Skip to main content
International Journal of Epidemiology logoLink to International Journal of Epidemiology
. 2015 Sep 24;44(5):1749–1754. doi: 10.1093/ije/dyv171

Reproducibility of Telomere Length Assessment - An International Collaborative Study

Carmen M Martin-Ruiz , Duncan Baird, Laureline Roger, Petra Boukamp, Damir Krunic, Richard Cawthon, Martin M Dokter, Pim van der Harst, Sofie Bekaert, Tim de Meyer, Goran Roos, Ulrika Svenson, Veryan Codd, Nilesh J Samani, Liane McGlynn, Paul G Shiels, Karen A Pooley, Alison M Dunning, Rachel Cooper, Andrew Wong, Andrew Kingston, Thomas von Zglinicki
PMCID: PMC6312091  PMID: 26403809

International Journal of Epidemiology 2014, doi: 10.1093/ije/dyu191

Key Messages

Rankings are similar if different laboratories measure telomere lengths in the same samples.

TLR values for Labs 3 and 4 in round 2 as shown in Tab. 2 were not calculated from the set of raw values shown in suppl. Tab. S2, and this error was propagated through the following analyses. In addition, in suppl. Tab. S3, Pearson correlations were shown instead of Spearman’s correlation coefficients. Results based on the set of raw data shown in suppl. Tab S2 are provided below.

We correct the following statements (corrections underlined):

Results: Absolute results from different laboratories differed widely and could thus not be compared directly, but most rankings of relative telomere lengths were highly correlated (correlation coefficients 0.25 −0.99).

Table 2.

TLR as measured in the participating labs and inter-lab CVs in round 1 (to) and round 2 (bottom)

Sample Round 1
Lab 1 Lab 2 Lab 3 Lab 4 Lab5 Lab6 Lab7 Lab8 Lab9 CV for All Labs CV for qPCR Labs CV for qPCR triplets (median) CV for South & STELA
South South STELA qPCR qPCR qPCR qPCR qPCR qPCR
A 1.19 1.07 1.35 1.13 1.06 1.23 1.44 0.91 1.10 13.78 15.60 14.75 11.67
B 1.15 1.34 1.28 0.65 1.18 1.14 1.21 1.34 1.16 17.89 21.44 19.23 7.68
C 1.91 1.61 1.85 1.51 1.72 1.53 2.35 1.55 1.78 15.17 18.40 13.66 8.92
D 1.08 1.26 1.07 0.59 0.66 0.83 1.13 0.83 0.63 27.45 25.75 18.78 9.38
E 0.63 0.87 0.44 0.22 0.36 0.79 0.13 0.31 57.53 61.59 52.56 22.86
F 0.63 0.79 0.79 0.39 0.19 0.28 0.46 0.39 0.14 53.46 40.54 42.92 12.80
G a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
H 0.64 0.68 0.75 0.17 0.31 0.33 0.31 0.13 57.93 36.87 41.46 7.79
I 0.91 1.11 0.94 1.30 1.52 1.10 1.80 1.39 1.79 25.44 18.65 16.93 10.86
J 0.90 0.95 0.94 0.88 0.86 0.83 1.15 0.89 0.89 10.21 12.83 9.74 2.68
Sample b Round 2
Lab 1 Lab 2 Lab 3 Lab 4 Lab5 Lab6 Lab7 Lab8 Lab9 Lab 10 Lab 10-2 CV for All Labs CV for qPCR Labs CV for qPCR triplets (median) CV for South & STELA
South South STELA qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR
B 1.39 1.37 1.51 0.84 1.13 1.06 1.43 1.20 0.99 0.85 0.98 20.62 18.48 16.69 5.54
C 1.52 1.54 1.54 1.73 1.54 1.50 1.64 1.73 1.93 1.55 11.06 12.32 8.93 6.31
C 1.56 1.53 1.76 1.78 1.55 1.59 1.61 1.53 1.65 1.63 0.99
K 0.99 1.05 1.12 0.68 0.58 0.74 1.04 0.84 0.50 0.73 0.74 24.89 22.15 19.87 6.24
L 0.99 1.08 0.83 0.42 0.60 0.65 0.60 0.65 0.45 0.73 30.74 22.21 19.99 6.53
G 0.94 0.95 0.94 1.02 0.97 1.03 0.94 1.00 1.01 1.11 1.33 7.93 8.64 4.64 3.25
G a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
H 0.57 0.69 0.75 0.38 0.19 0.31 0.31 0.11 0.12 0.19 62.27 39.89 37.06 13.19
H 0.56 0.71 0.77 0.38 0.21 0.27 0.28 0.29 0.11 0.13 0.21
I 0.87 0.99 0.99 1.28 1.48 1.25 1.32 1.37 1.61 1.44 1.38 18.17 8.48 7.26 7.67
Median 22.76 20.05 17.86 7.74

TLR, telomere length ratio; CVs, coefficients of variation.

a All TLR values were calculated as the ratio of the estimated telomere length for a particular sample, divided by the estimated telomere length for sample G.

b The second round of measurements was designed to enable inter-batch comparison and included 5 repeat samples from the first round (B, C, G, H, I), of which samples C, G and H were duplicated (for intra-batch comparison). CVs for qPCR labs were higher than those for Southern/STELA labs (p = 0.000, paired t-test).

