Abstract

Background Shorter leucocyte telomere length (LTL) is a promising marker of biological ageing. It is predicted by cumulative adverse conditions throughout life course, but few studies have data from the prenatal period when most developmental processes and cell replication take place. We studied whether body size at birth and underlying factors including severely preterm birth predict LTL in adult life.

Methods We used data from following three cohorts: (i) 1894 subjects (age: 56–69 years) from the Helsinki Birth Cohort Study (HBCS), representing normal variation in fetal growth; (ii) the Helsinki Study of Very Low Birth Weight Adults encompassing 164 subjects born preterm at very low birthweight (<1500 g; representing extreme pre- and neonatal conditions) and 170 term-born controls (18–27 years) and (iii) 248 twins (23–31 years) from the FinnTwin16 cohort, allowing comparisons between twin pairs. Relative telomere length was measured from leucocytes by real-time quantitative polymerase chain reaction.

Results Shorter LTL was associated with higher age in HBCS and among men in the Helsinki Study of Very Low Birth Weight Adults and with lower childhood socio-economic status in HBCS and FinnTwin16. LTL was not associated with weight, length or gestational age at birth in any cohort. LTL was similar in very-low-birthweight and control subjects.

Conclusions LTL is unlikely to be a useful marker of a mechanism linking body size at birth with individual differences in ageing in the general population.

Introduction

Telomeres consist of DNA repeats and associated proteins located at the ends of chromosomes. At each cell division, telomeres get shorter, unless their length is maintained by telomerase enzyme. Because telomerase activity in most somatic cells is low, telomere length carries the replication history of the cell.1,2 Leucocyte telomere length (LTL) has been suggested to serve as a biomarker capturing cumulative lifelong burden of adverse conditions associated with advanced cellular ageing, in particular conditions acting through oxidative stress and inflammation,3,4 two key elements of ageing-related processes.5,6 Indeed, one of the most consistent findings in epidemiological studies is the inverse association between LTL and age.2

A relationship between shorter LTL and individual differences in ageing is further supported by its associations with insulin resistance,3,7 type 18 and type 29,10 diabetes, dementia,11 markers of oxidative distress3 and the risk of developing cardiovascular disease.1215 Moreover, according to some,16,17 although not all,15,1820 studies, LTL predicts all-cause mortality.

The chain of evidence linking LTL with lifelong burden of adverse conditions is incomplete. A number of studies do support a relationship between cumulative lifelong psychosocial stress and shorter adult LTL,2124 but most studies are limited to data from childhood or adulthood onwards. Yet, most of the replication history of a cell takes place in the fetus, which undergoes some 42 successive mitotic divisions in progressing from a fertilized ovum to a term infant, with only five more divisions being necessary to achieve adult size.25 Adverse conditions during fetal period are well-known predictors of ageing-related disorders such as cardiovascular disease,26,27 type 2 diabetes28,29 and possibly cognitive impairments30,31 and their risk factors. Should LTL be a cumulative biomarker, it would also be expected to be associated with circumstances during the prenatal period. This has been suggested by two case–control studies. Among 132 Bangladeshi children aged 5 years, those with low birthweight had shorter LTL than those with normal birthweight.32 Shorter telomeres were also observed among 94 25-year-old Western Europeans who had as fetuses been exposed to severe maternal stress.33 In that study, shorter telomeres were also associated with lower birthweight. However, these findings await replication.

With this background, we hypothesized that adverse conditions during the fetal period are associated with shorter LTL in adulthood. We tested this in three complementary cohorts. The Helsinki Birth Cohort Study (HBCS) is a community sample of individuals (examined at an age of ∼60 years) encompassing the whole range of perinatal conditions, which allows us to assess possible variation in LTL across the whole normal variation in fetal growth. The Helsinki Study of Very Low Birth Weight Adults (HeSVA) is a case–control cohort of young adults born preterm at a very low birthweight (VLBW; <1500 g), constituting a model of severe early-life adversity, and a comparison group born at term. The third cohort was twins from the FinnTwin16 study, which enables comparisons within and between young adult twin pairs. Twins are often born prematurely and are, on average, 1 kg lower in birthweight than singletons.34

Materials and methods

Study cohorts

Helsinki Birth Cohort Study

Characteristics of the 1894 subjects from the HBCS are shown in Table 1.They came from a birth cohort of 8760 men and women, who were born as singletons at Helsinki University Central Hospital during 1934–44, attended child welfare clinics in Helsinki and were resident in Finland in 1971. Their birth records include weight, length and head circumference at birth; dates of birth and mother’s last menstrual period,26 based on which length of gestation was calculated.35 Ponderal index was calculated as kg/m3. Because no newborn growth reference exists for this period in Finland, we adjusted birth measurements for length of gestation by including both simultaneously in a regression model. As previously,35,36 we determined socio-economic status by father’s occupation, categorized in three groups (Table 1) according to the classification by Statistics Finland. For 928 subjects with adequate data, maternal hypertensive disorders were defined as described.37 The present study is based on a clinical visit between 2001 and 2004.36 Briefly, we invited 2902 subjects living in greater Helsinki, selected by random-number tables, to participate; 2003 of them agreed to do so. The examination included measurements of height and weight and assessment of smoking by questionnaire. DNA for measurement of LTL was available for 1960 subjects, of whom 1894 (65.3% of those invited) had adequate data for length of gestation and were included in the study.

