Research Article
Print
Research Article
Circulating microRNAs in the context of the lifestyle intervention program MultiPill-Exercise: a pilot study
expand article infoRebecca Bankamp‡, Angelika Schmitt‡, Annunziata Fragasso‡, Simone Schweda‡, Manuel Widmann‡, Andreas Nie߇§, Inga Krau߇§, Barbara Munz‡§
‡ University Hospital Tübingen, Medical Clinic, Department of Sports Medicine, Tübingen, Germany
§ Interfaculty Research Institute for Sport and Physical Activity, Eberhard Karls University of Tübingen, Tübingen, Germany

Corresponding author: Barbara Munz ( barbara.munz@med.uni-tuebingen.de )

© 2026 Rebecca Bankamp, Angelika Schmitt, Annunziata Fragasso, Simone Schweda, Manuel Widmann, Andreas Nieß, Inga Krauß, Barbara Munz.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation: Bankamp R, Schmitt A, Fragasso A, Schweda S, Widmann M, Nieß A, Krauß I, Munz B (2026) Circulating microRNAs in the context of the lifestyle intervention program MultiPill-Exercise: a pilot study. Folia Medica 68(3): e176338. https://doi.org/10.3897/folmed.68.e176338
Open Access

Abstract

Introduction: Multimorbidity is a common finding in the context of lifestyle-related diseases, such as obesity, hypertonia, diabetes mellitus type 2, or osteoarthritis. However, little is known about the effects of exercise on subjects with multimorbidity. Biomarkers, such as microRNAs (miRs), might be useful tools in this context.

Aim: To analyze miRNA patterns in subjects with multimorbidity and to correlate them with adaptations to the exercise-focused lifestyle intervention program “MultiPill-Exercise Pilot Study.

Patients and methods: Using RNA-Seq and qPCR analysis, we determined both baseline expression and training-associated changes of several miRs and correlated them with changes in clinical and physiological parameters during the intervention.

Results: We found a negative correlation between changes in miR-486-5p concentrations and changes in peak power output ΔPPO. Remarkably, several changes in miR concentrations were associated with changes in blood pressure readings (RR) during the intervention: ΔmiR-150a-5p (**), ΔmiR-223-3p, ΔmiR130-3p, and ΔmiR126-3p positively correlated with ΔsystRR, and there was also a positive correlation between ΔmiR-150a-5p and ΔdiastRR.

Conclusion: These pilot data will be useful in generating hypotheses for further testing the potential of miRs as biomarkers in the context of exercise adaptation in multimorbidity. For this purpose, confirmatory, larger, randomized, and controlled studies should be performed, employing correction factors for multiple testing.

Keywords

exercise, miRs, multimorbidity

Introduction

Lifestyle-associated chronic adverse conditions, such as overweight/obesity, cardiovascular disease, type 2 diabetes mellitus (T2DM), or osteoarthritis (OA), are major burdens of disease in Western societies. Lifestyle intervention programs, specifically those that focus on physical activity and exercise, are effective in both the prevention and treatment of lifestyle-related diseases.[1] When designing exercise-based lifestyle intervention programs, biomarkers can be useful tools. They might help to find the most effective training concept for each patient. In addition, they could be employed to monitor individual training adaptation and thus be early indicators of treatment success.

Due to common risk factors, most patients suffering from lifestyle-related diseases are affected by multimorbidity, harboring diagnoses for more than one of the diseases. However, so far, studies on the effects of exercise programs on subjects with multimorbidity are rare, warranting more extensive research.[2]

MicroRNAs (miRs) are a class of small-molecule RNAs involved in the regulation of gene expression.[3] Recent data suggest that miR patterns, assessed in tissues such as skeletal muscle or in the circulation, might be useful tools to predict training adaptation and/or to monitor training effects.[4] In particular, patterns of circulating miRs have potential as biomarkers in clinical practice, since they can easily, quickly, and precisely be assessed from small volumes of blood.

