Sci. Aging Knowl. Environ., 22 June 2005
Vol. 2005, Issue 25, p. pe18
[DOI: 10.1126/sageke.2005.25.pe18]


Endothelial Progenitor Cell Therapy for Atherosclerosis: The Philosopher's Stone for an Aging Population?

Julia Kravchenko, Pascal J. Goldschmidt-Clermont, Tiffany Powell, Eric Stallard, Igor Akushevich, Michael S. Cuffe, and Kenneth G. Manton

The authors are at the Center for Demographic Studies at Duke University, Durham, NC 27708, USA (J.K., E.S., I.A., and K.G.M) and in the Department of Medicine at Duke Medical Center, Duke University, Durham, NC 27710, USA (P.J.G.-C., T.P., and M.S.C). E-mail: krauchanka{at} (J.K.)

Key Words: atherosclerosis • cardiovascular disease mortality • endothelial progenitor cells • risk factor intervention • microsimulation


The risk of atherosclerosis and thromboembolic complications such as stroke and heart attack has been attributed to many factors directly or indirectly affecting the state of arterial vessels. Cigarette smoking, hypertension, dyslipidemia (abnormal serum concentrations of lipoproteins and triglycerides), diabetes mellitus, obesity, and other conventional risk factors promote a multistage inflammatory process in atherosclerosis. Atherosclerosis begins in childhood, and the risk factors for its clinical syndromes appear to determine, to a large degree, its rate of progression rather than its presence (1). Labarthe (2) suggested emphasizing the prevention or control of cardiovascular risk factors, concluding that fewer risk factors in a population would lead to less common and less extensive early atherosclerosis. This concept suggested the need for observational studies and intervention trials to determine the effectiveness of risk factor interventions.

However, even in the hypothetical situation in which conventional "classic" risk factors are eliminated completely, there remains a substantial risk of cardiovascular disease (CVD) mortality because of constitutive age-associated risk factors. Specifically, even assuming that the levels of hypertension, smoking, diabetes, dyslipidemia, and sedentary behavior remain constant from age 20 to age 60, the risk of a coronary event at age 60 would still be 10 to 100 times as high as the risk at age 20. It is believed that much of the extra risk at age 60 is attributable to age-related declines in the capacity of precursor cells to repair damage in the arterial endothelium (see Goldschmidt-Clermont Review and Edelberg Perspective). Bone marrow appears to contain endothelial progenitor cells (EPCs) that help to repair areas of vascular senescence, a function that, if lost as a result of aging and risk factor exposure, would lead to the acceleration of atherogenesis with age (3).

There are an estimated 58 million patients with CVD in the United States, a population that could potentially benefit from stem cell-based therapy (see Edelberg et al. Perspective) (4). To study the possible outcomes of progenitor cell therapy for atherosclerosis on a population level, we examined projections of interventions in which risk factors are maintained at specific levels selected on the basis of epidemiological and clinical studies, and compared the results to projections in which the repair capacity of the arteries is assumed to be improved by a strategy of progenitor cell therapy. Here, we describe the use of a mathematical model to simulate interventions of the prospective impact of such therapy on male and female CVD mortality. The health effects of progenitor cell therapy at the population level were compared with the effects of lifelong control of "classic" risk factors.

Study Design

Background and aims

The health effects of progenitor cell therapy were modeled using data from the 46-year follow-up of the Framingham Heart Study (see "Taking the Long View"). In 1950, 2336 males and 2873 females aged 29 to 62 years were enrolled in this study. From 1950 to 1996, 23 exams of each volunteer were conducted, one every 2 years. The parameters followed were age, diastolic blood pressure, pulse pressure (the difference between the systolic and diastolic pressure), serum cholesterol, vital capacity index (the maximum amount of air that can be exhaled after a maximum inhalation), blood glucose, body mass index (BMI), hematocrit, smoking, left ventricular hypertrophy, and pulse rate.