Table 1

Characteristics of the study cohorts

  HeSVA
 
 
 HBCS (n = 1894) VLBW (n = 164) Term (n = 170) FinnTwin16 (n = 248) 
Age at study, years (SD) 61.5 (2.9) 22.4 (2.1) 22.5 (2.2) 27.5 (2.0) 
Male, n (%) 885 (46.7) 69 (42.1) 68 (40.0) 134 (54.0) 
Ever smoker, n (%) 1125 (59.4) 74 (45.1) 108 (63.5) N/A 
Current daily smoker, n (%) 361 (19.1) 37 (22.6) 54 (31.8) 50 (20.5) 
BMI (kg/m227.6 (4.7) 22.2 (3.8) 23.0 (3.5) 25.3 (4.9) 
Length of gestation, weeks (SD) 39.8 (1.9) 29.2 (1.9) 40.2 (1.1) 36.8 (4.0) 
Measurements at birth     
    Weight, g (SD) 3404 (484) 1120 (222) 3598 (472) 2673 (503) 
    Length, cm (SD) 50.3 (2.0) 36.9 (2.4) 50.5 (1.8) 47.0 (2.5) 
    Head circumference, cm (SD) 35.1 (1.5) 26.2 (1.9) 35.1 (1.3)  
    Ponderal index, kg/m3 (SD) 26.7 (2.3) 22.0 (2.4) 27.9 (2.6) 25.5 (2.5) 
    Weight, SD score (SD) N/A −1.3 (1.5) 0.0 (1.0) N/A 
    Length, SD score (SD) N/A −1.3 (1.7) 0.0 (0.9) N/A 
    Head circumference, SD score (SD) N/A −1.1 (1.3) −0.1 (0.9) N/A 
Preterm, n (%) 112 (5.9) 164 (100) 102 (40.5) 
SGA, n (%) N/A 55 (33.5) N/A 
Twin, n (%) 24 (14.6) 248 (100) 
Triplet, n (%) 4 (2.4) 
Bronchopulmonary dysplasia, n (%) N/A 30 (18.3) N/A 
Neurodevelopmental disability, n (%) N/A 19 (11.6) 2 (1.2) N/A 
Maternal variables     
    Height, cm (SD) 159.6 (5.8) 164.3 (9.4) 165.2 (10.4) 164.0 (5.5) 
    BMI in late pregnancy, kg/m3 (SD) 26.5 (2.9) 25.4 (3.7) 27.1 (3.5) 28.2 (3.7) 
    Age at delivery, year (SD) 28.7 (5.5) 29.8 (4.8) 29.7 (5.0) N/A 
    Multiparous, n (%) 978 (51.6)   N/A 
    Smoking during pregnancy, n (%) N/A 31 (18.9) 26 (15.3) N/A 
Maternal hypertensive disorders in pregnancy, n (%) 
    Normotensive 581 (62.6) 123 (75.0) 133 (78.2) N/A 
    Gestational hypertension 278 (30.0) 7 (4.3) 24 (14.1) N/A 
    Pre-eclampsia 38 (4.1) 34 (20.7) 13 (7.6) N/A 
    Chronic hypertension 31 (3.3) N/A N/A N/A 
Father’s age at birth N/A 31.6 (5.5) 31.9 (5.9) N/A 
Father’s occupational status, n (%)     
    Upper middle 329 (17.4)    
    Lower middle 432 (22.8)    
    Manual worker 1119 (59.1)   16 (7.2) 
    Upper-level employee    36 (16.1) 
    Lower-level employee    53 (23.2) 
    Self-employed    82 (36.6) 
    Others    38 (17.0) 
Highest current education of either parent, n (%) 
    Elementary  17 (10.4) 11 (6.5)  
    High school  34 (20.7) 30 (17.6)  
    Intermediate  68 (41.5) 56 (32.9)  
    University  45 (27.4) 73 (42.9)  
  HeSVA
 
 
 HBCS (n = 1894) VLBW (n = 164) Term (n = 170) FinnTwin16 (n = 248) 
Age at study, years (SD) 61.5 (2.9) 22.4 (2.1) 22.5 (2.2) 27.5 (2.0) 
Male, n (%) 885 (46.7) 69 (42.1) 68 (40.0) 134 (54.0) 
Ever smoker, n (%) 1125 (59.4) 74 (45.1) 108 (63.5) N/A 
Current daily smoker, n (%) 361 (19.1) 37 (22.6) 54 (31.8) 50 (20.5) 
BMI (kg/m227.6 (4.7) 22.2 (3.8) 23.0 (3.5) 25.3 (4.9) 
Length of gestation, weeks (SD) 39.8 (1.9) 29.2 (1.9) 40.2 (1.1) 36.8 (4.0) 
Measurements at birth     
    Weight, g (SD) 3404 (484) 1120 (222) 3598 (472) 2673 (503) 
    Length, cm (SD) 50.3 (2.0) 36.9 (2.4) 50.5 (1.8) 47.0 (2.5) 
    Head circumference, cm (SD) 35.1 (1.5) 26.2 (1.9) 35.1 (1.3)  
    Ponderal index, kg/m3 (SD) 26.7 (2.3) 22.0 (2.4) 27.9 (2.6) 25.5 (2.5) 
    Weight, SD score (SD) N/A −1.3 (1.5) 0.0 (1.0) N/A 
    Length, SD score (SD) N/A −1.3 (1.7) 0.0 (0.9) N/A 
    Head circumference, SD score (SD) N/A −1.1 (1.3) −0.1 (0.9) N/A 
Preterm, n (%) 112 (5.9) 164 (100) 102 (40.5) 
SGA, n (%) N/A 55 (33.5) N/A 
Twin, n (%) 24 (14.6) 248 (100) 
Triplet, n (%) 4 (2.4) 
Bronchopulmonary dysplasia, n (%) N/A 30 (18.3) N/A 
Neurodevelopmental disability, n (%) N/A 19 (11.6) 2 (1.2) N/A 
Maternal variables     
    Height, cm (SD) 159.6 (5.8) 164.3 (9.4) 165.2 (10.4) 164.0 (5.5) 
    BMI in late pregnancy, kg/m3 (SD) 26.5 (2.9) 25.4 (3.7) 27.1 (3.5) 28.2 (3.7) 
    Age at delivery, year (SD) 28.7 (5.5) 29.8 (4.8) 29.7 (5.0) N/A 
    Multiparous, n (%) 978 (51.6)   N/A 
    Smoking during pregnancy, n (%) N/A 31 (18.9) 26 (15.3) N/A 
Maternal hypertensive disorders in pregnancy, n (%) 
    Normotensive 581 (62.6) 123 (75.0) 133 (78.2) N/A 
    Gestational hypertension 278 (30.0) 7 (4.3) 24 (14.1) N/A 
    Pre-eclampsia 38 (4.1) 34 (20.7) 13 (7.6) N/A 
    Chronic hypertension 31 (3.3) N/A N/A N/A 
Father’s age at birth N/A 31.6 (5.5) 31.9 (5.9) N/A 
Father’s occupational status, n (%)     
    Upper middle 329 (17.4)    
    Lower middle 432 (22.8)    
    Manual worker 1119 (59.1)   16 (7.2) 
    Upper-level employee    36 (16.1) 
    Lower-level employee    53 (23.2) 
    Self-employed    82 (36.6) 
    Others    38 (17.0) 
Highest current education of either parent, n (%) 
    Elementary  17 (10.4) 11 (6.5)  
    High school  34 (20.7) 30 (17.6)  
    Intermediate  68 (41.5) 56 (32.9)  
    University  45 (27.4) 73 (42.9)  

N/A, not available; SGA, small for gestational age (birthweight < −2 s.d.); SD, standard deviation.

