Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Leukocyte Telomere Dynamics: Longitudinal Findings Among Young Adults in the Bogalusa Heart Study

Leukocyte Telomere Dynamics: Longitudinal Findings Among Young Adults in the Bogalusa Heart Study American Journal of Epidemiology Vol. 169, No. 3 ª 2008 The Authors DOI: 10.1093/aje/kwn338 This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial Advance Access publication December 4, 2008 License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Original Contribution Leukocyte Telomere Dynamics: Longitudinal Findings Among Young Adults in the Bogalusa Heart Study Abraham Aviv, Wei Chen, Jeffrey P. Gardner, Masayuki Kimura, Michael Brimacombe, Xiaojian Cao, Sathanur R. Srinivasan, and Gerald S. Berenson Initially submitted April 30, 2008; accepted for publication September 22, 2008. Leukocyte telomere length (LTL) is ostensibly a biomarker of human aging. Cross-sectional analyses have found that LTL is relatively short in a host of aging-related diseases. These studies have also provided indirect estimates of age-dependent LTL shortening. In this paper, the authors report findings of the first comprehensive longitudinal study of 450 whites and 185 African Americans in Louisiana (aged 31.4 and 37.4 years at baseline (1995–1996) and follow-up (2001–2006) examinations, respectively) participating in the Bogalusa Heart Study. Rate of change in LTL was highly variable among individuals, with some displaying a paradoxical gain in LTL during the follow-up period. The most striking observation was that age-dependent LTL shortening was proportional to LTL at baseline examination. At both baseline and follow-up examinations, African Americans had longer LTLs than whites, and smokers had shorter LTLs than nonsmokers. The longer LTL in African Americans than in whites explained in part the faster rate of LTL shortening observed among African Americans. These findings underscore the complexity of leukocyte telomere dynamics in vivo and suggest that determinants in addition to the ‘‘end-replication problem’’ contribute to telomere shortening in vivo. aging; body mass index; leukocytes; oxidative stress; smoking; telomere Abbreviations: BMI, body mass index; bp, base pairs; LTL, leukocyte telomere length. Leukocyte telomere length (LTL) is highly variable at shortening was faster, in adult African Americans than in birth (1, 2) and afterward (3–9). In cross-sectional analyses whites (15). For these reasons, we examined in a longitudinal based primarily on comparisons of LTL in individuals of study the factors that contribute to variation in LTL shorten- different ages, it seems that LTL progressively shortens lin- ing among young African-American and white adults in the early with age (3–9). However, cross-sectional analyses Bogalusa Heart Study (16). Our findings indicate that age- hardly capture the rate and trajectory of age-dependent dependent LTL shortening is proportional to LTL. LTL shortening in an individual, as well as variations in these parameters among individuals. Deciphering patterns MATERIALS AND METHODS of LTL shortening and variation in the rate of LTL shorten- ing with age among individuals would be highly relevant Participants because shortened LTL has often been detected in individu- als with aging-related disorders, cardiovascular disease in The Bogalusa Heart Study is a long-term investigation of particular (5, 9–12). Factors that predispose to cardiovascu- the natural history of cardiovascular disease beginning in lar disease, including obesity (4, 13, 14), insulin resistance childhood in the biracial community (65% white, 35% (8, 13), and cigarette smoking (4, 6, 14), are also associated African American) of Bogalusa, Louisiana (16). Between with shortened LTL. Moreover, a recent cross-sectional 1973 and 2006, 7 cross-sectional surveys of children aged analysis observed that LTL was longer, but its age-dependent 4–17 years and 9 cross-sectional surveys of adults aged Correspondence to Prof. Abraham Aviv, The Center of Human Development and Aging, Room F-464, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103 (e-mail: avivab@umdnj.edu). 323 Am J Epidemiol 2009;169:323–329 324 Aviv et al. 18–46 years, who had been previously examined as chil- Statistical analysis dren, were conducted. In the ongoing Bogalusa Heart Study Mean values and standard deviations were computed for cohort, 1,203 adults aged 24–44 years were examined for baseline and follow-up LTL and other continuous covariate cardiovascular risk factors between 2004 and 2006. LTL variables separately for African Americans and whites. data were available for 635 participants, who had blood LTLs adjusted for age and body mass index (BMI) were samples collected on 2 occasions—a baseline examination calculated as least-squares means for smokers and non- in 1995–1996 and a follow-up examination in 2001–2006. smokers by holding age and/or BMI to their mean values. Compared with individuals not included, subjects with lon- Analysis of covariance was performed by using generalized gitudinal LTL measurements showed no significant differ- linear models to test differences in baseline and follow-up ences in the relevant study variables, except that they were LTL between African Americans and whites and between 0.8 years older (P ¼ 0.032). Self-administered question- smokers and nonsmokers. The relation between LTL attri- naires were used to obtain information on cigarette smok- tion and baseline LTL was examined by linear regression ing; those who smoked at least 1 cigarette per week during models using rate of change in LTL as the dependent vari- the past 12 months were defined as current smokers. The able, adjusting for race, age, sex, BMI, and smoking. Sta- study protocol was approved by the institutional review tistical analyses were performed by using SAS version 9.0 board of Tulane University (New Orleans, Louisiana), and software (SAS Institute, Inc., Cary, North Carolina). Three each participant gave written informed consent. outliers (2 African-American women and 1 white man) whose rate of LTL shortening was more than 250 base pairs LTL measurements (bp)/year or LTL lengthening was more than 250 bp/year were excluded from the analyses. DNA from frozen blood samples, obtained at baseline and follow-up examinations, was extracted by using the Gentra Puregene Blood Kit (QIAGEN Inc., Valencia, California). DNA integrity was assessed following electrophoresis of RESULTS 0.02 lg of DNA (200 V for 0.75 hours) on 1% agarose gels and staining with 2X SYBR Green I (Invitrogen Corpora- General characteristics tion, Carlsbad, California) for 45 minutes. Telomere-length measurements were performed by ob- Major characteristics of the cohort are summarized in taining the mean length of the terminal restriction fragments, Tables 1–3. The age ranges were 20.0–40.0 years at baseline using the Southern blot method, as previously described and 25.7–48.2 years at follow-up examinations. Smokers (1, 13). In brief, after overnight DNA digestion with 10 U constituted approximately a third of the cohort. African Hinf I and 10 U Rsa I restriction enzymes (Roche Diagnos- Americans had a higher BMI than whites. Compared with tics Corporation, Indianapolis, Indiana), samples of DNA whites, African Americans had longer LTLs at baseline (3 lg each) were resolved on 0.5% agarose gels at 50 V and follow-up examinations and a higher rate of LTL short- for 16 hours. DNA was then depurinated in 0.25 N hydro- ening. The follow-up period was slightly shorter for African chloric acid, denatured in 0.5 mol/L of sodium hydroxide Americans than for whites (5.7 years vs. 6.0 years), but this per 1.5 mol/L of sodium chloride, and neutralized in difference was not significant (P ¼ 0.1180). Table 2 sum- 0.5 mol/L of tris(hydroxymethyl)aminomethane (pH 8) marizes LTL parameters for smokers versus nonsmokers. per 1.5 mol/L of sodium chloride. After transfer to a posi- Age-, BMI-, and race-adjusted LTL was significantly shorter in smokers versus nonsmokers by 141 bp (P ¼ tively charged nylon membrane, DNA was hybridized to 0.0227) at baseline examination and by 164 bp (P ¼ a digoxigenin 3#-end labeled 5#-(CCCTAA) telomeric probe (overnight at 65C), washed 3 times in 2X saline- 0.0099) at follow-up examination. However, we found no sodium citrate/0.1% of sodium dodecyl sulfate, and washed significant difference in the rate of LTL shortening between once in 2X saline-sodium citrate. Probe was detected by the smokers and nonsmokers. digoxigenin luminescent detection procedure (Roche Diag- We next examined the contribution of