York,
NY: John Wiley; 1992.
19. Neter J, Wasserman W, Kutner MH. Applied Linear Statistical Models.
Homewood, Ill: Richard D. Irwin; 1985.
20. Franklin SS, Gustin W IV, Wong ND, Larson MG, Weber MA, Kannel
WB, Levy D. Hemodynamic patterns of age-related changes in blood
pressure: the Framingham Heart Study Circulation. 1997;96:308 –315.
21. Bramwell JC, Hill AV, McSwiney BA. The velocity of the pulse wave in
man in relation to age as measured by the hot-wire sphygmograph. Heart.
1923;10:233–255.
22. Avolio AP, Deng FQ, Li WQ, Luo YF, Huang ZD, Xing LF, O’Rourke
MF. Effects of aging on arterial distensibility in populations with high and
low prevalence of hypertension: comparison between urban and rural
communities in China. Circulation. 1985;71:202–210.
23. Gribbin B, Pickering TG, Sleight P. Arterial distensibility in normal and
hypertensive man. Clin Sci. 1979;56:413– 417.
24. Salomaa V, Riley W, Kark JD, Nardo C, Folsom AR. Non–insulindependent diabetes mellitus and fasting glucose and insulin concentrations are associated with arterial stiffness indexes: the ARIC study:
Atherosclerosis Risk in Communities Study. Circulation. 1995;91:
1432–1443.
25. Avolio AP, Clyde KM, Beard TC, Cooke HM, Ho KK, O’Rourke MF.
Improved arterial distensibility in normotensive subjects on a low salt
diet. Arteriosclerosis. 1986;6:166 –169.
26. Nichols WW, O’Rourke MF. McDonald’s Blood Flow in Arteries. Theoretical, Experimental and Clinical Principles. 4th ed. London/Sydney/
Auckland; Arnold: 1998:347–376.
27. Darne B, Girerd X, Safar M, Cambien F, Guize L. Pulsatile versus steady
component of blood pressure: a cross-sectional analysis on cardiovascular
mortality. Hypertension. 1989;13:392– 400.
200
Hypertension
August 2000
28. Benetos A, Rudnichi A, Safar M, Guize L. Pulse pressure and cardiovascular mortality in normotensive and hypertensive subjects. Hypertension.
1998;32:560 –564.
29. Verducchia P, Schillaci C, Borgioni C, Ciucci A, Pede S, Procellati C.
Ambulatory pulse pressure: a potent predictor of cardiovascular risk in
hypertension. Hypertension. 1998;32:983–988.
30. Domanski MJ, Davis BR, Pfeffer MA, Kastantin M, Mitchell GF. Isolated
systolic hypertension: prognostic information provided by pulse pressure.
Hypertension. 1999;34:375–380.
31. Westerhof N, Elzinga G. Normalized input impedance and arterial
decay-time over heart period are independent of animal size. Am J
Physiol. 1991;261:R126 –R133.
32. Milnor WR. Aortic wavelength as a determinant of the relation between
heart rate and body size in mammals. Am J Physiol. 1979;237:R3–R6.
33. Martens UM, Zijlmans JM, Poon SS, Dragowska W, Yui J, Chavez EA,
Ward RK, Lansdorp PM. Short telomeres on human chromosome 17p.
Nat Genet. 1998;18:76 – 80.
34. Okuda K, Khan MY, Skurnick J, Kimura M, Aviv H, Aviv A. Telomere
attrition of the human abdominal aorta: relationship with age and atherosclerosis. Atherosclerosis. In press.
35. Frenck RW Jr, Blackburn EH, Shannon KM. The rate of telomere
sequence loss in human leukocytes varies with age. Proc Natl Acad Sci
U S A. 1998;95:5607–5610.
36. Weng NP, Granger L, Hodes RJ. Telomere lengthening and telomerase
activation during human B cell differentiation. Proc Natl Acad Sci U S A.
1997;94:10827–10832.
37. Jeanclos E, Krolewski A, Skurnick J, Kimura M, Aviv H, Warram JH,
Aviv A. Shortened telomere length in white blood cells of patients with
IDDM. Diabetes. 1998;47:482– 486.
38. Ward R. Familial aggregation and genetic epidemiology of blood pressure.
In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis,
and Management. New York, NY: Raven Press; 1990:81–100.
39. Darlu P, Sagnier PP, Bois E. Evidences pour une transmission genetique
de la pulsatilte arterielle. C R Acad Sci III. 1994;317:62– 69.