Helsinki Study of Very Low Birth Weight Adults

Characteristics of the 164 VLBW and 170 term subjects from HeSVA are shown in Table 1. The cohort has been described in detail.3840 Briefly, the original VLBW cohort consisted of 335 consecutive VLBW infants who were born between 1978 and 1985 and were discharged alive from the neonatal intensive care unit of the Children’s Hospital at Helsinki University Central Hospital, the only tertiary neonatal care centre serving the province of Uusimaa, Finland. We selected a comparison group (group-matching) based on the charts of all consecutive births at their birth hospitals. For each VLBW survivor, we selected the next available singleton term (gestational age at birth ≥37 weeks) infant of the same sex who was not small for gestational age (birthweight >−2 s.d.).41 We invited 255 VLBW and 314 term-born individuals who were then residing in greater Helsinki to a clinical examination, which included blood sample for DNA extraction, measurement of height and weight and a questionnaire on medical history, smoking and socio-demographic data.38 A total of 166 VLBW and 172 control subjects participated; 164 (64.3% of those invited) and 170 (54.1%), respectively, had DNA available for measurement of LTL. Perinatal data came from birth records. Birth measurements relative to length of gestation were calculated in standard deviation (SD) scores by Finnish standards.41 Pre-eclampsia was defined with standard criteria42 and bronchopulmonary dysplasia according to the Northway criteria.43 As in previous publications,3840 childhood socio-economic status was indicated by the educational attainment of the higher-educated parent, obtained by questionnaire at the clinical examination.

FinnTwin16 Study

The 248 twins, aged 23.7–31.2 years, were selected from FinnTwin16, a population-based longitudinal study consisting of five consecutive birth cohorts of 5601 Finnish twins44 with four waves of questionnaire assessments (at 16, 17, 18.5 and 22–27 years). Because the samples came from a sub-study focusing on obesity,45 the sample of 124 twin pairs, 67 male and 57 female pairs, 41 monozygotic (MZ) and 83 same-sex dizygotic (DZ), was enriched by 13 MZ pairs and 49 DZ pairs who were discordant for body mass index (BMI; within-pair difference in BMI of ≥3 kg/m2). Clinical characteristics are shown in Table 1. The examination included measurements of height and weight and assessment of smoking by a questionnaire. Zygosity was confirmed by genotyping of 10 informative highly polymorphic genetic markers.46 Perinatal data came from birth records. Childhood socio-economic status was defined based on father’s reported occupation at twins’ age 16 years and classified into five groups (Table 1).

Study protocols were approved by ethics committees. All subjects signed an informed consent form. The principles outlined in the Declaration of Helsinki were followed.

Measurement of telomere length

Genomic DNA was extracted from EDTA-anti-coagulated whole peripheral blood by using QIAamp DNA Blood Maxi Kit (Qiagen®) in HBCS and HeSVA and by using an Autopure LS (Qiagen®) instrument in FinnTwin16. DNA concentration was measured by an absorbance-based method in HeSVA and HBCS and by a fluoresence-based (PicoGreen®) method in FinnTwin16.

Relative LTL was measured by a real-time quantitative polymerase chain reaction-based method47 as a ratio of telomere DNA signal intensity by haemoglobin beta single-copy gene signal intensity. Each sample was analysed as a separate batch. For HBCS and HeSVA, the analysis was performed as previously described,48 except that a synthetic oligomer (Sigma) dilution series,49 hgb-120-mer and tel14x (0.0002, 0.002, 0.02, 0.2, 1.0, 3.0 and 6.0 pg), was included on every plate to create reaction-specific standard curves. Plasmid DNA (pcDNA3.1) was added to each standard to maintain a constant 10 ng of total DNA concentration per reaction. Standard curves were used to perform absolute quantification of each individual sample reaction with 10 ng of template DNA. For FinnTwin16, LTL was also determined as previously described,24 except that 10 ng of template DNA was used. Each sample plate contained two to four control samples, which were used to calculate the inter-assay coefficients of variation. It was 21% for the HBCS (four control samples on each of the 22 plates), 28% for the HeSVA (two controls on each of the four plates) and 13% for FinnTwin16 (four controls on each of the four plates).

Data analysis

LTL (ratio of telomere DNA amount by haemoglobin beta single-copy gene amount) values were log-transformed to attain normality. For each of the three studies separately, the log-transformed value was adjusted for the dummy-coded plate number and stock DNA concentration by linear regression and expressed in SD units. This served as the dependent variable in the multiple linear regression assessing the associations of LTL with each predictor. All of these regression models included age at sampling and sex. We also adjusted for maternal age when the subject was born, as well as for socio-economic status (paternal occupational status in HBCS and FinnTwin16, parental educational attainment in HeSVA). In the HBCS, we further adjusted for current daily smoking, which was related to telomere length in this cohort.

In twin analyses, the multiple linear regressions that we used to assess the associations of LTL with each predictor in individual twins were also adjusted for clustered sampling of twin pairs. Therefore, the lack of statistical independence of co-twins was considered by computing robust variance estimators50 to yield correct P-values. Similar multiple linear regressions were then performed using within-pair differences of standardized LTL and within-pair differences of each of the birth measures (predictors). To compare TwinA with TwinB for their birth measures and LTL, t-tests were used. Wald tests for independent samples (t-tests adapted for clustered twin data) were used to compare LTL in the socio-economic groups.