The presence of fatty streaks in aorta in the first decade of life (5) and in coronary arteries in the second decade (6) were described about 45 years ago. More recently, the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) study showed that lesions to the inner lining of the vessel appeared in all aortas examined and in more than half of the right coronary arteries of the youngest age group (15 to 19 years) and increased in prevalence and extent with age through the oldest age group (30 to 34 years) (1). The National Heart, Lung, and Blood Institute (NHLBI) Atherosclerosis Risk in Communities (ARIC) surveillance study (1987-2000) found that the annual rate of first heart attack in males and females starts to increase at age 35 to 40 (7).

In projections of mortality caused by CVD, our goal was to control classic CVD risk factors over the individual's lifetime and then to project mortality, assuming that progenitor cell therapy for atherosclerosis is used at age 30 without changing the observed dynamics of conventional risk factors. In our projections, we used parameters characterizing the individuals' initial health conditions, 2-year changes in risk factors, and the age-dependent hazard function (the probability of the event, for example, mortality, occurring at a given time point) for CVD.


The Framingham Study includes persons aged 29 years and older, so we were able to study the effects of interventions on population health at age 30 and older. We assumed that it would be beneficial to initiate progenitor cell therapy shortly after age 30, that is, at an age at which the depletion of bone marrow cells' ability to repair arterial endothelium appears to be first manifest but before most individuals have clinical signs of CVD.

At present, no clinical information is available about the effective duration of a course of progenitor cell therapy in humans. We made an assumption about what this duration might be based on the finding that patients who have experienced myocardial infarction exhibit shorter telomeres than do healthy controls. Telomeres become progressively shorter with each cell division; this shortening occurs in many human tissues during aging and is the major cause of senescence in cultured human cells (see Hornsby Perspective and Aviv Perspective). Specifically, the difference in mean leukocyte telomere-restriction fragment (TRF) length (a measure of telomere length) between cases (patients who have undergone myocardial infarction) and controls represents a biological age gap of more than 11 years (8). Thus, we made our projections assuming that the effective period for progenitor cell therapy was close to 10 years, hypothesizing that its effect might reflect a "rejuvenation" of arterial endothelial cells, effectively decreasing their biological age to approximately that seen in healthy controls. Using this assumption, we modeled cardiovascular events (including mortality) and simulated the future effects of various interventions.

The effects of progenitor cell therapy could be larger than we predict because we did not consider two possibilities in our projections: (i) that more than one round of progenitor cell therapy might be efficacious and (ii) that there could be an interaction of risk factor control and cell therapy, such that the average biological rate of endothelium degeneration might be reduced by risk factor therapy, thereby increasing progenitor cell therapy effects.

The model

To calculate the population effect of progenitor cell therapy, one must specify a mathematical model of the increase in CVD with age and then estimate parameters of that model from available data. The logistic regression model, which is often used for longitudinal analyses, has no mechanism to describe changes in risk factor values that ordinarily occur in longitudinal studies (9). Simple regression models of changes in risk factor values might be considered for this purpose, but they do not describe the health effects (such as CVD mortality) of the age dynamics of risk factors. The model employed herein to calculate the population effect of progenitor cell therapy is constructed to describe risk factor dynamics and mortality as linked stochastic processes (10-18).

In this model, age projections are constructed using the microsimulation technique, which is based on the simulation of trajectories (that is, serial values of physiological parameters that define a health state of an individual in his or her life). Two laws govern this process: (i) how new values of physiological parameters are defined by a set of prior values, and (ii) under what conditions this life trajectory stops because of death. These two laws are probabilistic, that is, one can predict only distributions of changes in physiological parameters or times of death but not their exact values. Using data from human studies, parameters of these laws are estimated such that an artificial simulated cohort would likely reproduce these data. The probabilistic property of these laws is fundamental and reflects individual risk heterogeneity and competing disease risks.