race, age, smoking, nostics) after exposure on Hyperfilm (GE Healthcare, Chalfont BMI, and sex to variations in LTL at baseline and follow-up St. Giles, United Kingdom). Each sample was resolved in examinations. Table 3 shows that race was the main factor duplicate on different gels, and autoradiograms were explaining variations in LTL at baseline and follow-up ex- digitized for analysis of telomere-length measurement. aminations and that smoking and BMI also accounted for Baseline and follow-up terminal restriction fragments some of the variations. from the same individuals were resolved in adjacent lanes Figure 1 displays the distribution of rate of change in LTL of each gel, with samples run in duplicate on different gels. among participants. This parameter varied considerably. The intraclass correlation coefficient of LTL was 0.972 (P < The majority of individuals displayed LTL shortening (loss) 0.0001), and the coefficient of variation for the duplicate (85.9% of African Americans and 88.0% of whites). In ad- samples was 1.4%. The laboratory that conducted the LTL dition, 1 African American and 5 whites showed no change measurements was blinded to the identity and characteris- in LTL between baseline and follow-up examinations. How- tics of subjects. The rate of change (loss or gain) in LTL was ever, a subset of the cohort showed LTL gain. The means for computed as the difference between LTL value at follow-up the rates of change for African Americans and whites are minus that at baseline examinations divided by the number displayed in Table 1. For the entire cohort, the mean rate of of years of follow-up. change was 40.7 bp/year and was higher in African Am J Epidemiol 2009;169:323–329 Leukocyte Telomere Dynamics 325 Table 1. Major Characteristics of the Cohort of Young Adults in the Bogalusa Heart Study in Louisiana at Baseline (1995–1996) and Follow-up (2001–2006) Examinations Whites African Americans All P Value Parameter (n 5 450) (n 5 185) (n 5 635) (Race) Baseline examination Age, years 31.4 (5.0) 31.4 (5.3) 31.4 (5.1) 0.7904 BMI, kg/m 27.1 (6.4) 30.5 (8.4) 28.1 (5.1) <0.0001 Smokers, % 32.4 36.8 33.7 0.2962 LTL , bp 7,288 (735) 7,847 (734) 7,451 (777) <0.0001 Follow-up examination Age, years 37.4 (4.4) 37.0 (4.7) 37.3 (4.5) 0.5823 BMI, kg/m 28.3 (6.6) 32.0 (8.9) 29.4 (7.5) <0.0001 Smokers, % 30.7 30.8 30.7 0.9715 LTL , bp 7,076 (721) 7,603 (767) 7,230 (772) <0.0001 FU Change in LTL Follow-up, years 6.0 (2.4) 5.7 (2.5) 5.9 (2.4) 0.1180 LTL, bp/year 37.8 (41.3) 47.7 (55.3) 40.7 (46.0) 0.0181 Abbreviations: BMI, body mass index; bp, base pairs; LTL, leukocyte telomere length (expressed in bp); LTL , baseline LTL; LTL , follow-up LTL. B FU Unless otherwise specified, data are presented as mean (standard deviation). P values for race difference in LTL and change in LTL were adjusted for age, sex, and BMI. Rate of change in LTL (bp/year) ¼ (LTL – LTL )/follow-up years. FU B Americans than in whites (47.7 bp/year vs. 37.8 bp/year, (standard deviation, 47.6) bp/year and that for whites was respectively; P ¼ 0.0181). 45.9 (standard deviation, 34.7) bp/year (P ¼ 0.0003). Given that different mechanisms are likely to govern LTL Figure 2 displays the relation between the rate of LTL short- shortening versus gain, we separately examined LTL dy- ening that took place during the follow-up period and LTL namics in individuals who did not gain LTL during the derived from the baseline samples. A strong association was follow-up period. For these individuals (n ¼ 561), the observed between the rate of LTL shortening and baseline LTL shortening rate for African Americans was 60.0 LTL (a 17-bp/year increase in the rate of LTL shortening for each kilobase of LTL). To evaluate the potential influence of measurement error Table 2. Leukocyte Telomere Length Parameters of Interest in on the association of LTL and its rate of change, we set Smokers Versus Nonsmokers Participating in the Bogalusa Heart a conservative limit of quantification defined as the absolute Study in Louisiana at Baseline (1995–1996) and Follow-up (2001– value of the rate of change in LTL of less than 20 bp/year to 2006) Examinations denote detectable change. The association between baseline LTL Parameter Nonsmokers Smokers P Value LTL and the rate of change was then examined for the sub- group with smaller (<20 bp/year loss or gain, n ¼ 146) and Unadjusted LTL larger (20 bp/year loss or gain, n ¼ 489) changes in LTL. LTL , bp 7,481 7,392 0.0660 For the smaller-change group, the association was not sig- LTL , bp 7,270 7,150 0.0219 FU nificant (b ¼0.05, P ¼ 0.956), but, in the group with Age-adjusted LTL a larger change, the regression coefficient was significant LTL , bp 7,492 7,370 0.0479 B (b ¼13.7, P < 0.0001). This subgroup analysis provides confidence that measurement error did not play a significant LTL , bp 7,275 7,128 0.0196 FU role in the association noted in this study. Age- and BMI-adjusted LTL Since African Americans had longer LTL at baseline and LTL , bp 7,500 7,359 0.0227 follow-up examinations and a higher rate of LTL shortening, LTL , bp 7,280 7,116 0.0099 FU we examined the race effect and other pertinent factors on Change, bp/year 40 42 0.4630 the variations in LTL rate of change for the entire cohort and separately for those individuals who demonstrated no gain Abbreviations: BMI, body mass index; bp, base pairs; LTL, leuko- in LTL during the follow-up period. The findings are sum- cyte telomere length (expressed in bp); LTL , baseline LTL; LTL , B FU marized in Table 4. When considered alone, race signifi- follow-up LTL. cantly contributed to the variations in rate of change in Rate of change in LTL (bp/year) ¼ (LTL  LTL )/follow-up FU B years. LTL. However, when race and LTL at baseline examination Am J Epidemiol 2009;169:323–329 326 Aviv et al. Table 3. Regression of Baseline (1995–1996) and Follow-up (2001–2006) Leukocyte Telomere Length on Parameters of Interest for Participants in the Bogalusa Heart Study in Louisiana Baseline LTL Follow-up LTL Parameter 2 2 2 2 b Coefficient P Value Partial R Model R b Coefficient P Value Partial R Model R Race, African American 587.8 <0.0001 0.107 551.5 <0.0001 0.096 Age, years 13.4 0.0220 0.009 11.7 0.0715 0.005 BMI, kg/m 9.1 0.0316 0.006 8.6 0.0312 0.007 Smoking, yes/no 141.6 0.0227 0.006 163.5 0.0099 0.008 Sex, female 56.8 0.3456 0.002 0.130 37.0 0.5372 0.000 0.116 Abbreviations: BMI, body mass index; LTL, leukocyte telomere length (expressed in base pairs (bp)). The rate of change in LTL (bp/year) is negative; therefore, negative regression coefficients denote faster LTL attrition. were analyzed jointly, only LTL at baseline examination follow-up period (13). A previous cross-sectional analysis accounted significantly for the rate of change in LTL, mean- of LTL in a combined sample of participants from the ing that the longer LTL in African Americans was a major Bogalusa Heart Study and the NHLBI Family Heart Study factor in the higher rate of LTL shortening in African (age range, 19–93 years) found not only that LTL was longer Americans versus whites. Adding age, BMI, smoking, and in African Americans than in whites but also that the rate of sex at baseline examination to the model provided little age-dependent LTL shortening was faster in African explanation above that of baseline LTL for interindividual Americans (15). The present work indicates that the longer variation in LTL shortening rate. These observations held LTL in African Americans partially explained their faster when the data were analyzed for the entire cohort or after age-dependent LTL shortening compared with that of whites. exclusion of individuals who gained LTL between baseline Cross-sectional analyses of leukocyte telomere dynamics and follow-up examinations. have suggested that age-dependent LTL shortening might be slower (17, 18) or faster (19) in young than in older adults, but these studies were statistically underpowered to test DISCUSSION either of these suppositions (20). Note that, in the present study, cross-sectional analysis revealed only a small effect Our key finding in this biracial cohort of young adults was of age on LTL attrition (Table 3), but this finding was due to that age-dependent LTL shortening rate was proportional to the considerable interindividual variation in LTL and the LTL. In addition, we confirmed previous observations in relatively narrow age range of the cohort (20 years for a small sample of the Bogalusa Heart Study with a 10-year follow-up showing that the rate of LTL shortening was highly variable among individuals, with a subset of partic- ipants displaying paradoxical lengthening of LTL during the Figure 1. Distribution of the rate of change in leukocyte telomere length (LTL) in Louisiana study participants. A total of 561 participants Figure 2. Relation between rate of leukocyte telomere length (LTL) showed LTL shortening, and 74 showed either gain (n ¼ 68) or no shortening during follow-up (2001–2006) and LTL derived from the change (n ¼ 6) in LTL between baseline (1995–1996) and follow-up baseline samples (1995–1996) in Louisiana study participants. D LTL¼ (2001–2006) examinations. bp, base pairs. 