PASW Statistics 18.0 was used for HBCS and HeSVA and Stata 11.0 for the twin analyses.

Results

Helsinki Birth Cohort Study

There was an inverse association between LTL and age at sampling (Table 2). The 361 current daily smokers had 0.181 s.d. [95% confidence interval (CI) 0.067 to 0.294] shorter telomeres than the remaining study population; a history of ever smoking (P = 0.3) and current BMI (P = 0.2) were not related to LTL. As HBCS represents a community sample encompassing the whole range of perinatal conditions, our main focus was on continuous early-life variables as predictors of LTL in adult life. As early-life predictors, we first assessed body size at birth or maternal or perinatal variables. There was no association between weight, length, ponderal index, head circumference, length of gestation or maternal height or BMI in late pregnancy and LTL at adult age. Subjects whose mothers were older when giving birth had longer telomeres than subjects with younger mothers. Birth order did not explain this. Moreover, childhood socio-economic status was associated with LTL. As compared with subjects whose fathers were manual workers, subjects whose fathers had lower middle class occupations had 0.024 s.d. (95% CI − 0.085 to 0.133) and those whose fathers had upper middle class occupations had 0.156 s.d. (95% CI 0.035 to 0.277) longer telomeres (P for trend = 0.02).

Table 2

Change in telomere length (or mean difference) associated with one unit change in each independent birth or maternal continuous (or dichotomous) variable in the HBCS

 Change (z score unit) or mean difference in telomere length (95% CI) 
 Model 1 Model 2 
Current age, years −0.038 (−0.054 to −0.023)* −0.034 (−0.050 to −0.019)* 
Birth measurements   
    Weight, kg −0.047 (−0.150 to 0.056) −0.060 (−0.163 to 0.042) 
    Length, cm −0.012 (−0.038 to 0.013) −0.014 (−0.040 to 0.012) 
    Ponderal index, kg/m3 0.001 (−0.019 to 0.021) −0.001 (−0.021 to 0.019) 
    Head circumference, cm 0.008 (−0.025 to 0.041) 0.000 (−0.033 to 0.032) 
Length of gestation, weeks −0.016 (−0.040 to 0.008) −0.014 (−0.038 to 0.010) 
Preterm (<37 weeks) 0.161 (−0.028 to 0.350) 0.131 (−0.058 to 0.320) 
Maternal data   
    Height, cm −0.002 (−0.010 to 0.006) −0.001 (−0.009 to 0.007) 
    BMI in late pregnancy, kg/m2 −0.014 (−0.030 to 0.003) −0.023 (−0.040 to −0.007)* 
    Age at delivery, years 0.014 (0.005 to 0.022)* 0.014 (0.006 to 0.022)* 
    Multiparous 0.035 (−0.055 to 0.124) −0.022 (−0.118 to 0.074) 
 Change (z score unit) or mean difference in telomere length (95% CI) 
 Model 1 Model 2 
Current age, years −0.038 (−0.054 to −0.023)* −0.034 (−0.050 to −0.019)* 
Birth measurements   
    Weight, kg −0.047 (−0.150 to 0.056) −0.060 (−0.163 to 0.042) 
    Length, cm −0.012 (−0.038 to 0.013) −0.014 (−0.040 to 0.012) 
    Ponderal index, kg/m3 0.001 (−0.019 to 0.021) −0.001 (−0.021 to 0.019) 
    Head circumference, cm 0.008 (−0.025 to 0.041) 0.000 (−0.033 to 0.032) 
Length of gestation, weeks −0.016 (−0.040 to 0.008) −0.014 (−0.038 to 0.010) 
Preterm (<37 weeks) 0.161 (−0.028 to 0.350) 0.131 (−0.058 to 0.320) 
Maternal data   
    Height, cm −0.002 (−0.010 to 0.006) −0.001 (−0.009 to 0.007) 
    BMI in late pregnancy, kg/m2 −0.014 (−0.030 to 0.003) −0.023 (−0.040 to −0.007)* 
    Age at delivery, years 0.014 (0.005 to 0.022)* 0.014 (0.006 to 0.022)* 
    Multiparous 0.035 (−0.055 to 0.124) −0.022 (−0.118 to 0.074) 

CI, confidence interval.

Model 1: adjusted for sex, age and length of gestation.

Model 2: adjusted for model 1 variables and also for family socio-economic status (based on father’s occupation) and maternal age at delivery.

*P < 0.05.

We repeated the analyses after adjustment for socio-economic status and maternal age at birth. The results were similar, except an inverse association between maternal BMI and LTL became stronger (Table 2). When further adjusted for current daily smoking status, the results remained similar.

We assessed possible non-linear associations with body size at birth by including a quadratic term in regression equation. No such association was found.

Of the subjects, 928 (49.0%) had data on maternal hypertensive disorders in pregnancy (Table 1). Maternal hypertensive disorders were not related to LTL (data not shown).

Helsinki Study of Very Low Birth Weight Adults

As HeSVA is a prospective case–control cohort, our main focus was on comparisons between VLBW and term-born groups and between subgroups of VLBW subjects (Table 3). LTL was similar between young adults born at VLBW and controls. Comparisons within the VLBW group revealed no differences in LTL according to being born small for gestational age, maternal pre-eclampsia or gestational hypertension, maternal smoking during pregnancy, multiple pregnancy, bronchopulmonary dysplasia or neurodevelopmental disability (P-values > 0.3).