Because the microsimulation model represents "real life in a computer," it opens a broad range of possibilities for analyzing "what-if" scenarios in computer-based experiments. Practically, such an experiment (medical experiments are often called "analysis of interventions" or simply interventions) is planned to reflect a real experiment with individuals. It possesses a set of convenient properties: It is very quick, inexpensive, and presents no problems associated with using human subjects. Practically, an intervention is performed by making changes in the two laws of the model to reflect the properties of a study. Comparison of simulation results with and without an intervention defines the effect at a population level.

A key quantity in this modeling approach is the mortality rate µ(x,t), which is defined to be a sum of three competing risks of mortality due to cancer, CVD, and "other" causes (Fig. 1) .

View larger version (3K):
[in this window]
[in a new window]
Fig. 1. Competitive risk model for mortality rate, where t is age and x is a set of 11 risk factors/parameters (defined in the text) measured or modeled at age t.

We considered three types of interventions in our analysis. The first situation is designed so that "classic" CVD risk factors (cholesterol and glucose levels, pulse pressure, diastolic blood pressure, BMI, cigarette smoking, and pulse rate) are kept within selected limits to model current clinical recommendations. The second is designed to describe the effects of progenitor cell therapy. It is assumed that after therapy the entire cardiovascular system is "rejuvenated" by 10 years, that is, the CVD mortality component is modified as µcvd(x,t) -> µcvd(x,t – 10). The third type of intervention is the simple elimination of the cancer competing risk term, reflecting the hypothetical situation in which cancer is completely "beaten."

Results of Projections

In our projections, we assumed that it was possible to control or fix "classic" CVD risk factors during an individual's lifetime and compared these effects to the simulated effects of progenitor cell therapy and to observed age-dependent, sex-specific mortality rates. The results are presented in Fig. 2 in which mortality, µ, is plotted against age.

View larger version (14K):
[in this window]
[in a new window]
Fig. 2. Age-specific CVD mortality for four scenarios: without intervention (black), progenitor cell therapy (red), reduction of CVD risk factors (blue), and "ideal" restriction (green) of CVD risk factors.

The black curve shows the age-specific CVD mortality rate without intervention. The red curve represents age-specific CVD mortality for a situation in which progenitor cell therapy is performed at age 30, with the effect assumed to be a 10-year delay in age-related atherosclerosis progression. The observed risk factor dynamics are used in this example.

The blue curve represents the CVD mortality rate obtained when risk factors are minimized for the selected time frames, but not to "ideal" values. In this situation, it is presumed that medicines to control hypertension, serum lipid concentrations, and glucose levels are administered and that diet and physical activity are controlled over the person's lifetime, but the result of the joint efforts of the physician and patient do not produce clinically "ideal" values for the patient, resulting in pulse pressure of 30 to 55 mm Hg, diastolic blood pressure of 70 to 90 mm Hg, a BMI of 18.5 to 29.9 kg/m2, serum cholesterol levels of 180 to 240 mg/dl, serum glucose levels of 70 to 124 mg/dl, cigarette smoking at a rate of 0 to 10 cigarettes/day, and a pulse rate of 72 to 90 beats/minute. Finally, the green curve represents the CVD mortality rate if CVD risk factors are restricted to "ideal" ranges to simulate a situation in which antihypertensive, lipid-lowering, and glucose control therapy, as well as diet and physical activity, kept individuals in "ideal shape," such that they display a pulse pressure of 30 to 40 mm Hg, diastolic blood pressure of 70 to 85 mm Hg, a BMI of 18.5 to 24.9 kg/m2, cholesterol levels of 180 to 200 mg/dl, glucose levels of 70 to 100 mg/dl, a complete lack of smoking, and a pulse rate of 72 to 84 beats/minute.