The MultiPill-Exercise pilot study was an exercise-based lifestyle intervention program for patients with lifestyle-related diseases and multimorbidity, carried out at the Department of Sports Medicine of University Hospital Tübingen in 2019-2020.[5]

Aim

To analyze patterns of circulating miRs and their correlation with training-associated adaptation processes, such as changes in physical fitness or health-related clinical parameters, in subjects with multimorbidity.

Patients and methods

MultiPill-Exercise pilot study

The study design, as well as the physiological and clinical findings, have been previously described.[5-7] Briefly, subjects diagnosed with at least two out of four lifestyle-related diseases (or displaying relevant risk factors) were enrolled in the study. Medication was documented at an initial screening appointment as well as at all diagnostics timepoints (T0, T1, and T2, see below). Subjects were eligible for the study when medication for study-relevant diseases had been unchanged for at least three months prior to enrollment. Otherwise, there were no restrictions with regard to medication types or subclasses, except cases in which the screening physician considered a specific form of medication an exclusion criterion for study participation.[5,6] After enrollment, subjects underwent a 24-week exercise-focused lifestyle intervention program divided into two blocks of twelve weeks, with the first block (T0 → T1) being closely supervised, whereas the second block (T1 → T2) was more self-directed. Training regimens were designed as combinations of a “general” exercise sessions, combined with disease-specific and elective elements.[5,6] At baseline (T0) and after each of the two study blocks (T1 and T2), an extensive diagnostics battery was carried out, including collection of blood specimens. The study was carried out with two groups of patients (wave #1 and wave #2), starting in 08-2019 and 01-2020, respectively. The study was approved by the Ethics Committee of University Medicine Tübingen (Grant number: 298/2019BO2, 06/04/2019). Before blood drawing, written informed consent was obtained from all subjects.

Patients

Initially, 39 subjects were enrolled in the study, 20 for wave #1 and 19 for wave #2. Of the latter, 4 subjects were excluded at T0, so that for T0, data of 35 subjects were available. For T1, only data of 20 subjects could be evaluated, since, due to the COVID 19 pandemic, the T1 diagnostics timepoint was cancelled for subjects of wave #2. Thus, in the miR analysis described here, only data obtained from subjects of wave #1 were included. In addition, (T0 → T1) data were only available for 18 subjects, since for medical reasons (herniated intervertebral disc and hip pain, respectively) two subjects (MP1906 and MP1913) of wave #1 did not complete spiroergometry at the T1 timepoint. Finally, for 2 subjects (MP1901 - missing value at T0 - and MP1925 - missing value at T1), ΔFibrinogen could not be determined due to a technical problem in the laboratory.

RNA-Seq

RNA-Seq was carried out by GenXPro GmbH (Frankfurt / Main, Germany), using the TrueQuant method for analysis of small RNA expression as previously described.[8] For this purpose, the Small RNA Sequencing Kit (v. 1.0) and Illumina®NGS technology were employed.

RNA isolation from plasma and qPCR analysis

RNA isolation was carried out using the miRNeasy Serum/Plasma Advanced Kit (Qiagen). qPCR was carried out using the miRCURY LNA RT Kit and the miRCURY LNA miRNA PCR Assay (Qiagen), in combination with pre-designed primers (Qiagen). For normalization, employing RefFinder software[9], the three most stable housekeeping genes, miRs, 590-5p, 26b-5p und 320d, were chosen from a set of seven.

Statistical analysis

Statistical analysis was carried out using SPSS Statistics Version 26.0 (IBM Corporation, Armonk, New York). Data were tested for normal distribution using Kolmogo- rov–Smirnov and Shapiro–Wilk statistical tests. Since the respective prerequisites were not met by the data, Wilcoxon testing was employed to assess possible differences. For correlation analyses, Spearman rank correlation coefficient r was calculated.[10,11] Statistical significance was accepted at p<0.05 (*) and p<0.01 (**). Since this was an exploratory, hypothesis-generating pilot study, no corrections for multiple testing were carried out.