Table 3

Mean telomere lengths and mean differences in telomere lengths between young adults born at VLBW and term-born controls and between subgroups of those born at VLBW

 Mean difference in telomere length (95% CI)
 
Mean telomere length in z score units Model 1 Model 2 
VLBW (n = 164) Term (n = 170)   
0.010 −0.009 0.022 (−0.192 to 0.236) 0.013 (−0.234 to 0.230) 
VLBW SGA (n = 55) VLBW AGA (n = 109)   
0.013 0.008 0.003 (−0.323 to 0.328) 0.036 (−0.290 to 0.363) 
VLBW pre-eclampsia (n = 34) VLBW no pre-eclampsia (n = 130)   
0.113 −0.017 0.127 (−0.251 to 0.504) 0.164 (−0.214 to 0.542) 
 Mean difference in telomere length (95% CI)
 
Mean telomere length in z score units Model 1 Model 2 
VLBW (n = 164) Term (n = 170)   
0.010 −0.009 0.022 (−0.192 to 0.236) 0.013 (−0.234 to 0.230) 
VLBW SGA (n = 55) VLBW AGA (n = 109)   
0.013 0.008 0.003 (−0.323 to 0.328) 0.036 (−0.290 to 0.363) 
VLBW pre-eclampsia (n = 34) VLBW no pre-eclampsia (n = 130)   
0.113 −0.017 0.127 (−0.251 to 0.504) 0.164 (−0.214 to 0.542) 

AGA, appropriate for gestational age (birthweight ≥ −2 s.d.).

Model 1: adjusted for sex and age.

Model 2: adjusted for model 1 variables and family socio-economic status (based on educational attainment of the highest educated parent).

We also assessed whether length of gestation or birthweight, length at birth or head circumference relative to length of gestation, ponderal index at birth or maternal height or BMI at late pregnancy predicted LTL. There was no association, neither separately among the VLBW and term groups nor among the whole study population (P-values ≥ 0.05).

Older age at sampling was associated with shorter telomeres, although this association was evident among men only (1 year higher age corresponded to −0.084 s.d. shorter telomeres; 95% CI 0.010 to 0.164). LTL was not associated with ever smoking (P = 0.7), current smoking (P = 0.08), BMI (P = 0.6), childhood socio-economic status (P = 0.6) or with maternal (P = 0.9) or paternal (P = 0.9) age at birth.

FinnTwin16 Study

In the twin sample, there was no correlation between LTL and the fairly narrow age range in young adulthood. LTL was also unrelated to current daily smoking or to BMI in within-pair and between-pair models (P-values > 0.1). There was no association between LTL and weight, length or ponderal index at birth (Table 4).

Table 4

Change in telomere length associated with one unit change in each independent birth variable in individual twins and in twin pairs from FinnTwin16 study

 Change in z score units in telomere length (95% CI)
 
 Individual twins (n = 248)
 
Twin pairsa(n = 124)
 
 Model 1 Model 2 Model 1b Model 2b 
Length of gestation (weeks) −0.0030 −0.0031 – – 
 (−0.016 to 0.0098) (−0.017 to 0.011)   
Birth measurements     
    Weight, kg −1.826 −1.487 −1.417 −0.950 
 (−4.018 to 0.366) (−3.641 to 0.665) (−4.922 to 2.087) (−3.718 to 1.819) 
    Length, cm 0.339 0.275 0.182 0.118 
 (−0.010 to 0.688) (−0.070 to 0.619) (−0.400 to 0.763) (−0.338 to 0.574) 
    Ponderal index, kg/m3 0.184 0.157 0.065 0.076 
 (−0.042 to 0.411) (−0.067 to 0.380) (−0.270 to 0.400) (−0.202 to 0.353) 
 Change in z score units in telomere length (95% CI)
 
 Individual twins (n = 248)
 
Twin pairsa(n = 124)
 
 Model 1 Model 2 Model 1b Model 2b 
Length of gestation (weeks) −0.0030 −0.0031 – – 
 (−0.016 to 0.0098) (−0.017 to 0.011)   
Birth measurements     
    Weight, kg −1.826 −1.487 −1.417 −0.950 
 (−4.018 to 0.366) (−3.641 to 0.665) (−4.922 to 2.087) (−3.718 to 1.819) 
    Length, cm 0.339 0.275 0.182 0.118 
 (−0.010 to 0.688) (−0.070 to 0.619) (−0.400 to 0.763) (−0.338 to 0.574) 
    Ponderal index, kg/m3 0.184 0.157 0.065 0.076 
 (−0.042 to 0.411) (−0.067 to 0.380) (−0.270 to 0.400) (−0.202 to 0.353) 

CI, confidence interval.

Model 1: adjusted for sex and age.

Model 2: adjusted for model 1 variables and family socio-economic status (based on father’s occupation).

aWithin-pair differences of each variable are used.

bWithin twin pairs also adjusted for zygosity (separate models for monozygotic and dizygotic twins did not differ and were therefore pooled).

Co-twins were then ordered based on their birthweights. Mean (SD) birthweights of TwinA were 2491 (487) g and 2548 (472) g and those of TwinB were 2774 (521) g and 2838 (484) g in MZ twins and in DZ twins, respectively. LTL did not differ between the lower vs higher birthweight co-twins in either of the zygosity groups. Further, within-pair differences in LTL and those of birth measures were not associated in the multivariate models (Table 4).

In the twins, childhood socio-economic status was associated with LTL. LTL was shortest in twins whose fathers were lower-level employees (s.d. −0.391, 95% CI −0.791 to 0.0099) and longest (s.d. 0.187, 95% CI −0.068 to 0.442) in those whose fathers were self-employed. Mean LTL between these two groups was different (P = 0.012). In the multivariate models, neither childhood socio-economic status nor birth measures were related to LTL.

Discussion

We used three cohorts of different and complementary design to test our hypothesis that adverse conditions during fetal period are associated with shorter LTL in adult life. We found no association in any of these cohorts between LTL in adulthood and with size at birth, length of gestation or preterm birth at VLBW. Of the three cohorts, the HBCS allows assessment of continuous effects across the whole normal range of births, whereas the HeSVA focuses on extreme conditions associated with VLBW, and the twin study enables comparisons between twin pairs. Therefore, our findings represent a strong argument that LTL is unlikely to serve as a marker of a mechanism linking conditions underlying body size at birth with individual differences in ageing in the general population.