As can be seen in Fig. 2, males receiving progenitor cell therapy for atherosclerosis had the lowest projected CVD mortality rate (probability of death within 2 years), compared with those receiving other interventions (including "ideal" lifetime control of CVD risk factors). This effect is more striking at age 60 and above. Females who received progenitor cell therapy have CVD mortality rates lower than those in the "non-ideally" controlled risk factors scenario. The most effective intervention for females was the "ideal" control of risk factors, but that tendency only becomes obvious at ages 80 and above. In females, we observe the possibility of estrogen-associated differences in the effectiveness of progenitor cell therapy: The effectiveness of progenitor cell therapy was about 11% higher in women at ages 36 to 44 compared with those age 50 and older; postmenopausal women display decreases in CVD mortality similar to males. Life expectancies for females and males at age 30 with and without intervention are presented in Fig. 3.

View larger version (9K):
[in this window]
[in a new window]
Fig. 3. Life expectancy in females and males at age 30 when various interventions are applied. EYL (extra years of life) is the number of years that life expectancy is increased as a result of different interventions.

The effects of progenitor cell therapy on life expectancy in females at age 30 years are comparable to the effect of totally eliminating cancer, resulting in a life expectancy of 53.42 versus 53.12 years, respectively--an additional 3.67 and 3.37 years of life as compared with the situation with no intervention. In males at age 30, progenitor cell therapy is far more effective than the elimination of cancer, resulting in a life expectancy of 49.58 versus 46.50 years, respectively--an additional 5.94 and 2.86 years of life.

Analysis of the CVD survival function (Fig. 4) shows that the implementation of progenitor cell therapy might considerably increase female and male survival. The best survival in both genders is reached when progenitor cell therapy is performed at age 30 rather than at 40 or 50.

View larger version (18K):
[in this window]
[in a new window]
Fig. 4. CVD survival function without intervention (black) and for progenitor cell therapy at age 30 (red) and 40 and 50 (blue) years old.

Implications of Results

The results of this analysis suggest a potentially larger effect of progenitor cell therapy on CVD mortality than the control of "classic" CVD risk factors. It has been previously demonstrated that levels of circulating EPCs were a better predictor of vascular reactivity than the presence or absence of conventional risk factors (19). Vascular reactivity refers to changes in vascular tone in response to factors released by the endothelium; compromised vasodilation as a result of endothelial dysfunction is often associated with CVD. The stochastic process model (17) has been used to analyze 25-year follow-up data from the Finnish Cohorts of the Seven Countries Study. (In this study, cohorts of eastern and western Finnish men were analyzed to determine whether area differences in CVD mortality in Finland could be completely explained by the differences in pulse pressure, BMI, total cholesterol, vital capacity index, cigarette smoking, and heart rate.) Application of the model showed that, of the 2.4-year difference in life expectancy between the west and the east at age 40, 29% was due to differences in risk factor dynamics (differences between the west and the east in baseline risk factor levels and variance, and their changes with age) and 71% (1.7 years) was accounted for by the effects of risk factor interactions on CVD mortality. (For example, smoking not only increases CVD mortality as an independent risk factor but also influences CVD mortality by increasing arterial blood pressure, increasing the risk of diabetes, and decreasing BMI.) (20). This result suggests that a complex multivariant process underlies atherosclerosis, which is accelerated when multiple risk factors are elevated. Standard risk factors may be better used to explain mortality differences between populations when their role in such a multifactorial process is well understood and appropriately modeled (20).


One of the most discussed "classic" CVD risk factors is a high serum cholesterol level. Control of this parameter might substantially decrease CVD mortality. The existence of interactions between EPCs and a high cholesterol level is a topic for further research. Our results show that even "ideal" lifetime control of all conventional CVD risk factors, including cholesterol levels, had effects on CVD mortality that were comparable to (in females) or smaller than (in males) the effects of progenitor cell therapy. The effects of progenitor cell therapy have been tested in Apoe–/– mice, which lack the cholesterol transport protein apoE (see Raber Review), display an abnormally high serum cholesterol level, and serve as a model of atherosclerosis. The atheroprotective effect of such therapy is not likely a result of the elimination of the vascular injury caused by high cholesterol, as was previously suggested (21), but rather a result of repair in response to injury in these mice (3). In humans, a strong correlation was observed between the number of circulating EPCs and the subjects' combined Framingham risk factor score (19). Chen et al. (22) showed that the number of EPCs is inversely correlated with total cholesterol and low-density lipoprotein cholesterol levels. The interaction of risk factors suggests that the joint elevation of risk factors accelerates atherogenesis. Conversely, controlling risk factors slows the process by denying it conditions necessary for progression. This may be why risk factors operate multiplicatively, that is, the inhibition of a risk factor may represent a reduction in a "rate-limiting" stage of the process (20).