0.017 baseline LTL – 0.073; r ¼ 0.326, P < 0.0001. bp, base pairs. Am J Epidemiol 2009;169:323–329 Leukocyte Telomere Dynamics 327 Table 4. Regression of the Rate of Change in Leukocyte Telomere Length on Parameters of Interest for Participants in the Bogalusa Heart Study a,b in Louisiana Entire Cohort (n 5 635) Subjects With no Gain in LTL (n 5 561) 2 2 2 2 b Coefficient P Value Partial R Model R b Coefficient P Value Partial R Model R Model I Race, African American 9.9 0.0135 0.010 0.010 14.1 <0.0001 0.026 0.026 Model II Race, African American 4.6 0.2772 0.010 5.9 0.1070 0.026 LTL ,bp 9.6 <0.0001 0.023 0.033 15.6 <0.0001 0.084 0.110 Model III Race, African American 3.4 0.4309 0.010 4.4 0.2424 0.026 LTL ,bp 9.9 <0.0001 0.023 16.1 <0.0001 0.084 Age, years 0.6 0.1120 0.003 0.6 0.0605 0.007 BMI, kg/m 0.2 0.5464 0.000 0.3 0.1288 0.004 Smoking, yes/no 3.6 0.3533 0.000 0.5 0.8760 0.000 Sex, female 3.6 0.3378 0.002 0.038 1.3 0.6885 0.000 0.121 Abbreviations: BMI, body mass index; bp, base pairs; LTL, leukocyte telomere length (expressed in bp); LTL , baseline LTL; LTL , follow-up B FU LTL. The rate of change in LTL (bp/year) ¼ (LTL  LTL )/follow-up years. FU B The rate of change in LTL is negative; therefore, negative regression coefficients denote faster LTL attrition. baseline examination and 23 years for follow-up because longer telomeres are bigger targets for free radicals, examinations) (20). which attack the G triplets on the telomeres. What, then, might be the reason for the dependency of The following 2 suppositions are at the heart of studies leukocyte telomere shortening on LTL itself? In cultured exploring the use of LTL as a biomarker of human aging: somatic cells, telomeres become shortened with cell divi- 1) oxidative stress and inflammation are key elements in the sion. The same apparently holds for somatic cells in vivo, biology of aging and in aging-related disorders (35, 36); and given that telomere length is substantially longer in skeletal 2) leukocyte telomere dynamics register the accruing burden muscle, a postmitotic tissue, than in proliferative tissues (21). of oxidative stress and inflammation over the life course of The inability of DNA polymerase to replicate nuclear DNA the individual (reviewed by Aviv (37)). Thus, the shortened to its terminus—the so-called end-replication problem— LTL displayed by individuals who suffer from aging-related would result in telomere shortening with somatic cell di- diseases presumably results from an accelerated rate of age- vision (22, 23), a process that is autonomous of telomere dependent LTL shortening. This presumption is based on length. However, both theoretical considerations and empir- findings that oxidative stress augments telomere attrition ical data suggest that the ‘‘end-replication problem’’ ac- per replication (31–34) and the inflammation heightens the counts for only a fraction of telomere shortening with turnover of leukocytes, which would further increase LTL replication (24–26). shortening with age. Such a concept supports the tenet that Telomere length is variable not only among different the shortened LTL in persons with atherosclerotic cardio- chromosomes (27) but also between homologous chromo- vascular disease (5, 9–12) is due to a protracted increase in somes (28), and it shortens faster in the homologous chro- the inflammatory and oxidative stress burdens (38, 39). mosome with the longer telomeres. This finding suggests The findings of the present study support previous obser- that telomere shortening might depend on telomere length. vations that LTL was shorter in smokers than in nonsmokers Telomere shortening that is dependent on telomere length (4, 6, 14) and was inversely correlated with BMI (4, 9, would maintain the length proportionality among different 13–15), presumably because smoking and obesity are asso- telomeres. In contrast, replication-mediated clipping of ciated with accelerated LTL shortening. However, as shown a fixed stretch of telomere repeats from telomeres of differ- in Table 2, we could not demonstrate a statistically signifi- ent lengths would result in loss of proportionality between cant difference in the rate of LTL shortening for smokers telomere length in a single chromosome and the mean telo- versus nonsmokers. One potential explanation might be that mere length of all chromosomes. This possibility does not our method was insufficiently sensitive to detect differences seem to be the case; proportionality between the length of in the rates of LTL shortening between smokers and non- telomeres in single chromosomes and the mean length of smokers, although the gap in LTL between the 2 groups did telomeres in all chromosomes is maintained not only in increase with age. Theoretically, LTL might be rapidly sperm (29) but also in somatic cells such as leukocytes from shortened after an individual starts smoking. Such rapid donors of a wide age range (30). The factor that might cause shortening would attenuate the rate of shortening thereafter proportional telomere shortening is oxidative stress (31–34) because of the proportionality of LTL loss to LTL itself. Am J Epidemiol 2009;169:323–329 328 Aviv et al. As indicated in our previous work (13), we are uncertain telomere dynamics in vivo and explain in part the faster rate as to the reason for the gain in LTL in a small subset of the of LTL shortening in African Americans than in whites. LTL cohort. We doubt that it relates to technical matters. Age- itself accounts for approximately 10% of the variation in LTL dependent LTL shortening would largely mirror telomere shortening, meaning that a host of other factors influence attrition in hematopoietic stem cells and progenitor cells, leukocyte telomere dynamics in both African Americans assuming that telomere shortening downstream of hemato- and whites. However, the effect of LTL on its own shortening poietic stem cells/progenitor cells is relatively constant. rate is sizable and requires careful consideration in longitu- However, this constancy might not hold all the time. For dinal studies that examine leukocyte telomere dynamics. instance, acute infection might transiently increase prolifera- Such studies might provide valuable information about the tion of peripheral mononuclear cells, resulting in a tempo- biology of human aging. rary increase in the difference between LTL and telomere length in hematopoietic stem cells/progenitor cells. This transient effect would not shorten by much telomere length ACKNOWLEDGMENTS in hematopoietic stem cells/progenitor cells over the life course of the individual, but it might affect LTL results in Author affiliations: The Center of Human Development a longitudinal study of short duration. Another alternative and Aging, University of Medicine and Dentistry of New relates to the status of hematopoietic stem cells within bone Jersey, New Jersey Medical School, Newark, New Jersey marrow niches. Evidently, hematopoietic stem cells residing (Abraham Aviv, Jeffrey P. Gardner, Masayuki Kimura, in the osteoblastic niche are largely quiescent; their mobi- Michael Brimacombe, Xiaojian Cao); and Tulane Center lization into the vascular niche, which might arise from for Cardiovascular Health, Tulane University Health Sciences a host of factors, promotes their transformation into prolif- Center, New Orleans, Louisiana (The Bogalusa Heart Study) erative hematopoietic stem cells (40). It might be possible (Wei Chen, Sathanur R. Srinivasan, Gerald S. Berenson). that, in a small subset of participants, a crop of quiescent Supported by R01 grants AG16592 and AG020132 from hematopoietic stem cells were mobilized into the vascular the National Institute on Aging. niche between baseline and follow-up examinations, reset- The authors thank Dr. Woodring Wright for his insightful ting LTL above its length at the baseline examination. Both suggestions. of these possibilities, if present, would not only account for Conflict of interest: none declared. LTL gain but also contribute to variation in LTL shortening during the follow-up period. The dependence of LTL shortening on LTL suggests that the rate of LTL shortening declines in older individuals. REFERENCES However, the factors that affect LTL dynamics in the in- 1. Okuda K, Bardeguez A, Gardner JP, et al. Telomere length in dividual are anything but constant throughout adulthood the newborn. Pediatr Res. 2002;52(3):377–381. because the precursors of inflammation and oxidative stress 2. Akkad A, Hastings R, Konje JC, et al. Telomere length in progressively increase with age (41–43). Individually or small-for-gestational-age babies. BJOG. 2006;113(3): jointly, inflammation and oxidative stress might therefore 318–323. sustain or even accelerate the rate of LTL shortening in 3. Slagboom PE, Droog S, Boomsma DI. Genetic determination the elderly, even though LTL progressively shortens with of telomere size in humans: a twin study of three age groups. age. Accordingly, only longitudinal evaluations of LTL Am J Hum Genet. 