Results

In an initial, untargeted screening approach, 12 samples (taken at T0 and T1) of 6 subjects of wave #1 were screened by RNA-Seq. They were selected as a maximally homogeneous group with regard to sex and BMI to minimize scattering of data. Furthermore, female sex was chosen since the majority of subjects in the entire cohort (27/39) was female. Mean age of selected subjects (55.17±6.49 years) corresponded well with mean age of the entire cohort, which was 55.37±10.62 years. For subjects’ characteristics, please refer to Supplementary Tables S1 and S2. )

Baseline data: RNA-Seq-based screening of 6 subjects

From the RNA-Seq data set, 13 miRs were selected for further analysis based on the following criteria: a minimum of 50 “reads” in each sample, and a correlation coefficient of at least ±0.6 between baseline levels and changes in one of the 16 clinical parameters analyzed in this study (Supplementary Table S3). In some cases, there were similar (or inverted) correlation readings for different miRs. This is due to the algorithm of Spearman rank correlation, which, for small n numbers, only allows a limited amount of permutations.

Baseline data: qPCR-based data of 18 out of 20 subjects

For the selected miRs, concentrations at T0 and T1 were analyzed in the 20 subjects of wave #1 via qPCR. Similarly as for RNA-Seq data, we again analyzed correlations between miRNA concentrations and changes in clinical parameters. As shown in Supplementary Table S4, miR correlation patterns as analyzed by qPCR only partially corresponded to those obtained in the RNA-Seq analysis. In addition, most correlations were only trends and not significant.

Significant positive correlations were found between baseline expression of miR-130b-3p and ΔCRP and that of miR-486-5p with Δtriglycerides. A significant negative correlation could be observed between miR-103a-3p baseline concentration and changes in BMI. However, only in case of miR-486-5p, data from RNA-Seq and qPCR analysis were consistent (Supplementary Tables S3 and S4).

Progression: qPCR data

Based on these results, we analyzed changes in miR patterns during the intervention and correlated them with changes in clinical parameters over time. Supplementary Table S5 shows fold changes of the 13 selected miRs after 12 weeks of training. During the first twelve weeks of the intervention, miR concentrations increased or declined. Several of these changes in expression correlated with changes in physiological and/or clinical parameters (Supplementary Table S6). Again, similarly as in the qPCR analysis of baseline concentrations, for ΔmiR-103a-3p, there was a negative correlation with ΔBMI. Furthermore, ΔmiR-190a-5p negatively correlated with ΔV˙O2max, whereas ΔmiR-486-5p (Fig. 1) and ΔmiR-210-3p concentrations (negatively) paralleled ΔPPO. In addition, there was a positive correlation between ΔmiR-143-3p and Δinsulin. Remarkably, several changes in miR concentrations were associated with changes in blood pressure readings (RR) during the intervention: ΔmiR-150a-5p (**), ΔmiR-223-3p, ΔmiR130-3p, and ΔmiR126-3p positively correlated with ΔsystRR, and there was also a positive correlation between ΔmiR-150a-5p and ΔdiastRR. Fig. 2 illustrates correlations between ΔmiR-150a-5p and ΔsytRR as well as ΔdiastRR.

Figure 1.

Correlation between changes in miR-486-5p concentrations and changes in PPO between T0 and T1.

Figure 2.

Correlation between changes in miR-150a-5p concentrations and changes in both diastolic and systolic blood pressure between T0 and T1.

Taken together, our data suggest that miRs might be promising biomarkers in the context of lifestyle intervention programs for multimorbid subjects.

Discussion

We could demonstrate that circulating miRs, assessed in subjects with multimorbidity, might be predictive as well as monitoring biomarkers in the context of an exercise-based lifestyle intervention program. Still, this was a pilot-character, hypothesis-generating study with a limited number of participants, thus not employing correction algorithms for multiple testing, a factor limiting the clinical significance of the data at this time.

Initially, to identify candidate marker miRs, we carried out RNA-Seq-based screening of 6 study participants, followed by qPCR analysis in a larger number of subjects (18 out of 20). Obviously, miR correlation patterns as analyzed by qPCR only partially corresponded to those obtained in the initial RNA-Seq screening analysis, indicating merely moderate efficacy of our screening procedure. In particular, significant but opposed correlations were found for miR-130b-3p and ΔCRP and for miR-103a-3p and ΔBMI. Thus, in future, comparable studies, RNA-Seq-based screening should potentially be carried out with more than six subjects, and/or a more refined strategy should be applied to identify the samples best suited for screening, i.e., with the highest degree of representativeness for the whole cohort. In addition, with regard to miRs as “monitoring” biomarkers, i.e., for changes during the intervention, a separate screening should be carried out, since it is likely that here, different miR markers are most efficient.

