Response to ‘Regression to the mean, apparent data errors, and biologically extraordinary results’

We thank Dr. Smith and colleagues for their careful review of published medical literature. We agree that scientific rigor is of greatest importance. However, we disagree with their review of ‘Changes in Telomere Length 3–5 years after Gastric Bypass Surgery’.1 We believe our findings are free of data errors, are more expected than extraordinary and are sufficient to support our conclusions.

Smith describes our results as ‘extraordinary’, implying they are unusual and unexpected. Given existing published research, we believe it is unreasonable to describe our study as ‘extraordinary’. We are not the first to report telomere lengthening after bariatric surgery. In 2016, Laimer et al.2 reported increases in telomere length (TL) after bariatric surgery. Laimer evaluated changes in 142 bariatric patients and TL increased at 10 years after bariatric surgery. Our study found similar results, but at 3–5 years post surgery. We argue that our results are probable and not extraordinary.

Smith suggests the original article includes a numerical error. However, no data error exists as the medians that they state are unequal were identical. We agree that the median values for TL of the intermediate group should agree with the median value for baseline TL of the overall group, but only for the baseline time point. In our study (Figure 1), these baseline values agree (both with median values of 5). However, since we categorized based on baseline values and not follow-up values, the median values of the intermediate and overall group for the follow-up TL are free to diverge without mathematical constraints.

Figure 1
Figure 1

Figure adapted from Dershem et al.,1 demonstrating that medians were equal and numerical error does not exist.

Smith references Steenstrup et al.3 when stating concerns for measurement error of TL. We agree with a recent article by Bateson and Nettle4, which disputed Steenstrup’s results and concluded that the current data do not support the notion of telomere lengthening being an artifact of measurement error. Regardless, our study was not measuring TL in healthy individuals who would experience typical telomere attrition. It has been suggested that ‘cell aging can be slowed or reversed in vivo over short periods’ under certain conditions.5 Some well-designed experiments also indicate that the high turnover rate and the changes in distribution of leukocytes can impact TL and produced comparable TL data.5, 6 We measured TL by following a well-established method7 and conducted the experiments in a controlled research environment. Samples for the study were run in triplicate and any sample that fell outside of an acceptable standard deviation was reevaluated to minimize any error. Therefore, quality-control procedures applied in our experimental design minimized measurement error and support the conclusion that telomere lengthening is a true biological phenomenon.

Finally, we agree with Smith that regression to the mean (RTM) is important and we acknowledged RTM and the lack of a control group as a weakness of our study. Due to the possibility of RTM, we agree that caution should be used in interpreting these conclusions. However, because similar published results show telomere lengthening after bariatric surgery2, we believe it would be irresponsible to totally disregard these conclusions as a byproduct of RTM.

In summary, although the criticism of Smith and colleagues is well intended, they do not impact our conclusions or the relevance of this research. Their criticism is based on a data error that we find does not exist, the misapplied use of the word ‘extraordinary’, and a study limitation that we acknowledged within the published manuscript. We believe that our findings are sufficient to support our conclusions and we look forward to additional contributions with larger sample sizes to further expand these preliminary findings in this important area of research.


  1. 1.

    , , , , , . Changes in telomere length 3-5 years after gastric bypass surgery. Int J Obes 2017; 41911: 1718–1720.

  2. 2.

    , , , , , et al. Telomere length increase after weight loss induced by bariatric surgery: results from a 10-year prospective study. Int J Obes 2016; 40: 773–778.

  3. 3.

    , , , , . The telomere lengthening conundrum-artifact or biology? Nucleic Acids Res 2013; 41: e131.

  4. 4.

    , . The telomere lengthening conundrum- it could be biology. Aging Cell 2017; 16: 312–319.

  5. 5.

    . How “reversible” is telomeric aging? Cancer Prev Res 2012; 5: 1163–1168.

  6. 6.

    , , , , , et al. Leukocyte telomere length and cardiovascular disease in the cardiovascular health study. Am J Epidemiol 2006; 165: 14–21.

  7. 7.

    , . A quantitative PCR method for measuring absolute telomere length. Biological Procedures 2011; 13: 3.

Download references

Author information


  1. Geisinger Obesity Research Institute, Geisinger, Danville, PA, USA

    • R Dershem
    • , X Chu
    • , G C Wood
    • , P Benotti
    •  & C D Still
  2. Weis Center for Research, Geisinger, Danville, PA, USA

    • R Dershem
  3. Department of Internal Medicine, Geisinger, Danville, PA, USA

    • D D Rolston


  1. Search for R Dershem in:

  2. Search for X Chu in:

  3. Search for G C Wood in:

  4. Search for P Benotti in:

  5. Search for C D Still in:

  6. Search for D D Rolston in:

Competing interests

The authors declare no conflict of interest.

Corresponding author

Correspondence to G C Wood.