1994;55(5):876–882. shortening rates in different age groups and diverse sample 4. Valdes A, Andrew T, Gardner JP, et al. Obesity, cigarette smoking, and telomere length in women. Lancet. 2005; populations would provide a reliable appraisal of leukocyte 366(9486):662–664. telomere dynamics as humans get older. 5. Brouilette SW, Moore JS, McMahon AD, et al. Telomere Regarding potential regression to the mean effect, base- length, risk of coronary heart disease, and statin treatment in line telomere length was not relevant to the inclusion of the West of Scotland Primary Prevention Study: a nested case- subjects in the study. In terms of average change, any mask- control study. Lancet. 2007;369(9556):107–114. ing effect of LTL dynamics was not an issue because strong 6. Nawrot TS, Staessen JA, Gardner JP, et al. Telomere length biologic arguments support LTL shortening with age. The and possible link to X chromosome. Lancet. 2004;363(9408): potential latent effect of race may underlie some of our 507–510. results, but race was independent of inclusionary criteria 7. Njajou OT, Cawthon RM, Damcott CM, et al. Telomere length in the study and was used to further stratify the analysis. is paternally inherited and is associated with parental lifespan. Proc Natl Acad Sci U S A. 2007;104(29):12135–12139. In terms of induced correlation between baseline and change 8. Demissie S, Levy D, Benjamin EJ, et al. Insulin resistance, from baseline, age adjustment of only those changes from oxidative stress, hypertension, and leukocyte telomere length baseline, as well as the existing high level of variation in in men from the Framingham Heart Study. Aging Cell. 2006; LTL at baseline and changes from baseline, limited this 5(4):325–330. effect. The use of analysis of covariance also limited un- 9. Fitzpatrick AL, Kronmal RA, Gardner JP, et al. Leukocyte foreseen regression to the mean effects. telomere length and cardiovascular disease in the Cardiovas- In conclusion, this comprehensive longitudinal study of cular Health Study. Am J Epidemiol. 2007;165(1):14–21. age-dependent LTL shortening underscores the complexity 10. Benetos A, Gardner JP, Zureik M, et al. Short telomeres are of leukocyte telomere dynamics. Its findings support the role associated with increased carotid artery atherosclerosis in of mechanisms additional to the end-replication problem in hypertensive subjects. Hypertension. 2004;43(2):182–185. Am J Epidemiol 2009;169:323–329 Leukocyte Telomere Dynamics 329 11. Cawthon RM, Smith KR, O’Brien E, et al. Association be- 26. Proctor CJ, Kirkwood TB. Modeling telomere shortening and tween telomere length in blood and mortality in people aged the role of oxidative stress. Mech Ageing Dev. 2002;123(4): 60 years or older. Lancet. 2003;361(9355):393–395. 351–363. 12. van der Harst P, van der Steege G, de Boer RA, et al. Telomere 27. Martens UM, Zijlmans JM, Poon SS, et al. Short telomeres on length of circulating leukocytes is decreased in patients with human chromosome 17p. Nat Genet. 1998;18(1):76–80. chronic heart failure. J Am Coll Cardiol. 2007;49(13): 28. Goldman F, Bouarich R, Kulkarni S, et al. The effect of TERC 1459–1464. haploinsufficiency on the inheritance of telomere length. Proc 13. Gardner JP, Li S, Srinivasan SR, et al. Rise in insulin resistance Natl Acad Sci U S A. 2005;102(47):17119–17124. and gain in adiposity is associated with escalated telomere 29. Baird DM, Britt-Compton B, Rowson J, et al. Telomere in- attrition. Circulation. 2005;111(17):2171–2177. stability in the male germline. Hum Mol Genet. 2006;15(1): 14. Vasan RS, Demissie S, Kimura M, et al. Association of leu- 45–51. kocyte telomere length with circulating biomarkers of the 30. Kimura M, Barbieri M, Gardner JP, et al. Leukocytes of ex- renin-angiotensin-aldosterone system: the Framingham Heart ceptionally old persons display ultra-short telomeres. Am J Study. Circulation. 2008;117(4):1138–1144. Physiol Regul Integr Comp Physiol. 2007;293(6): 15. Hunt SC, Chen W, Gardner JP, et al. Leukocyte telomeres are R2210–R2217. longer in African Americans than in whites: the National 31. Forsyth NR, Evans AP, Shay JW, et al. Developmental dif- Heart, Lung, and Blood Institute Family Heart Study and the ferences in the immortalization of lung fibroblasts by telo- Bogalusa Heart Study. Aging Cell. 2008;7(4):451–458. merase. Aging Cell. 2003;2(5):235–243. 16. The Bogalusa Heart Study 20th Anniversary Symposium. Am 32. Richter T, von Zglinicki T. A continuous correlation between J Med Sci. 1995;310:S1–S138. oxidative stress and telomere shortening in fibroblasts. Exp 17. Frenck RW Jr, Blackburn EH, Shannon KM. The rate of Gerontol. 2007;42(11):1039–1042. telomere sequence loss in human leukocytes varies with age. 33. Tchirkov A, Lansdorp PM. Role of oxidative stress in telomere Proc Natl Acad Sci U S A. 1998;95(10):5607–5610. shortening in cultured fibroblasts from normal individuals and 18. Nordffja ¨ll K, Larefalk A, Limdgern P, et al. Telomere length patients with ataxia-telangiectasia. Hum Mol Genet. 2003; and heredity: indications of paternal inheritance. Proc Natl 12(3):227–232. Acad Sci U S A. 2005;102(45):16374–16378. 34. Sitte N, Saretzki G, von Zglinicki T. Accelerated telomere 19. Unryn BM, Cook LS, Riabowol KT. Paternal age is positively shortening in fibroblasts after extended periods of confluency. linked to telomere length of children. Aging Cell. 2005;4(2): Free Radic Biol Med. 1998;24(6):885–893. 97–101. 35. Balaban RS, Nemoto S, Finkel T. Mitochondria, oxidants, and 20. Aviv A, Valdes AM, Spector TD. Human telomere biology: aging. Cell. 2005;120(4):483–495. pitfalls of moving from the laboratory to epidemiology. Int J 36. Finch C, Crimmins EM. Inflammatory exposure and historical Epidemiol. 2006;35(6):1424–1429. changes in human life-spans. Science. 2004;305(5691): 21. Gardner JP, Kimura M, Chai W, et al. Telomere dynamics in 1736–1739. macaques and humans. J Gerontol A Biol Sci Med Sci. 2007; 37. Aviv A. Telomeres and human somatic fitness. J Gerontol A 62(4):367–374. Biol Sci Med Sci. 2006;61(8):871–873. 22. Olovnikov AM. A theory of marginotomy. The incomplete 38. Hansson GK, Libby P. The immune response in atheroscle- copying of template margin in enzymatic synthesis of poly- rosis; a double-edged sword. Nat Rev Immunol. 2006;6(7): nucleotides and biological significance of the phenomenon. 508–519. J Theor Biol. 1973;41(1):181–190. 39. Harrison D, Griendling KK, Landmesser U, et al. Role of 23. Watson ID. Origin of concatemeric T7 DNA. Nat New Biol. oxidative stress in atherosclerosis. Am J Cardiol. 2003;91(3A): 1972;239(94):197–201. 7A–11A. 24. op den Buijs J, van den Bosch PP, Musters MW, et al. 40. Yin T, Li L. The stem cell niches in bone. J Clin Invest. Mathematical modeling confirms the length-dependency 2006;116(5):1195–1201. of telomere shortening. Mech Ageing Dev. 2004;125(6): 41. Prelog M. Aging of the immune system: a risk factor for 437–444. autoimmunity? Autoimmun Rev. 2006;5(2):136–139. 25. op den Buijs J, Musters M, Verrips T, et al. Mathematical 42. Caruso C, Lio D, Cavallone L, et al. Aging, longevity, in- modeling of vascular endothelial layer maintenance: the role flammation, and cancer. Ann N Y Acad Sci. 2004;1028:1–13. of endothelial cell division, progenitor cell homing, and telo- 43. Dro ¨ ge W, Schipper HM. Oxidative stress and aberrant sig- mere shortening. Am J Physiol Heart Circ Physiol. 2007; naling in aging and cognitive decline. Aging Cell. 2007; 287(6):H2651–H2658. 6(3):361–370. Am J Epidemiol 2009;169:323–329 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Epidemiology Pubmed Central

Leukocyte Telomere Dynamics: Longitudinal Findings Among Young Adults in the Bogalusa Heart Study

Loading next page...