With regard to correlations of miR patterns (both baseline and changes/Δs) and changes in physiological and clinical parameters during the intervention, a negative correlation of baseline miR-486-5p concentrations with Δtriglycerides was found. On average, expression of this miR decreased during the intervention, as has previously been demonstrated in the context of the HERITAGE study, where up-to-then sedentary subjects underwent a 20-week endurance training program.[12] Interestingly, associations of this miR with fatty acid metabolism had previously been demonstrated.[13,14]

As mentioned and in contrast to baseline levels, changes (Δs) in miR concentrations between T0 and T1 were only assessed by qPCR, which limits the validity of the respective set of data. Nevertheless, several interesting correlations were found. Specifically, for ΔmiR-103a-3p, there was a negative correlation with ΔBMI. Interestingly, this miR has previously been implicated in metabolic disorders.[15] Furthermore, ΔmiR-190a-5p negatively correlated with ΔV˙O2max, whereas ΔmiR-486-5p (see above) and ΔmiR-210-3p concentrations (negatively) paralleled ΔPPO. For miR-190a-5p, associations with a lower metabolic risk and a higher degree of physical activity have been demonstrated in older adults[16], which might contradict our data but might also be due to the fact that in this study, inter-individual differences were assessed, not, as in our study, changes over time. Interestingly, miR-210-3p has been implicated in the regulation of angiogenesis and diabetes.[17] In addition, there was a positive correlation between ΔmiR-143-3p and Δinsulin. Remarkably, this miR species has been widely implicated in the regulation of insulin sensitivity and metabolic syndrome.[18] In particular, increasing concentrations of miR-143-3p have previously been associated with low improvement of glycemic control in response to an 8-week program of regular physical exercise[19], however, in the respective study, skeletal muscle miR was assessed, so that results might not be directly comparable with our c-miR analysis.

Finally and most interestingly, several changes in miR concentrations significantly correlated with changes in blood pressure readings (RR) during the intervention: ΔmiR-150a-5p (**), ΔmiR-223-3p, ΔmiR130b-3p, and ΔmiR126-3p positively correlated with ΔsystRR, and there was also a positive correlation between ΔmiR-150a-5p and ΔdiastRR. Of these, miR-223-3p[20] and miR-126-3p[21] have already been linked to hypertension. However, it is very likely that changes in blood pressure readings during the intervention might not exclusively be associated with increased physical activity or, more generally, lifestyle changes: antihypertensive agents were the most common type of medication already at enrollment (24/39[5]) and it is very likely that, with more focus on health issues, patients’ medication might also have been adapted, especially with regard to dosage and type. Thus, correlations with miR patterns will have to be tested in better-controlled settings in the future.

Conclusion

Taken together, our data suggest that c-miRs might be easily accessible biomarkers in the context of lifestyle intervention programs for subjects with multimorbidity. However, this was an exploratory, pilot-character, hypothesis-generating study. Its major limitation was the small sample size in both RNA-Seq-based screening and qPCR analysis. In addition, to initially cover a maximum of potentially interesting miR species and corresponding to the pilot character of the study, no correction for multiple testing was carried out. Thus, candidate miRs identified in this study should be tested in larger, confirmatory, possibly multicentric, prospective, randomized, and controlled studies, such as the MultiPill-Exercise Main Study, which was started at our institution in 2022.[6] In this larger sample, in addition to analysis of the complete study population, it will also be possible to perform subgroup analyses with specific patient subcohorts, e.g., the “obesity cohort.” In the future, the inclusion of a healthy, age-matched control group might also be interesting, allowing discrimination between “general” and disease-specific effects. Finally, it might be possible to define patterns of specific miR markers as determinants for the design and supervision of individualized training regimens (“personalized medicine”).

Acknowledgements

The authors would like to express their gratitude to all members of the Department of Sports Medicine and the complete MultiPill-Exercise team for their support.