We are aware of only two studies assessing the association between body size at birth and LTL in later life. A study of 132 subjects aged 5 years in Matlab, Bangladesh, showed that those born at low birthweight had shorter telomeres than those born at normal birthweight.32 This finding was interpreted as a sign of rapid T-cell turnover in a tropical rural setting with high infectious exposure and may not be comparable with our findings in adults. Another study included 94 young adults in Trier, Germany, half of whom had shorter telomeres, as they as fetuses had been exposed to severe maternal stress. In that study, severe maternal stress and lower birthweight adjusted for length of gestation were both independently associated with shorter leucocyte telomeres.33 Other longitudinal studies have investigated exposures such as childhood maltreatment21 and institutional care,22 chronic stress among caregiver mothers of chronically ill children23 and the number of adverse life events.24 Although our study did not focus on psychosocial stress as an exposure, in two of our three cohorts, we found shorter telomeres in subjects who grew up in families of low socio-economic status, a rough proxy of early-life adversities. It is possible that LTL in adult life could be more sensitive to long-term effects of psychosocial stress during early life than to other early-life adversities.

The strengths of the study include three well-characterized cohorts with complementary designs, a large number of subjects to detect any meaningful associations, should such associations exist, and a robust method for measurement of LTL. We have previously discussed the limitations of each of the cohorts36,3840,44 and published detailed non-participant analyses of the HBCS and HeSVA.38,51 Because our results are based on internal comparisons within each cohort, participation bias would be expected only if the association between early-life exposures and LTL was different among participants and non-participants. This is unlikely but cannot be excluded. DNA samples available for the study were obtained from aged individuals (HBCS) or young adults (HeSVA and FinnTwin16). Therefore, we were not able to measure LTL directly after birth or to assess telomere dynamics since birth. Other factors affecting LTL between birth and DNA sampling might have therefore affected the results. Moreover, birthweight is a crude summary measure reflecting a variety of intrauterine exposures and thus unlikely to show strong associations with any adult phenotype. Nevertheless, the lack of association with LTL should be interpreted against the fact that among virtually the same HBCS participants, low birthweight does predict a number of adult diseases and their risk factors, including impaired glucose tolerance,28 type 2 diabetes,28 hypertension,52 the use of lipid-lowering medication51 and depressive symptoms.53 Associations with many of these characteristics have also been seen among the HeSVA38 and FinnTwin16.45,54,55 Our findings are based on the Finnish population and do not necessarily generalize to other populations or specific exposures such as severe under-nutrition, high infectious exposure or severe psychosocial stress. The measurement of LTL was performed and standardized separately for each study cohort, precluding the use of pooled data. However, the expected associations, such as that between age and shorter telomeres, could be observed within the narrow age range of the HBCS and HeSVA.

In conclusion, in three cohorts of adults of different ages, we found no association between LTL and body size at birth or underlying factors, including severely preterm birth. Although LTL in adult life may in part reflect long-term effects of specific exposures such as severe psychosocial stress, our results imply that it is not likely to serve as a marker for a mechanism linking fetal and neonatal conditions reflected in body size at birth with individual differences in ageing-related processes in the general population.

Funding

This work was supported by grants from the Finnish Foundation for Pediatric Research, Finska Läkaresällskapet, the Finnish Special Governmental Subsidy for Health Sciences, Academy of Finland (grants 127437, 129306, 130326 and 134791 to E.K.; 132700 to K.W.; 140278, 129911, 135132 and 129457 to K.R.; 135252 to S.A.; 213506, 129680, 100499 and 141054 to J.K.; 129369, 129907, 135072, 129255 and 126775 to J.G.E.; 124940 and 140747 to I.H.), the Emil Aaltonen Foundation, the Finnish Medical Society Duodecim, the Jalmari and Rauha Ahokas Foundation, the Juho Vainio Foundation, the Novo Nordisk Foundation, the Finnish Foundation for Cardiovascular Research, the Biomedicum Foundation, the Päivikki and Sakari Sohlberg Foundation, the Pediatric Graduate School, the Clinical Graduate School in Paediatrics and Obstetrics/Gynaecology, University of Helsinki, the Research Foundation for the Orion Corporation, the Signe and Ane Gyllenberg Foundation, the Sigrid Jusélius Foundation and the Yrjö Jahnsson Foundation.

KEY MESSAGES

  • Shorter LTL is predicted by cumulative adverse conditions throughout life course and may serve as a marker of biological ageing. Previous studies have suggested an association between shorter telomeres and low birthweight.

  • This study of three large cohorts showed that LTL is not associated with weight, length or gestational age at birth.

  • LTL is unlikely to be a useful marker of a mechanism linking body size at birth with individual differences in ageing.

Acknowledgements

The authors thank Kaisa Manninen for technical help in telomere length measurement.

Conflict of interest: None declared.

References

1
Chan
SR
Blackburn
EH
Telomeres and telomerase
Philos Trans R Soc Lond B Biol Sci
 , 
2004
, vol. 
359
 (pg. 
109
-
21
)
2
Eisenberg
DT
An evolutionary review of human telomere biology: the thrifty telomere hypothesis and notes on potential adaptive paternal effects
Am J Hum Biol
 , 
2011
, vol. 
23
 (pg. 
149
-
67
)
3
Demissie
S
Levy
D
Benjamin
EJ
, et al.  . 
Insulin resistance, oxidative stress, hypertension, and leukocyte telomere length in men from the Framingham Heart Study
Aging Cell
 , 
2006
, vol. 
5
 (pg. 
325
-
30
)
4
Aviv
A
Telomeres and human aging: facts and fibs
Sci Aging Knowledge Environ
 , 
2004
, vol. 
2004
 pg. 
pe43
 