Figure 1.

Figure 1.

Telomere length ratios (TLRs) by laboratory, round and sample. TLRs are normalized to sample G, first round. Symbols indicate laboratories and techniques. Green indicates SOUTH, blue indicates STELA and pink symbols indicate qPCR. ▪ Lab 1 South; ▴ Lab 2 South; ✗ Lab 3 STELA; ▴ Lab 4 qPCR; ♦ Lab 5 qPCR; * Lab 6 qPCR; ▪ Lab 7 qPCR; Δ Lab 8 qPCR; ♦ Lab 9 qPCR; • Lab 10 qPCR duplex; ○ Lab 10-2 qPCR monoplex.

Figure 2.

Figure 2.

Figure 3.

Figure 3.

(a–c).

Table 3.

Intra-batch CVs per laboratory

Sample Lab 1 Lab 2 Lab 3 Lab 4 Lab5 Lab6 Lab7 Lab8 Lab9 Lab 10 Lab 10-2
name South South STELA qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR
C 1.702 0.178 10.294 7.799 1.903 4.771 4.566 3.354 11.934 31.299
G 4.614 3.481 4.374 1.331 2.162 2.156 4.721 0.324 0.470 7.095 20.089
H 1.083 2.007 1.489 1.304 7.018 8.985 3.861 0.000 2.404 6.264

Table 4.

Inter-batch CVs per laboratory

Sample Lab 1 Lab 2 Lab 3 Lab 4 Lab 5 Lab 6 Lab 7 Lab 8 Lab 9
name Tech 1 Tech 1 Tech 2 Tech 3 Tech 3 Tech 3 Tech 3 Tech 3 Tech 3
B 13.388 1.499 11.627 17.989 3.046 5.215 11.522 7.431 11.314
C 15.305 3.368 3.772 6.497 3.564 1.652 28.906 1.709 3.973
G 2.270 1.719 2.154 0.669 1.073 1.086 2.322 0.162 0.235
H 8.813 2.980 1.259 11.650 8.925 7.144 0.850 13.671
I 3.877 7.991 3.755 0.897 2.175 8.620 22.052 1.093 7.395

Suppl. Table S3.

Spearman’s rank correlation coefficients between participating laboratories

Round 1 Lab 1 South Lab 2 South Lab 3 STELA Lab 4 qPCR Lab 5 qPCR Lab 6 qPCR Lab 7 qPCR Lab 8 qPCR Lab 9 qPCR
Lab 1 South 1.000
Lab 2 South 0.855 1.000
Lab 3 STELA 0.983 0.867 1.000
Lab 4 qPCR 0.650 0.600 0.524 1.000
Lab 5 qPCR 0.770 0.855 0.750 0.900 1.000
Lab 6 qPCR 0.879 0.818 0.867 0.867 0.939 1.000
Lab 7 qPCR 0.770 0.842 0.750 0.883 0.952 0.915 1.000
Lab 8 qPCR 0.770 0.806 0.700 0.867 0.952 0.867 0.867 1.000
Lab 9 qPCR 0.709 0.818 0.667 0.883 0.988 0.903 0.939 0.939 1.000
Round 2 Lab 1 South Lab 2 South Lab 3 STELA Lab 4 qPCR Lab 5 qPCR Lab 6 qPCR Lab 7 qPCR Lab 8 qPCR Lab 9 qPCR Lab 10 qPCR Lab 10-2 qPCR
Lab 1 South 1.000
Lab 2 South 0.933 1.000
Lab 3 STELA 0.929 0.950 1.000
Lab 4 qPCR 0.700 0.685 0.483 1.000
Lab 5 qPCR 0.783 0.855 0.667 0.927 1.000
Lab 6 qPCR 0.690 0.733 0.476 0.933 0.967 1.000
Lab 7 qPCR 0.883 0.927 0.817 0.806 0.927 0.917 1.000
Lab 8 qPCR 0.800 0.842 0.650 0.915 0.988 0.967 0.939 1.000
Lab 9 qPCR 0.695 0.693 0.477 0.985 0.936 0.917 0.802 0.936 1.000
Lab 10 qPCR 0.683 0.733 0.500 0.976 0.952 0.933 0.830 0.939 0.985 1.000
Lab 10-2 qPCR 0.450 0.576 0.250 0.867 0.855 0.750 0.661 0.842 0.912 0.927 1.000

Suppl. Table S4.

z-scored results from all participating laboratories in round 1 (top) and 2 (bottom) and inter-laboratory variation in z scores (as standard deviation) between all laboratories and separated by technique