In our model, we observed a tendency for a larger effect of progenitor cell therapy on CVD mortality reduction in women age 36 to 44 than in those age 50 and older. This difference might be a result of the effect of estrogen on EPCs. Estrogen increases the production and survival of EPCs in the bone marrow, which results in an increase in the number of circulating cells (23). The cellular and molecular events involved in regulation of EPC numbers are not clear. The involvement of the induction of nitric oxide, phosphatidylinositol 3-kinase in the endothelium (24), and the apoptosis-associated caspase-8 pathway (23) has been discussed. It is unknown how agents such as estrogens or statins (drugs that reduce cholesterol and triglyceride concentrations; see "Greasing Aging's Downward Slide") influence the production, mobilization, egress, and adhesion of EPCs at target sites. The gender differences we observed in the effects of progenitor cell therapy and risk factor control need further investigation.


To increase the predictive value of our model and to provide deeper insight into atherosclerosis progression with age, the analysis of telomeres may be of importance. Having shorter than average mean leukocyte TRF length increases the risk of myocardial infarction by a factor of about 3 (8). No significant effect of a history of hypertension, diabetes mellitus, or smoking on mean leukocyte TRF length explains the observed association. Individuals born with relatively short telomeres may be at higher risk of premature coronary heart disease (CHD). To date, investigation of the molecular basis of the genetic contribution to the risk of premature CHD has focused on the role of individual gene function rather than on a global property of the genetic material. Interindividual variation in telomere length (see Aviv et al. Perspective) could explain some of the variability in age of onset of CHD that exists even when conventional risk factors are taken into account (8).

Apoe–/– mice have shorter telomeres than healthy age-matched mice, whereas telomeres of cells of the intima (the inner lining of arteries covered by a thin layer of endothelial cells) in Apoe–/– mice that received combined hematopoietic- and stromal-enriched bone marrow cells were significantly longer than those of untreated Apoe–/– mice (3). It is likely that an increased rate of cell turnover in blood vessels, including the coronary arteries, that are subject to disturbed flow accelerates telomere loss and exacerbates associated inherited abnormalities. Telomere attrition might affect the function of a subset of genes long before the replicative capacity of the cells themselves is impaired. Such a situation might have relevance to chronic pathological processes such as atherosclerosis (8).

Life expectancy and health expenditures

A country's health status has historically been measured by life expectancy. In our simulations, progenitor cell therapy shows an effect on life expectancy comparable to the elimination of all cancers. According to our calculations, progenitor cell therapy might add 3.67 years to a female's life and 5.94 years to a male's life. This is less than the modeled 10-year delay in CVD progression because non-CVD conditions would intervene to cause the death of treated people before the full 10 years had elapsed.

Health expenditures on CVD diseases and stroke in the United States in 2005 are estimated to be $242 billion (7). Roughly, the reduction in health expenditures for CVD as a result of progenitor cell therapy may be approximated by the projected 36% reduction of the CVD mortality rate, that is, $85 to $90 billion per year. Some portion of that reduction would be offset by the cost of the progenitor cell therapy intervention. However, the largest benefit of the intervention would be an increase in human capital as a result of the creation of five additional healthy life years for each person in the population. To consider this for the simple case of the United States birth cohort of 1975 with a size of approximately 4 million persons, we note that a year of life has been valued by health economists at $100,000 (25). Assuming that, for an intervention at age 30, 5 years of life expectancy are gained, the total gain in person-years of life is 20 million (that is, 4 million x 5). At an estimated value of $100,000 per person-year, the benefit over the lifetime of the birth cohort is $2 trillion. This figure clearly exceeds by a large margin any likely cost of an intervention. The gains achieved for any period of time would be the human capital gained summed over all birth cohorts involved in the intervention. For the current year, the effect would be less than $2 trillion, because older birth cohorts are smaller than 4 million persons. For future years, the gain would be larger than $2 trillion, because more birth cohorts are larger than 4 million persons.