5
Crimmins
EM
Finch
CE
Infection, inflammation, height, and longevity
Proc Natl Acad Sci U S A
 , 
2006
, vol. 
103
 (pg. 
498
-
503
)
6
Finkel
T
Holbrook
NJ
Oxidants, oxidative stress and the biology of ageing
Nature
 , 
2000
, vol. 
408
 (pg. 
239
-
47
)
7
Gardner
JP
Li
S
Srinivasan
SR
, et al.  . 
Rise in insulin resistance is associated with escalated telomere attrition
Circulation
 , 
2005
, vol. 
111
 (pg. 
2171
-
77
)
8
Jeanclos
E
Krolewski
A
Skurnick
J
, et al.  . 
Shortened telomere length in white blood cells of patients with IDDM
Diabetes
 , 
1998
, vol. 
47
 (pg. 
482
-
86
)
9
Adaikalakoteswari
A
Balasubramanyam
M
Ravikumar
R
Deepa
R
Mohan
V
Association of telomere shortening with impaired glucose tolerance and diabetic macroangiopathy
Atherosclerosis
 , 
2007
, vol. 
195
 (pg. 
83
-
89
)
10
Sampson
MJ
Winterbone
MS
Hughes
JC
Dozio
N
Hughes
DA
Monocyte telomere shortening and oxidative DNA damage in type 2 diabetes
Diabetes Care
 , 
2006
, vol. 
29
 (pg. 
283
-
89
)
11
von Zglinicki
T
Serra
V
Lorenz
M
, et al.  . 
Short telomeres in patients with vascular dementia: an indicator of low antioxidative capacity and a possible risk factor?
Lab Invest
 , 
2000
, vol. 
80
 (pg. 
1739
-
47
)
12
Brouilette
SW
Moore
JS
McMahon
AD
, et al.  . 
Telomere length, risk of coronary heart disease, and statin treatment in the West of Scotland Primary Prevention Study: a nested case–control study
Lancet
 , 
2007
, vol. 
369
 (pg. 
107
-
14
)
13
Fitzpatrick
AL
Kronmal
RA
Gardner
JP
, et al.  . 
Leukocyte telomere length and cardiovascular disease in the cardiovascular health study
Am J Epidemiol
 , 
2007
, vol. 
165
 (pg. 
14
-
21
)
14
Yang
Z
Huang
X
Jiang
H
, et al.  . 
Short telomeres and prognosis of hypertension in a chinese population
Hypertension
 , 
2009
, vol. 
53
 (pg. 
639
-
45
)
15
Fyhrquist
F
Silventoinen
K
Saijonmaa
O
, et al.  . 
Telomere length and cardiovascular risk in hypertensive patients with left ventricular hypertrophy: the LIFE study
J Hum Hypertens
 , 
2011
, vol. 
25
 (pg. 
711
-
18
)
16
Bakaysa
SL
Mucci
LA
Slagboom
PE
, et al.  . 
Telomere length predicts survival independent of genetic influences
Aging Cell
 , 
2007
, vol. 
6
 (pg. 
769
-
74
)
17
Cawthon
RM
Smith
KR
O’Brien
E
Sivatchenko
A
Kerber
RA
Association between telomere length in blood and mortality in people aged 60 years or older
Lancet
 , 
2003
, vol. 
361
 (pg. 
393
-
95
)
18
Martin-Ruiz
CM
Gussekloo
J
van Heemst
D
von Zglinicki
T
Westendorp
RG
Telomere length in white blood cells is not associated with morbidity or mortality in the oldest old: a population-based study
Aging Cell
 , 
2005
, vol. 
4
 (pg. 
287
-
90
)
19
Njajou
OT
Hsueh
WC
Blackburn
EH
, et al.  . 
Association between telomere length, specific causes of death, and years of healthy life in health, aging, and body composition, a population-based cohort study
J Gerontol A Biol Sci Med Sci
 , 
2009
, vol. 
64
 (pg. 
860
-
64
)
20
Strandberg
TE
Saijonmaa
O
Tilvis
RS
, et al.  . 
Association of telomere length in older men with mortality and midlife body mass index and smoking
J Gerontol A Biol Sci Med Sci
 , 
2011
, vol. 
66
 (pg. 
815
-
20
)
21
Tyrka
AR
Price
LH
Kao
HT
Porton
B
Marsella
SA
Carpenter
LL
Childhood maltreatment and telomere shortening: preliminary support for an effect of early stress on cellular aging
Biol Psychiatry
 , 
2010
, vol. 
67
 (pg. 
531
-
34
)
22
Drury
SS
Theall
K
Gleason
MM
, et al.  . 
Telomere length and early severe social deprivation: linking early adversity and cellular aging
Mol Psychiatry
 , 
2012
, vol. 
17
 (pg. 
719
-
27
)
23
Epel
ES
Blackburn
EH
Lin
J
, et al.  . 
Accelerated telomere shortening in response to life stress
Proc Natl Acad Sci U S A
 , 
2004
, vol. 
101
 (pg. 
17312
-
15
)
24
Kananen
L
Surakka
I
Pirkola
S
, et al.  . 
Childhood adversities are associated with shorter telomere length at adult age both in individuals with an anxiety disorder and controls
PLoS One
 , 
2010
, vol. 
5
 pg. 
e10826
 