 
/lp/pubmed-central/leukocyte-telomere-dynamics-longitudinal-findings-among-young-adults-PM30hkQdbx

References (50)

Publisher
Pubmed Central
Copyright
American Journal of Epidemiology © 2008 The Authors
ISSN
0002-9262
eISSN
1476-6256
DOI
10.1093/aje/kwn338
Publisher site
See Article on Publisher Site

Abstract

American Journal of Epidemiology Vol. 169, No. 3 ª 2008 The Authors DOI: 10.1093/aje/kwn338 This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial Advance Access publication December 4, 2008 License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Original Contribution Leukocyte Telomere Dynamics: Longitudinal Findings Among Young Adults in the Bogalusa Heart Study Abraham Aviv, Wei Chen, Jeffrey P. Gardner, Masayuki Kimura, Michael Brimacombe, Xiaojian Cao, Sathanur R. Srinivasan, and Gerald S. Berenson Initially submitted April 30, 2008; accepted for publication September 22, 2008. Leukocyte telomere length (LTL) is ostensibly a biomarker of human aging. Cross-sectional analyses have found that LTL is relatively short in a host of aging-related diseases. These studies have also provided indirect estimates of age-dependent LTL shortening. In this paper, the authors report findings of the first comprehensive longitudinal study of 450 whites and 185 African Americans in Louisiana (aged 31.4 and 37.4 years at baseline (1995–1996) and follow-up (2001–2006) examinations, respectively) participating in the Bogalusa Heart Study. Rate of change in LTL was highly variable among individuals, with some displaying a paradoxical gain in LTL during the follow-up period. The most striking observation was that age-dependent LTL shortening was proportional to LTL at baseline examination. At both baseline and follow-up examinations, African Americans had longer LTLs than whites, and smokers had shorter LTLs than nonsmokers. The longer LTL in African Americans than in whites explained in part the faster rate of LTL shortening observed among African Americans. These findings underscore the complexity of leukocyte telomere dynamics in vivo and suggest that determinants in addition to the ‘‘end-replication problem’’ contribute to telomere shortening in vivo. aging; body mass index; leukocytes; oxidative stress; smoking; telomere Abbreviations: BMI, body mass index; bp, base pairs; LTL, leukocyte telomere length. Leukocyte telomere length (LTL) is highly variable at shortening was faster, in adult African Americans than in birth (1, 2) and afterward (3–9). In cross-sectional analyses whites (15). For these reasons, we examined in a longitudinal based primarily on comparisons of LTL in individuals of study the factors that contribute to variation in LTL shorten- different ages, it seems that LTL progressively shortens lin- ing among young African-American and white adults in the early with age (3–9). However, cross-sectional analyses Bogalusa Heart Study (16). Our findings indicate that age- hardly capture the rate and trajectory of age-dependent dependent LTL shortening is proportional to LTL. LTL shortening in an individual, as well as variations in these parameters among individuals. Deciphering patterns MATERIALS AND METHODS of LTL shortening and variation in the rate of LTL shorten- ing with age among individuals would be highly relevant Participants because shortened LTL has often been detected in individu- als with aging-related disorders, cardiovascular disease in The Bogalusa Heart Study is a long-term investigation of particular (5, 9–12). Factors that predispose to cardiovascu- the natural history of cardiovascular disease beginning in lar disease, including obesity (4, 13, 14), insulin resistance childhood in the biracial community (65% white, 35% (8, 13), and cigarette smoking (4, 6, 14), are also associated African American) of Bogalusa, Louisiana (16). Between with shortened LTL. Moreover, a recent cross-sectional 1973 and 2006, 7 cross-sectional surveys of children aged analysis observed that LTL was longer, but its age-dependent 4–17 years and 9 cross-sectional surveys of adults aged Correspondence to Prof. Abraham Aviv, The Center of Human Development and Aging, Room F-464, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103 (e-mail: avivab@umdnj.edu). 323 Am J Epidemiol 2009;169:323–329 324 Aviv et al. 18–46 years, who had been previously examined as chil- Statistical analysis dren, were conducted. In the ongoing Bogalusa Heart Study Mean values and standard deviations were computed for cohort, 1,203 adults aged 24–44 years were examined for baseline and follow-up LTL and other continuous covariate cardiovascular risk factors between 2004 and 2006. LTL variables separately for African Americans and whites. data were available for 635 participants, who had blood LTLs adjusted for age and body mass index (BMI) were samples collected on 2 occasions—a baseline examination calculated as least-squares means for smokers and non- in 1995–1996 and a follow-up examination in 2001–2006. smokers by holding age and/or BMI to their mean values. Compared with individuals not included, subjects with lon- Analysis of covariance was performed by using generalized gitudinal LTL measurements showed no significant differ- linear models to test differences in baseline and follow-up ences in the relevant study variables, except that they were LTL between African Americans and whites and between 0.8 years older (P ¼ 0.032). Self-administered question- smokers and nonsmokers. The relation between LTL attri- naires were used to obtain information on cigarette smok- tion and baseline LTL was examined by linear regression ing; those who smoked at least 1 cigarette per week during models using rate of change in LTL as the dependent vari- the past 12 months were defined as current smokers. The able, adjusting for race, age, sex, BMI, and smoking. Sta- study protocol was approved by the institutional review tistical analyses were performed by using SAS version 9.0 board of Tulane University (New Orleans, Louisiana), and software (SAS Institute, Inc., Cary, North Carolina). Three each participant gave written informed consent. outliers (2 African-American women and 1 white man) whose rate of LTL shortening was more than 250 base pairs LTL measurements (bp)/year or LTL lengthening was more than 250 bp/year were excluded from the analyses. DNA from frozen blood samples, obtained at baseline and follow-up examinations, was extracted by using the Gentra Puregene Blood Kit (QIAGEN Inc., Valencia, California). DNA integrity was assessed following electrophoresis of RESULTS 0.02 lg of DNA (200 V for 0.75 hours) on 1% agarose gels and staining with 2X SYBR Green I (Invitrogen Corpora- General characteristics tion, Carlsbad, California) for 45 minutes. Telomere-length measurements were performed by ob- Major characteristics of the cohort are summarized in taining the mean length of the terminal restriction fragments, Tables 1–3. The age ranges were 20.0–40.0 years at baseline using the Southern blot method, as previously described and 25.7–48.2 years at follow-up examinations. Smokers (1, 13). In brief, after overnight DNA digestion with 10 U constituted approximately a third of the cohort. African Hinf I and 10 U Rsa I restriction enzymes (Roche Diagnos- Americans had a higher BMI than whites. Compared with tics Corporation, Indianapolis, Indiana), samples of DNA whites, African Americans had longer LTLs at baseline (3 lg each) were resolved on 0.5% agarose gels at 50 V and follow-up examinations and a higher rate of LTL short- for 16 hours. DNA was then depurinated in 0.25 N hydro- ening. The follow-up period was slightly shorter for African chloric acid, denatured in 0.5 mol/L of sodium hydroxide Americans than for whites (5.7 years vs. 6.0 years), but this per 1.5 mol/L of sodium chloride, and neutralized in difference was not significant (P ¼ 0.1180). Table 2 sum- 0.5 mol/L of tris(hydroxymethyl)aminomethane (pH 8) marizes LTL parameters for smokers versus nonsmokers. per 1.5 mol/L of sodium chloride. After transfer to a posi- Age-, BMI-, and race-adjusted LTL was significantly shorter in smokers versus nonsmokers by 141 bp (P ¼ tively charged nylon membrane, DNA was hybridized to 0.0227) at baseline examination and by 164 bp (P ¼ a digoxigenin 3#-end labeled 5#-(CCCTAA) telomeric probe (overnight at 65C), washed 3 times in 2X saline- 0.0099) at follow-up examination. However, we found no sodium citrate/0.1% of sodium dodecyl sulfate, and washed significant difference in the rate of LTL shortening between once in 2X saline-sodium citrate. Probe was detected by the smokers and nonsmokers. digoxigenin luminescent detection