Modeling the interaction of CVD risk factors and progenitor cell therapy for atherosclerosis will be the subject of further study. This is necessary to understand the mechanisms of atherosclerosis progression and to guide the medical management of risk factors in different sex and age groups. Conventional risk factors may better explain mortality differences between populations when their role in such multifactorial processes is better understood and appropriately modeled.

June 22, 2005
  1. J. P. Strong, G. T. Malcom, C. A. McMahan, R. E. Tracy, W. P. Newman III, E. E. Herderick, J. F. Cornhill, Prevalence and extent of atherosclerosis in adolescents and young adults: Implications for prevention from the Pathobiological Determinants of Atherosclerosis in Youth Study. JAMA 281, 727-735 (1999).[CrossRef][Medline]
  2. D. R. Labarthe, Prevention of cardiovascular risk factors in the first place. Prev. Med. 29, S72-S79 (1999).[CrossRef][Medline]
  3. F. M. Rauscher, P. J. Goldschmidt-Clermont, B. H. Davis, T. Wang, D. Gregg, P. Ramaswami, A. M. Pippen, B. H. Annex, C. Dong, D. A. Taylor, Aging, progenitor cell exhaustion, and atherosclerosis. Circulation 29, 457-463 (2003).
  4. D. Perry, Patients' voices: The powerful sound in the stem cell debate. Science 287, 1423 (2000).[Abstract/Free Full Text]
  5. R. L. Holman, H. C. McGill, J. P. Strong, J. C. Geer, The natural history of atherosclerosis. Am. J. Pathol. 34, 209-235 (1958).[Medline]
  6. J. P. Strong, H. C. McGill. The natural history of coronary atherosclerosis. Am. J. Pathol. 40, 37-49 (1962).[Medline]
  7. Heart Disease and Stroke Statistics--2005 Update. (American Heart Association, Dallas, TX)
  8. S. Brouilette, R. K. Singh, J. R. Thompson, A. H. Goodall, N. J. Samani, White cell telomere length and risk of premature myocardial infarction. Arterioscler. Thromb. Vasc. Biol. 23, 842-846 (2003).[Abstract/Free Full Text]
  9. M. A. Woodbury, K. G. Manton, E. Stallard, Longitudinal models for chronic disease risk: An evaluation of logistic multiple regression and alternatives. Int. J. Epidemiol. 10, 187-197 (1981).[Abstract/Free Full Text]
  10. M. A. Woodbury, K. G. Manton, A random walk model of human mortality and aging. Theor. Popul. Biol. 11, 37-48 (1977).[CrossRef][Medline]
  11. M. A. Woodbury, K. G. Manton, A mathematical model of the physiological dynamics of aging and correlated mortality selection. I. Theoretical development and critiques. J. Gerontol. 38, 398-405 (1983).[Abstract/Free Full Text]
  12. M. A. Woodbury, K. G. Manton, A theoretical model of the physiological dynamics of circulatory disease in human populations. Human Biol. 55, 417-441 (1983).[Medline]
  13. K. G. Manton, E. Stallard, Chronic Disease Modeling: Measurement and Evaluation of the Chronic Disease Processes (Charles Griffin & Co., London, 1988).
  14. K. G. Manton, E. Stallard, M. A. Woodbury, J. E. Dowd, Time-varying covariates in models of human mortality and aging: Multidimensional generalizations of the Gompertz. J. Gerontol. 49, B169-B190 (1994).[Abstract]
  15. J. C. M Witteman, D. E. Grobbee, H. A. Valkenburg, A. M. van Hemert, T. Stijnen, H. Burger, A. Hofman, J-shaped relation between change in diastolic blood pressure and aortic atherosclerosis. Lancet 343, 504-507 (1994).[CrossRef][Medline]