25
Milner
RDG
Gluckman
PD
Gluckman
PD
Heymann
MA
Regulation of intrauterine growth
Pediatrics & Perinatology: The Scientific Basis
 , 
1996
London
Arnold
(pg. 
284
-
89
)
26
Barker
DJ
Osmond
C
Forsén
TJ
Kajantie
E
Eriksson
JG
Trajectories of growth among children who have coronary events as adults
N Engl J Med
 , 
2005
, vol. 
353
 (pg. 
1802
-
09
)
27
Huxley
R
Owen
CG
Whincup
PH
, et al.  . 
Is birth weight a risk factor for ischemic heart disease in later life?
Am J Clin Nutr
 , 
2007
, vol. 
85
 (pg. 
1244
-
50
)
28
Eriksson
JG
Osmond
C
Kajantie
E
Forsén
TJ
Barker
DJ
Patterns of growth among children who later develop type 2 diabetes or its risk factors
Diabetologia
 , 
2006
, vol. 
49
 (pg. 
2853
-
58
)
29
Whincup
PH
Kaye
SJ
Owen
CG
, et al.  . 
Birth weight and risk of type 2 diabetes: a systematic review
JAMA
 , 
2008
, vol. 
300
 (pg. 
2886
-
97
)
30
Kajantie
E
Early life events. Effect on aging
Hormones
 , 
2008
, vol. 
7
 (pg. 
101
-
13
)
31
Räikkönen
K
Forsén
T
Henriksson
M
, et al.  . 
Growth trajectories and intellectual abilities in young adulthood: the Helsinki Birth Cohort study
Am J Epidemiol
 , 
2009
, vol. 
170
 (pg. 
447
-
55
)
32
Raqib
R
Alam
DS
Sarker
P
, et al.  . 
Low birth weight is associated with altered immune function in rural Bangladeshi children: a birth cohort study
Am J Clin Nutr
 , 
2007
, vol. 
85
 (pg. 
845
-
52
)
33
Entringer
S
Epel
ES
Kumsta
R
, et al.  . 
Stress exposure in intrauterine life is associated with shorter telomere length in young adulthood
Proc Natl Acad Sci U S A
 , 
2011
, vol. 
108
 (pg. 
E513
-
18
)
34
Pietiläinen
KH
Kaprio
J
Räsänen
M
Rissanen
A
Rose
RJ
Genetic and environmental influences on the tracking of body size from birth to early adulthood
Obes Res
 , 
2002
, vol. 
10
 (pg. 
875
-
84
)
35
Kajantie
E
Osmond
C
Barker
DJP
Eriksson
JG
Preterm birth—a risk factor for type 2 diabetes? The Helsinki Birth Cohort Study
Diabetes Care
 , 
2010
, vol. 
33
 (pg. 
2623
-
25
)
36
Ylihärsilä
H
Kajantie
E
Osmond
C
Forsen
T
Barker
DJ
Eriksson
JG
Birth size, adult body composition and muscle strength in later life
Int J Obes (Lond)
 , 
2007
, vol. 
31
 (pg. 
1392
-
99
)
37
Kajantie
E
Eriksson
JG
Osmond
C
Thornburg
KL
Barker
DJ
Preeclampsia is associated with increased risk of stroke in the adult offspring: the Helsinki Birth Cohort Study
Stroke
 , 
2009
, vol. 
40
 (pg. 
1176
-
80
)
38
Hovi
P
Andersson
S
Eriksson
JG
, et al.  . 
Glucose regulation in young adults with very low birth weight
N Engl J Med
 , 
2007
, vol. 
356
 (pg. 
2053
-
63
)
39
Kajantie
E
Hovi
P
Räikkönen
K
, et al.  . 
Young adults with very low birth weight: leaving the parental home and sexual relationships—Helsinki Study of Very Low Birth Weight Adults
Pediatrics
 , 
2008
, vol. 
122
 (pg. 
e62
-
72
)
40
Räikkönen
K
Pesonen
AK
Heinonen
K
, et al.  . 
Depression in young adults with very low birth weight: the Helsinki Study of Very Low Birth Weight Adults
Arch Gen Psychiatry
 , 
2008
, vol. 
65
 (pg. 
290
-
96
)
41
Pihkala
J
Hakala
T
Voutilainen
P
Raivio
K
Uudet suomalaiset sikiön kasvukäyrät [Characteristics of recent fetal growth curves in Finland]
Duodecim
 , 
1989
, vol. 
105
 (pg. 
1540
-
46
)
42
NHBPEP
Report of the National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy
Am J Obstet Gynecol
 , 
2000
, vol. 
183
 (pg. 
S1
-
22
)
43
Northway
WH
Jr
Rosan
RC
Porter
DY
Pulmonary disease following respirator therapy of hyaline-membrane disease. Bronchopulmonary dysplasia
N Engl J Med
 , 
1967
, vol. 
276
 (pg. 
357
-
68
)
44
Kaprio
J
Twin studies in Finland 2006
Twin Res Hum Genet
 , 
2006
, vol. 
9
 (pg. 
772
-
77
)
45
Pietiläinen
KH
Söderlund
S
Rissanen
A
, et al.  . 
HDL subspecies in young adult twins: heritability and impact of overweight
Obesity (Silver Spring)
 , 
2009
, vol. 
17
 (pg. 
1208
-
14
)
46
Pietiläinen
KH
Rissanen
A
Laamanen
M
, et al.  . 
Growth patterns in young adult monozygotic twin pairs discordant and concordant for obesity
Twin Res
 , 
2004
, vol. 
7
 (pg. 
421
-
29
)
47
Cawthon
RM
Telomere measurement by quantitative PCR
Nucleic Acids Res
 , 
2002
, vol. 
30
 pg. 
e47
 
48
Eerola
J
Kananen
L
Manninen
K
Hellström
O
Tienari
PJ
Hovatta
I
No evidence for shorter leukocyte telomere length in Parkinson’s disease patients
J Gerontol A Biol Sci Med Sci
 , 
2010
, vol. 
65
 (pg. 
1181
-
84
)
49
O’Callaghan
N
Dhillon
V
Thomas
P
Fenech
M
A quantitative real-time PCR method for absolute telomere length
Biotechniques
 , 
2008
, vol. 
44
 (pg. 
807
-
09
)
50
Williams
RL
A note on robust variance estimation for cluster-correlated data
Biometrics
 , 
2000
, vol. 
56
 (pg. 
645
-
46
)
51
Kajantie
E
Barker
DJ
Osmond
C
Forsén
T
Eriksson
JG
Growth before 2 years of age and serum lipids 60 years later: the Helsinki Birth Cohort Study
Int J Epidemiol
 , 
2008
, vol. 
37
 (pg. 
280
-
89
)
52
Eriksson
JG
Forsén
TJ
Kajantie
E
Osmond
C
Barker
DJ
Childhood growth and hypertension in later life
Hypertension
 , 
2007
, vol. 
49
 (pg. 
1415
-
21
)
53
Räikkonen
K
Pesonen
AK
Kajantie
E
, et al.  . 
Length of gestation and depressive symptoms at age 60 years
Br J Psychiatry
 , 
2007
, vol. 
190
 (pg. 
469
-
74
)
54
Pietiläinen
KH
Rissanen
A
Kaprio
J
, et al.  . 
Acquired obesity is associated with increased liver fat, intra-abdominal fat, and insulin resistance in young adult monozygotic twins
Am J Physiol Endocrinol Metab
 , 
2005
, vol. 
288
 (pg. 
E768
-
74
)
55
Pietiläinen
KH
Róg
T
Seppänen-Laakso
T
, et al.  . 
Association of lipidome remodeling in the adipocyte membrane with acquired obesity in humans
PLoS Biol
 , 
2011
, vol. 
9
 pg. 
e1000623