procedure (Roche Diag- We next examined the contribution of race, age, smoking, nostics) after exposure on Hyperfilm (GE Healthcare, Chalfont BMI, and sex to variations in LTL at baseline and follow-up St. Giles, United Kingdom). Each sample was resolved in examinations. Table 3 shows that race was the main factor duplicate on different gels, and autoradiograms were explaining variations in LTL at baseline and follow-up ex- digitized for analysis of telomere-length measurement. aminations and that smoking and BMI also accounted for Baseline and follow-up terminal restriction fragments some of the variations. from the same individuals were resolved in adjacent lanes Figure 1 displays the distribution of rate of change in LTL of each gel, with samples run in duplicate on different gels. among participants. This parameter varied considerably. The intraclass correlation coefficient of LTL was 0.972 (P < The majority of individuals displayed LTL shortening (loss) 0.0001), and the coefficient of variation for the duplicate (85.9% of African Americans and 88.0% of whites). In ad- samples was 1.4%. The laboratory that conducted the LTL dition, 1 African American and 5 whites showed no change measurements was blinded to the identity and characteris- in LTL between baseline and follow-up examinations. How- tics of subjects. The rate of change (loss or gain) in LTL was ever, a subset of the cohort showed LTL gain. The means for computed as the difference between LTL value at follow-up the rates of change for African Americans and whites are minus that at baseline examinations divided by the number displayed in Table 1. For the entire cohort, the mean rate of of years of follow-up. change was 40.7 bp/year and was higher in African Am J Epidemiol 2009;169:323–329 Leukocyte Telomere Dynamics 325 Table 1. Major Characteristics of the Cohort of Young Adults in the Bogalusa Heart Study in Louisiana at Baseline (1995–1996) and Follow-up (2001–2006) Examinations Whites African Americans All P Value Parameter (n 5 450) (n 5 185) (n 5 635) (Race) Baseline examination Age, years 31.4 (5.0) 31.4 (5.3) 31.4 (5.1) 0.7904 BMI, kg/m 27.1 (6.4) 30.5 (8.4) 28.1 (5.1) <0.0001 Smokers, % 32.4 36.8 33.7 0.2962 LTL , bp 7,288 (735) 7,847 (734) 7,451 (777) <0.0001 Follow-up examination Age, years 37.4 (4.4) 37.0 (4.7) 37.3 (4.5) 0.5823 BMI, kg/m 28.3 (6.6) 32.0 (8.9) 29.4 (7.5) <0.0001 Smokers, % 30.7 30.8 30.7 0.9715 LTL , bp 7,076 (721) 7,603 (767) 7,230 (772) <0.0001 FU Change in LTL Follow-up, years 6.0 (2.4) 5.7 (2.5) 5.9 (2.4) 0.1180 LTL, bp/year 37.8 (41.3) 47.7 (55.3) 40.7 (46.0) 0.0181 Abbreviations: BMI, body mass index; bp, base pairs; LTL, leukocyte telomere length (expressed in bp); LTL , baseline LTL; LTL , follow-up LTL. B FU Unless otherwise specified, data are presented as mean (standard deviation). P values for race difference in LTL and change in LTL were adjusted for age, sex, and BMI. Rate of change in LTL (bp/year) ¼ (LTL – LTL )/follow-up years. FU B Americans than in whites (47.7 bp/year vs. 37.8 bp/year, (standard deviation, 47.6) bp/year and that for whites was respectively; P ¼ 0.0181). 45.9 (standard deviation, 34.7) bp/year (P ¼ 0.0003). Given that different mechanisms are likely to govern LTL Figure 2 displays the relation between the rate of LTL short- shortening versus gain, we separately examined LTL dy- ening that took place during the follow-up period and LTL namics in individuals who did not gain LTL during the derived from the baseline samples. A strong association was follow-up period. For these individuals (n ¼ 561), the observed between the rate of LTL shortening and baseline LTL shortening rate for African Americans was 60.0 LTL (a 17-bp/year increase in the rate of LTL shortening for each kilobase of LTL). To evaluate the potential influence of measurement error Table 2. Leukocyte Telomere Length Parameters of Interest in on the association of LTL and its rate of change, we set Smokers Versus Nonsmokers Participating in the Bogalusa Heart a conservative limit of quantification defined as the absolute Study in Louisiana at Baseline (1995–1996) and Follow-up (2001– value of the rate of change in LTL of less than 20 bp/year to 2006) Examinations denote detectable change. The association between baseline LTL Parameter Nonsmokers Smokers P Value LTL and the rate of change was then examined for the sub- group with smaller (<20 bp/year loss or gain, n ¼ 146) and Unadjusted LTL larger (20 bp/year loss or gain, n ¼ 489) changes in LTL. LTL , bp 7,481 7,392 0.0660 For the smaller-change group, the association was not sig- LTL , bp 7,270 7,150 0.0219 FU nificant (b ¼0.05, P ¼ 0.956), but, in the group with Age-adjusted LTL a larger change, the regression coefficient was significant LTL , bp 7,492 7,370 0.0479 B (b ¼13.7, P < 0.0001). This subgroup analysis provides confidence that measurement error did not play a significant LTL , bp 7,275 7,128 0.0196 FU role in the association noted in this study. Age- and BMI-adjusted LTL Since African Americans had longer LTL at baseline and LTL , bp 7,500 7,359 0.0227 follow-up examinations and a higher rate of LTL shortening, LTL , bp 7,280 7,116 0.0099 FU we examined the race effect and other pertinent factors on Change, bp/year 40 42 0.4630 the variations in LTL rate of change for the entire cohort and separately for those individuals who demonstrated no gain Abbreviations: BMI, body mass index; bp, base pairs; LTL, leuko- in LTL during the follow-up period. The findings are sum- cyte telomere length (expressed in bp); LTL , baseline LTL; LTL , B FU marized in Table 4. When considered alone, race signifi- follow-up LTL. cantly contributed to the variations in rate of change in Rate of change in LTL (bp/year) ¼ (LTL  LTL )/follow-up FU B years. LTL. However, when race and LTL at baseline examination Am J Epidemiol 2009;169:323–329 326 Aviv et al. Table 3. Regression of Baseline (1995–1996) and Follow-up (2001–2006) Leukocyte Telomere Length on Parameters of Interest for Participants in the Bogalusa Heart Study in Louisiana Baseline LTL Follow-up LTL Parameter 2 2 2 2 b Coefficient P Value Partial R Model R b Coefficient P Value Partial R Model R Race, African American 587.8 <0.0001 0.107 551.5 <0.0001 0.096 Age, years 13.4 0.0220 0.009 11.7 0.0715 0.005 BMI, kg/m 9.1 0.0316 0.006 8.6 0.0312 0.007 Smoking, yes/no 141.6 0.0227 0.006 163.5 0.0099 0.008 Sex, female 56.8 0.3456 0.002 0.130 37.0 0.5372 0.000 0.116 Abbreviations: BMI, body mass index; LTL, leukocyte telomere length (expressed in base pairs (bp)). The rate of change in LTL (bp/year) is negative; therefore, negative regression coefficients denote faster LTL attrition. were analyzed jointly, only LTL at baseline examination follow-up period (13). A previous cross-sectional analysis accounted significantly for the rate of change in LTL, mean- of LTL in a combined sample of participants from the ing that the longer LTL in African Americans was a major Bogalusa Heart Study and the NHLBI Family Heart Study factor in the higher rate of LTL shortening in African (age range, 19–93 years) found not only that LTL was longer Americans versus whites. Adding age, BMI, smoking, and in African Americans than in whites but also that the rate of sex at baseline examination to the model provided little age-dependent LTL shortening was faster in African explanation above that of baseline LTL for interindividual Americans (15). The present work indicates that the longer variation in LTL shortening rate. These observations held LTL in African Americans partially explained their faster when the data were analyzed for the entire cohort or after age-dependent LTL shortening compared with that of whites. exclusion of individuals who gained LTL between baseline Cross-sectional analyses of leukocyte telomere dynamics and follow-up examinations. have suggested that age-dependent LTL shortening might be slower (17, 18) or faster (19) in young than in older adults, but these studies were statistically underpowered to test DISCUSSION either of these suppositions (20). Note that, in the present study, cross-sectional analysis revealed only a small effect Our key finding in this biracial cohort of young adults was of age on LTL attrition (Table 3), but this finding was due to that age-dependent LTL shortening rate was proportional to the considerable interindividual variation in LTL and the LTL. In addition, we confirmed previous observations in relatively narrow age range of the cohort (20 years for a small sample of the Bogalusa Heart Study with a 10-year follow-up showing that the rate of LTL shortening was highly variable among individuals, with a subset of partic- ipants displaying paradoxical lengthening of LTL during the Figure 1. Distribution of the rate of change in leukocyte telomere length (LTL) in Louisiana study participants. A total of 561 participants Figure 2. Relation between rate of leukocyte telomere length (LTL) showed LTL shortening, and 74 showed either gain (n ¼ 68) or no shortening during follow-up (2001–2006) and LTL derived from the change (n ¼ 6) in LTL between baseline (1995–1996) and follow-up baseline samples (1995–1996) in Louisiana study participants. D LTL¼ (2001–2006) examinations. bp, base pairs. 