  16. K. G. Manton, A. I. Yashin, Mechanisms of Aging and Mortality: Searches for New Paradigms (Odense Univ. Press, Odense, Denmark, 2000).
  17. K. G. Manton, E. Stallard, B. H. Singer, Projecting the future size and health status of the U.S. elderly population. Int. J. Forecasting 8, 433-458 (1992).
  18. I. Akushevich, A. Kulminski, K. G. Manton, Life tables with covariates: Dynamic model for nonlinear analysis of longitudinal data. Math. Popul. Stud. 12, 51-80 (2005).
  19. J. M. Hill, G. Zalos, J. P. Halcox, W. H. Schenke, M. A. Waclawiw, A. A. Quyyumi, T. Finkel, Circulating endothelial progenitor cells, vascular function, and cardiovascular risk. N. Engl. J. Med. 348, 593-600 (2003).[CrossRef][Medline]
  20. J. Pekkanen, K. G. Manton, E. Stallard, A. Nissinen, M. J. Karvonen, Risk factor dynamics, mortality, and life expectancy differences between eastern and western Finland: The Finnish cohorts of the seven countries study. Int. J. Epidemiol. 21, 406-419 (1992).[Abstract/Free Full Text]
  21. W. A. Boisvert, J. Spangenberg, L. K. Curtiss, Treatment of severe hypercholesterolemia in apolipoprotein E-deficient mice by bone marrow transplantation. J. Clin. Invest. 96, 1118-1124 (1995).[Medline]
  22. J. Z. Chen, F. R. Zhang, Q. M. Tao, X. X. Wang, J. H. Zhu, Number and activity of endothelial progenitor cells from peripheral blood in patients with hypercholesterolemia. Clin. Sci. (Lond.) 107, 273-280 (2004).[Medline]
  23. K. Strehlow, N. Werner, J. Berweiler, A. Link, U. Dirnagl, J. Priller, K. Laufs, L. Ghaeni, M. Milosevic, M. Bohm, G. Nickenig, Estrogen increases bone marrow-derived endothelial progenitor cell production and diminishes neointima formation. Circulation 107, 3059-3065 (2003).[Abstract/Free Full Text]
  24. M. E. Mendelsohn, R. H. Karas, The protective effects of estrogen on the cardiovascular system. N. Engl. J. Med. 340, 1801-1811 (1999).[CrossRef][Medline]
  25. D. M. Cutler, E. Richardson, Measuring the Health of the United States Population. Brookings Papers on Economic Activity. Microeconomics 1997, 217-271 (1997).
  26. Support for the research presented in this paper was provided by NIH through grant number 1R01 AG023073-01A1 8/15/04 (P.J.G.-C.) and by the National Institute on Aging through grant numbers P01AG17937 and R01AG01179 (K.G.M., E.S, and I.A.).
  27. The Framingham Heart Study (FHS) is conducted and supported by NHLBI in collaboration with FHS study investigators. This paper was prepared using a limited-access data set obtained by NHLBI and does not necessarily reflect the opinions or views of the FHS or NHLBI.
Citation: J. Kravchenko, P. J. Goldschmidt-Clermont, T. Powell, E. Stallard, I. Akushevich, M. S. Cuffe, K. G. Manton, Endothelial Progenitor Cell Therapy for Atherosclerosis: The Philosopher's Stone for an Aging Population? Sci. Aging Knowl. Environ. 2005 (25), pe18 (2005).

NIH funding trajectories and their correlations with US health dynamics from 1950 to 2004.
K. G. Manton, X.-L. Gu, G. Lowrimore, A. Ullian, and H. D. Tolley (2009)
PNAS 106, 10981-10986
   Abstract »    Full Text »    PDF »

Science of Aging Knowledge Environment. ISSN 1539-6150