References

  • 3. Makarova JA, Shkurnikov MU, Wicklein D, et al. Intracellular and extracellular microRNA: An update on localization and biological role. Prog Histochem Cytochem 2016; 51(3-4):33–49. doi: 10.1016/j.proghi.2016.06.001; PMID: 27396686.
  • 4. Domańska-Senderowska D, Laguette MN, Jegier A, et al. MicroRNA profile and adaptive response to exercise training: a review. Int J Sports Med 2019; 40(4):227–35. doi: 10.1055/a-0824-4813; PMID: 30791082.
  • 5. Schweda S, Munz B, Burgstahler C, et al. Proof of concept of a 6-month person-oriented exercise intervention ‘MultiPill-Exercise’ among patients at risk of or with multiple chronic diseases: results of a one-group pilot trial. Int J Environ Res Public Health 2022; 19(15):9469. doi: 10.3390/ijerph19159469; PMID: 35954825; PMCID: PMC9368673.
  • 6. Schweda S, Müller G, Munz B, et al. Implementation and evaluation of an individualized physical exercise promotion program in people with manifested risk factors for multimorbidity (MultiPill-Exercise): a study protocol for a pragmatic randomized controlled trial. BMC Public Health 2022; 22(1):1174. doi: 10.1186/s12889-022-13400-9; PMID: 35698074; PMCID: PMC9190168.
  • 7. Maturana FM, Rolf R, Schweda S, et al. Adipokines as predictive biomarkers for training adaptation in subjects with multimorbidity - a hypothesis-generating study. J Clin Med 2023; 12(13):4376. doi: 10.3390/jcm12134376; PMID: 37445411; PMCID: PMC10342946.
  • 9. Finotello F, Di Camillo B. Measuring differential gene expression with RNA-seq: challenges and strategies for data analysis. Brief Funct Genomics 2015; 14(2):130–42.
  • 10. Sachs L. [Applied statistics: application of statistical methods]. 11th ed. Berlin: Springer; 2004. 889 p. (Springer eBook Collection: Life Science and Basic Disciplines). [German].
  • 11. Pospeschill M. [Statistical methods: structures, basics, applications in psychology and social sciences]. 1st ed. München: Elsevier, Spektrum Akademischer Verlag; 2006. 484 p. [German].
  • 13. Ortega FJ, Cardona-Alvarado MI, Mercader JM, et al. Circulating profiling reveals the effect of a polyunsaturated fatty acid-enriched diet on common microRNAs. J Nutr Biochem 2015; 26(10):1095–101. doi: 10.1016/j.jnutbio.2015.05.001; PMID: 26092372.
  • 18. Xihua L, Shengjie T, Weiwei G, et al. Circulating miR-143-3p inhibition protects against insulin resistance in metabolic syndrome via targeting of the insulin-like growth factor 2 receptor. Transl Res 2019; 205:33–43. doi: 10.1016/j.trsl.2018.09.006; PMID: 30392876.
  • 19. Dreher SI, Höckele S, Huypens P, et al. TGF-β induction of miR-143/145 is associated to exercise response by influencing differentiation and insulin signaling molecules in human skeletal muscle. Cells 2021; 10(12):3443. doi: 10.3390/cells10123443; PMID: 34943951; PMCID: PMC8700369.
  • 21. Solís-Toro D, Mosquera Escudero M, García-Perdomo HA. Association between circulating microRNAs and the metabolic syndrome in adult populations: A systematic review. Diabetes Metab Syndr 2022; 16(1):102376. doi: 10.1016/j.dsx.2021.102376; PMID: 34954485.