Lab 1 Lab 2 Lab 3 Lab 4 Lab 5 Lab 6 Lab 7 Lab 8 Lab 9 Lab 10 lab 10-2 SD SD SD South/
South South STELA qPCR qPCR qPCR qPCR qPCR qPCR qPCR qPCR all qPCR STELA
round 1
0.453 −0.012 0.751 0.480 0.306 0.703 0.663 −0.033 0.319 0.288 0.271 0.385
0.344 0.928 0.540 −0.666 0.527 0.492 0.239 0.876 0.417 0.463 0.524 0.296
2.414 1.894 2.280 1.395 1.538 1.419 2.370 1.330 1.484 0.453 0.389 0.270
0.165 0.670 −0.094 −0.795 −0.426 −0.242 0.083 −0.210 −0.491 0.426 0.298 0.390
−1.075 −0.727 −1.172 −1.245 −1.358 −0.561 −1.699 −1.030 0.357 0.379 0.246
−1.073 −1.004 −0.958 −1.280 −1.303 −1.541 −1.176 −1.159 −1.312 0.181 0.138 0.058
−0.061 −0.264 −0.320 0.177 0.201 0.165 −0.162 0.161 0.147 0.209 0.137 0.136
−1.035 −1.416 −1.092 −1.333 −1.474 −1.423 −1.327 −1.337 0.158 0.067 0.205
−0.295 0.129 −0.506 0.891 1.166 0.413 1.338 0.998 1.496 0.723 0.382 0.324
−0.338 −0.457 −0.519 −0.117 −0.054 −0.236 0.124 −0.066 −0.049 0.213 0.117 0.092
round 2
0.998 1.029 1.242 −0.217 0.435 0.300 0.643 0.591 0.124 −0.143 0.140 0.483 0.317 0.133
1.349 1.635 1.454 1.549 1.455 0.778 1.521 1.386 1.624 1.397 0.398 0.414 0.244
1.449 1.622 1.987 2.030 1.215 1.554 0.974 1.302 1.250 1.131 0.169
−0.093 −0.098 0.043 −0.592 −0.572 −0.455 −0.095 −0.177 −0.705 −0.336 −0.365 0.247 0.210 0.080
−0.313 −0.073 −0.231 −0.880 −0.794 −0.817 −0.704 −0.444 −0.803 −0.387 0.289 0.242 0.170
−0.233 −0.434 −0.503 0.222 0.146 0.239 −0.283 0.151 0.158 0.273 0.919 0.310 0.245 0.157
−0.061 −0.264 −0.320 0.177 0.201 0.165 −0.162 0.161 0.147 0.100 0.195
−1.228 −1.349 −1.077 −1.313 −1.295 −1.446 −1.317 −1.376 −1.336 −1.566 0.138 0.104 0.123
−1.252 −1.278 −1.028 −1.296 −1.258 −1.561 −1.517 −1.352 −1.376 −1.329 −1.527
−0.428 −0.292 −0.351 0.853 1.081 0.754 0.429 0.953 1.193 0.819 1.024 0.615 0.235 0.068
median 0.310 0.245 0.170

Table 5.

Test results

Analysis
original value in the paper corrected value
Spearman’s rank correlation coefficients (abstract and results p4 1 st para) Range: 0.63 −0.99 Range: 0.25 −0.99
Paired T-test CVs (SB+STELA) vs CVs qPCR (Results p.4 2 nd para) p = 0.001 p = 1.8x10 −7
Linear regression of LTRs South/STELA vs qPCR (p6 1 st para) Offset: −0.55 ± 0.32 Offset: −0.49 ± 0.32
Slope: 1.38 ± 0.30 Slope: 1.30 ± 0.30
Intra-batch CV values (Table 3, p6 2 nd para) Differences between labs Labs 1 to 10-1 ANOVA, p = 0.299 Labs 1 to 10.1 ANOVA, p = 0.299
Labs 1 to 10-2 Kruskal-Wallis, p = 0.089
Median intra-batch CVs per technique 1.86% (SB); 2.83% (STELA); 4.57% (qPCR) 1.86% (SB); 2.93% (STELA); 4.57% (qPCR)
Differences between techniques (Kruskal-Wallis) p = 0.161 p = 0.201
Differences between techniques with SOUTH and STELA combined (Mann-Whitney) p = 0.075 p = 0.082
Analysis original value in the paper corrected value
Inter-batch CV values (Table 4, p7) Differences between labs (Kruskal-Wallis) p = 0.195 p = 0.190
Median inter-batch CVs per technique 3.62% (SB); 4.78% (STELA); 4.65% (qPCR) 3.62% (SB); 3.76% (STELA); 3.97% (qPCR)
Differences between techniques (Kruskal-Wallis) p = 0.840 p = 0.842
Intra- & Inter-batch CV values combined (p7, 2 nd para) Differences between labs (Kruskal-Wallis) p = 0.060 p = 0.052
GLM (p8) Null hypothesis of equal variance between all groups F = 1.650; p = 0.096 F = 1.998; p = 0.036
Partial eta-squared coefficients: Partial eta-squared coefficients:
LAB .013 LAB .014
TECHN .000 TECHN .000
LAB * TECHN .000 LAB * TECHN .000

Articles from International Journal of Epidemiology are provided here courtesy of Oxford University Press

RESOURCES