0.017 baseline LTL – 0.073; r ¼ 0.326, P < 0.0001. bp, base pairs. Am J Epidemiol 2009;169:323–329 Leukocyte Telomere Dynamics 327 Table 4. Regression of the Rate of Change in Leukocyte Telomere Length on Parameters of Interest for Participants in the Bogalusa Heart Study a,b in Louisiana Entire Cohort (n 5 635) Subjects With no Gain in LTL (n 5 561) 2 2 2 2 b Coefficient P Value Partial R Model R b Coefficient P Value Partial R Model R Model I Race, African American 9.9 0.0135 0.010 0.010 14.1 <0.0001 0.026 0.026 Model II Race, African American 4.6 0.2772 0.010 5.9 0.1070 0.026 LTL ,bp 9.6 <0.0001 0.023 0.033 15.6 <0.0001 0.084 0.110 Model III Race, African American 3.4 0.4309 0.010 4.4 0.2424 0.026 LTL ,bp 9.9 <0.0001 0.023 16.1 <0.0001 0.084 Age, years 0.6 0.1120 0.003 0.6 0.0605 0.007 BMI, kg/m 0.2 0.5464 0.000 0.3 0.1288 0.004 Smoking, yes/no 3.6 0.3533 0.000 0.5 0.8760 0.000 Sex, female 3.6 0.3378 0.002 0.038 1.3 0.6885 0.000 0.121 Abbreviations: BMI, body mass index; bp, base pairs; LTL, leukocyte telomere length (expressed in bp); LTL , baseline LTL; LTL , follow-up B FU LTL. The rate of change in LTL (bp/year) ¼ (LTL  LTL )/follow-up years. FU B The rate of change in LTL is negative; therefore, negative regression coefficients denote faster LTL attrition. baseline examination and 23 years for follow-up because longer telomeres are bigger targets for free radicals, examinations) (20). which attack the G triplets on the telomeres. What, then, might be the reason for the dependency of The following 2 suppositions are at the heart of studies leukocyte telomere shortening on LTL itself? In cultured exploring the use of LTL as a biomarker of human aging: somatic cells, telomeres become shortened with cell divi- 1) oxidative stress and inflammation are key elements in the sion. The same apparently holds for somatic cells in vivo, biology of aging and in aging-related disorders (35, 36); and given that telomere length is substantially longer in skeletal 2) leukocyte telomere dynamics register the accruing burden muscle, a postmitotic tissue, than in proliferative tissues (21). of oxidative stress and inflammation over the life course of The inability of DNA polymerase to replicate nuclear DNA the individual (reviewed by Aviv (37)). Thus, the shortened to its terminus—the so-called end-replication problem— LTL displayed by individuals who suffer from aging-related would result in telomere shortening with somatic cell di- diseases presumably results from an accelerated rate of age- vision (22, 23), a process that is autonomous of telomere dependent LTL shortening. This presumption is based on length. However, both theoretical considerations and empir- findings that oxidative stress augments telomere attrition ical data suggest that the ‘‘end-replication problem’’ ac- per replication (31–34) and the inflammation heightens the counts for only a fraction of telomere shortening with turnover of leukocytes, which would further increase LTL replication (24–26). shortening with age. Such a concept supports the tenet that Telomere length is variable not only among different the shortened LTL in persons with atherosclerotic cardio- chromosomes (27) but also between homologous chromo- vascular disease (5, 9–12) is due to a protracted increase in somes (28), and it shortens faster in the homologous chro- the inflammatory and oxidative stress burdens (38, 39). mosome with the longer telomeres. This finding suggests The findings of the present study support previous obser- that telomere shortening might depend on telomere length. vations that LTL was shorter in smokers than in nonsmokers Telomere shortening that is dependent on telomere length (4, 6, 14) and was inversely correlated with BMI (4, 9, would maintain the length proportionality among different 13–15), presumably because smoking and obesity are asso- telomeres. In contrast, replication-mediated clipping of ciated with accelerated LTL shortening. However, as shown a fixed stretch of telomere repeats from telomeres of differ- in Table 2, we could not demonstrate a statistically signifi- ent lengths would result in loss of proportionality between cant difference in the rate of LTL shortening for smokers telomere length in a single chromosome and the mean telo- versus nonsmokers. One potential explanation might be that mere length of all chromosomes. This possibility does not our method was insufficiently sensitive to detect differences seem to be the case; proportionality between the length of in the rates of LTL shortening between smokers and non- telomeres in single chromosomes and the mean length of smokers, although the gap in LTL between the 2 groups did telomeres in all chromosomes is maintained not only in increase with age. Theoretically, LTL might be rapidly sperm (29) but also in somatic cells such as leukocytes from shortened after an individual starts smoking. Such rapid donors of a wide age range (30). The factor that might cause shortening would attenuate the rate of shortening thereafter proportional telomere shortening is oxidative stress (31–34) because of the proportionality of LTL loss to LTL itself. Am J Epidemiol 2009;169:323–329 328 Aviv et al. As indicated in our previous work (13), we are uncertain telomere dynamics in vivo and explain in part the faster rate as to the reason for the gain in LTL in a small subset of the of LTL shortening in African Americans than in whites. LTL cohort. We doubt that it relates to technical matters. Age- itself accounts for approximately 10% of the variation in LTL dependent LTL shortening would largely mirror telomere shortening, meaning that a host of other factors influence attrition in hematopoietic stem cells and progenitor cells, leukocyte telomere dynamics in both African Americans assuming that telomere shortening downstream of hemato- and whites. However, the effect of LTL on its own shortening poietic stem cells/progenitor cells is relatively constant. rate is sizable and requires careful consideration in longitu- However, this constancy might not hold all the time. For dinal studies that examine leukocyte telomere dynamics. instance, acute infection might transiently increase prolifera- Such studies might provide valuable information about the tion of peripheral mononuclear cells, resulting in a tempo- biology of human aging. rary increase in the difference between LTL and telomere length in hematopoietic stem cells/progenitor cells. This transient effect would not shorten by much telomere length ACKNOWLEDGMENTS in hematopoietic stem cells/progenitor cells over the life course of the individual, but it might affect LTL results in Author affiliations: The Center of Human Development a longitudinal study of short duration. Another alternative and Aging, University of Medicine and Dentistry of New relates to the status of hematopoietic stem cells within bone Jersey, New Jersey Medical School, Newark, New Jersey marrow niches. Evidently, hematopoietic stem cells residing (Abraham Aviv, Jeffrey P. Gardner, Masayuki Kimura, in the osteoblastic niche are largely quiescent; their mobi- Michael Brimacombe, Xiaojian Cao); and Tulane Center lization into the vascular niche, which might arise from for Cardiovascular Health, Tulane University Health Sciences a host of factors, promotes their transformation into prolif- Center, New Orleans, Louisiana (The Bogalusa Heart Study) erative hematopoietic stem cells (40). It might be possible (Wei Chen, Sathanur R. Srinivasan, Gerald S. Berenson). that, in a small subset of participants, a crop of quiescent Supported by R01 grants AG16592 and AG020132 from hematopoietic stem cells were mobilized into the vascular the National Institute on Aging. niche between baseline and follow-up examinations, reset- The authors thank Dr. Woodring Wright for his insightful ting LTL above its length at the baseline examination. Both suggestions. of these possibilities, if present, would not only account for Conflict of interest: none declared. LTL gain but also contribute to variation in LTL shortening during the follow-up period. The dependence of LTL shortening on LTL suggests that the rate of LTL shortening declines in older individuals. REFERENCES However, the factors that affect LTL dynamics in the in- 1. Okuda K, Bardeguez A, Gardner JP, et al. Telomere length in dividual are anything but constant throughout adulthood the newborn. Pediatr Res. 2002;52(3):377–381. because the precursors of inflammation and oxidative stress 2. Akkad A, Hastings R, Konje JC, et al. Telomere length in progressively increase with age (41–43). Individually or small-for-gestational-age babies. BJOG. 2006;113(3): jointly, inflammation and oxidative stress might therefore 318–323. sustain or even accelerate the rate of LTL shortening in 3. Slagboom PE, Droog S, Boomsma DI. Genetic determination the elderly, even though LTL progressively shortens with of telomere size in humans: a twin study of three age groups. age. Accordingly, only longitudinal evaluations of LTL Am J Hum Genet. 1994;55(5):876–882. shortening rates in different age groups and diverse sample 4. Valdes A, Andrew T, Gardner JP, et al. Obesity, cigarette smoking, and telomere length in women. Lancet. 2005; populations would provide a reliable appraisal of leukocyte 366(9486):662–664. telomere dynamics as humans get older. 5. Brouilette SW, Moore JS, McMahon AD, et al. Telomere Regarding potential regression to the mean effect, base- length, risk of coronary heart disease, and statin treatment in line telomere length was not relevant to the inclusion of the West of Scotland Primary Prevention Study: a nested case- subjects in the study. In terms of average change, any mask- control study. Lancet. 2007;369(9556):107–114. ing effect of LTL dynamics was not an issue because strong 6. Nawrot TS, Staessen JA, Gardner JP, et al. Telomere length biologic arguments support LTL shortening with age. The and possible link to X chromosome. Lancet. 2004;363(9408): potential latent effect of race may underlie some of our 507–510. results, but race was independent of inclusionary criteria 7. Njajou OT, Cawthon RM, Damcott CM, et al. Telomere length in the study and was used to further stratify the analysis. is paternally inherited and is associated with parental lifespan. Proc Natl Acad Sci U S A. 2007;104(29):12135–12139. In terms of induced correlation between baseline and change 8. Demissie S, Levy D, Benjamin EJ, et al. Insulin resistance, from baseline, age adjustment of only those changes from oxidative stress, hypertension, and leukocyte telomere length baseline, as well as the existing high level of variation in in men from the Framingham Heart Study. Aging Cell. 2006; LTL at baseline and changes from baseline, limited this 5(4):325–330. effect. The use of analysis of covariance also limited un- 9. Fitzpatrick AL, Kronmal RA, Gardner JP, et al. Leukocyte foreseen regression to the mean effects. telomere length and cardiovascular disease in the Cardiovas- In conclusion, this comprehensive longitudinal study of cular Health Study. Am J Epidemiol. 2007;165(1):14–21. age-dependent LTL shortening underscores the complexity 10. Benetos A, Gardner JP, Zureik M, et al. Short telomeres are of leukocyte telomere dynamics. Its findings support the role associated with increased carotid artery atherosclerosis in of mechanisms additional to the end-replication problem in hypertensive subjects. Hypertension. 2004;43(2):182–185. Am J Epidemiol 2009;169:323–329 Leukocyte Telomere Dynamics 329 11. Cawthon RM, Smith KR, O’Brien E, et al. Association be- 26. Proctor CJ, Kirkwood TB. Modeling telomere shortening and tween telomere length in blood and mortality in people aged the role of oxidative stress. Mech Ageing Dev. 2002;123(4): 60 years or older. Lancet. 2003;361(9355):393–395. 351–363. 12. van der Harst P, van der Steege G, de Boer RA, et al. Telomere 27. Martens UM, Zijlmans JM, Poon SS, et al. Short telomeres on length of circulating leukocytes is decreased in patients with human chromosome 17p. Nat Genet. 1998;18(1):76–80. chronic heart failure. J Am Coll Cardiol. 2007;49(13): 28. Goldman F, Bouarich R, Kulkarni S, et al. The effect of TERC 1459–1464. haploinsufficiency on the inheritance of telomere length. Proc 13. Gardner JP, Li S, Srinivasan SR, et al. Rise in insulin resistance Natl Acad Sci U S A. 2005;102(47):17119–17124. and gain in adiposity is associated with escalated telomere 29. Baird DM, Britt-Compton B, Rowson J, et al. Telomere in- attrition. Circulation. 2005;111(17):2171–2177. stability in the male germline. Hum Mol Genet. 2006;15(1): 14. Vasan RS, Demissie S, Kimura M, et al. Association of leu- 45–51. kocyte telomere length with circulating biomarkers of the 30. Kimura M, Barbieri M, Gardner JP, et al. Leukocytes of ex- renin-angiotensin-aldosterone system: the Framingham Heart ceptionally old persons display ultra-short telomeres. Am J Study. Circulation. 2008;117(4):1138–1144. Physiol Regul Integr Comp Physiol. 2007;293(6): 15. Hunt SC, Chen W, Gardner JP, et al. Leukocyte telomeres are R2210–R2217. longer in African Americans than in whites: the National 31. Forsyth NR, Evans AP, Shay JW, et al. Developmental dif- Heart, Lung, and Blood Institute Family Heart Study and the ferences in the immortalization of lung fibroblasts by telo- Bogalusa Heart Study. Aging Cell. 2008;7(4):451–458. merase. Aging Cell. 2003;2(5):235–243. 16. The Bogalusa Heart Study 20th Anniversary Symposium. Am 32. Richter T, von Zglinicki T. A continuous correlation between J Med Sci. 1995;310:S1–S138. oxidative stress and telomere shortening in fibroblasts. Exp 17. Frenck RW Jr, Blackburn EH, Shannon KM. The rate of Gerontol. 2007;42(11):1039–1042. telomere sequence loss in human leukocytes varies with age. 33. Tchirkov A, Lansdorp PM. Role of oxidative stress in telomere Proc Natl Acad Sci U S A. 1998;95(10):5607–5610. shortening in cultured fibroblasts from normal individuals and 18. Nordffja ¨ll K, Larefalk A, Limdgern P, et al. Telomere length patients with ataxia-telangiectasia. Hum Mol Genet. 2003; and heredity: indications of paternal inheritance. Proc Natl 12(3):227–232. Acad Sci U S A. 2005;102(45):16374–16378. 34. Sitte N, Saretzki G, von Zglinicki T. Accelerated telomere 19. Unryn BM, Cook LS, Riabowol KT. Paternal age is positively shortening in fibroblasts after extended periods of confluency. linked to telomere length of children. Aging Cell. 2005;4(2): Free Radic Biol Med. 1998;24(6):885–893. 97–101. 35. Balaban RS, Nemoto S, Finkel T. Mitochondria, oxidants, and 20. Aviv A, Valdes AM, Spector TD. Human telomere biology: aging. Cell. 2005;120(4):483–495. pitfalls of moving from the laboratory to epidemiology. Int J 36. Finch C, Crimmins EM. Inflammatory exposure and historical Epidemiol. 2006;35(6):1424–1429. changes in human life-spans. Science. 2004;305(5691): 21. Gardner JP, Kimura M, Chai W, et al. Telomere dynamics in 1736–1739. macaques and humans. J Gerontol A Biol Sci Med Sci. 2007; 37. Aviv A. Telomeres and human somatic fitness. J Gerontol A 62(4):367–374. Biol Sci Med Sci. 2006;61(8):871–873. 22. Olovnikov AM. A theory of marginotomy. The incomplete 38. Hansson GK, Libby P. The immune response in atheroscle- copying of template margin in enzymatic synthesis of poly- rosis; a double-edged sword. Nat Rev Immunol. 2006;6(7): nucleotides and biological significance of the phenomenon. 508–519. J Theor Biol. 1973;41(1):181–190. 39. Harrison D, Griendling KK, Landmesser U, et al. Role of 23. Watson ID. Origin of concatemeric T7 DNA. Nat New Biol. oxidative stress in atherosclerosis. Am J Cardiol. 2003;91(3A): 1972;239(94):197–201. 7A–11A. 24. op den Buijs J, van den Bosch PP, Musters MW, et al. 40. Yin T, Li L. The stem cell niches in bone. J Clin Invest. Mathematical modeling confirms the length-dependency 2006;116(5):1195–1201. of telomere shortening. Mech Ageing Dev. 2004;125(6): 41. Prelog M. Aging of the immune system: a risk factor for 437–444. autoimmunity? Autoimmun Rev. 2006;5(2):136–139. 25. op den Buijs J, Musters M, Verrips T, et al. Mathematical 42. Caruso C, Lio D, Cavallone L, et al. Aging, longevity, in- modeling of vascular endothelial layer maintenance: the role flammation, and cancer. Ann N Y Acad Sci. 2004;1028:1–13. of endothelial cell division, progenitor cell homing, and telo- 43. Dro ¨ ge W, Schipper HM. Oxidative stress and aberrant sig- mere shortening. Am J Physiol Heart Circ Physiol. 2007; naling in aging and cognitive decline. Aging Cell. 2007; 287(6):H2651–H2658. 6(3):361–370. Am J Epidemiol 2009;169:323–329

Journal

American Journal of EpidemiologyPubmed Central

Published: Dec 4, 2008

There are no references for this article.