Appendix

Table A1.
Download as
CSV
XLSX

Overview of the complete study cohort

Demographic data
Sex F (n=25) M (n=10) Total (n=35)
Age (years) Mean=56.04 SD±2.06 Mean=53.7 SD±3.71 Mean=55.37 SD±10.62
Height (m) Mean=1.66 SD±0.01 Mean=1.82 SD±0.02 Mean=1.7 SD±0.09
Weight, BMI and ΔV˙O2max
T0 (n=35) T1 (n=20) T2 (n=30)
F M F M F M
Weight (kg) Mean=85.73 SD±1.96 Mean=102.29 SD±3.88 Mean=82.22 SD±10.33 Mean=103.96 SD±16.06 Mean=84.83 SD±9.64 Mean=98.53 SD±15.83
Mean=90.46 SD±12.86 Mean=88.95 SD±14.7 Mean=88.94 SD±13.02
BMI (kg/m2) Mean=31.12 SD±0.69 Mean=30.98 SD±0.86 Mean=29.97 SD±3.54 Mean=30.67 SD±2.28 Mean=30.76 SD±3.68 Mean=29.81 SD±3.01
Mean=31.08 SD±3.21 Mean=30.09 SD±3.18 Mean=30.47 SD±3.47
Relative ΔV˙O2max (mL/kg·min) Mean=18.89 SD±0.79 Mean=23.2 SD±1.15 Mean=21.92 SD±5.04 Mean=26.8 SD±3.27 Mean=20.71 SD±5.59 Mean=26.33 SD±3.01
Mean=20.12 SD±4.28 Mean=23.28 SD±5.05 Mean=22.41 SD±5.56
Inclusion criteria
Diseases No Risk Manifested
Obesity n=1 n=8 n=26
T2DM n=7 n=25 n=3
Hypertonia n=8 n=2 n=25
OA n=9 n=9 n=17
Table A2.
Download as
CSV
XLSX

Overview of the six subjects selected for RNA-Seq analysis

Subject ID Sex Age BMI (T0) (kg/m2) Overweight / Obesity Hypertonia T2DM Osteoarthritis IC
MP1902 F 54 27.12 Yes Yes No Risk 3
MP1903 F 46 31.24 Yes Yes Risk No 3
MP1907 F 57 32.72 Yes No Risk Yes 3
MP1912 F 61 32.79 Yes Yes Risk No 3
MP1915 F 63 26.62 No Yes No Yes 2
MP1919 F 50 28.87 Yes Yes Risk Yes 4
Mean 55.17 29.89 3
SD 6.49 2.74 0.63
MD 55.51 30.05 3

IC: number of inclusion criteria met
Table A3.
Download as
CSV
XLSX

RNA-Seq data (n=6). Correlations between miR baseline concentrations and changes in physiological and clinical data between T0 and T1

Δ clinical parameters between T0 and T1 Statistics miR baseline concentration (reads) at t0
16-5p 486-5p 210-3p 146a-5p 150a-5p 190a-5p 223-3p 143-3p 130b-3p 505-3p 21-5p 126-3p 103a-3p
Physical performance ΔrelV˙O2max (ml/kg·min) Correlation Coefficient (𝝆) 0.200 0.429 0.314 −0.200 −0.143 0.200 0.086 0.086 0.600 0.371 0.086 0.143 0.143
Sign. (2-sided) 0.704 0.397 0.544 0.704 0.787 0.704 0.872 0.872 0.208 0.468 0.872 0.787 0.787
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔPPO (W‎/kg) Correlation Coefficient (𝝆) 0.174 0.290 0.348 −0.232 0.232 0.174 −0.493 −0.406 0.116 −0.087 0.087 −0.232 −0.174
Sign. (2-sided) 0.742 0.577 0.499 0.658 0.658 0.742 0.321 0.425 0.827 0.870 0.870 0.658 0.742
n 6 6 6 6 6 6 6 6 6 6 6 6 6
Body weight ΔBMI (kg/m2) Correlation Coefficient (𝝆) −0.886* −0.657 −0.771 0.600 −0.029 −0.886* −0.143 0.257 −0.829* −0.771 −0.829* 0.029 −0.486
Sign. (2-sided) 0.019 0.156 0.072 0.208 0.957 0.019 0.787 0.623 0.042 0.072 0.042 0.957 0.329
n 6 6 6 6 6 6 6 6 6 6 6 6 6
Inflammation ΔIL6 (ng‎/L) Correlation Coefficient (𝝆) 0.486 0.086 −0.086 −0.029 −0.600 0.486 −0.086 0.314 0.086 −0.086 0.714 0.600 0.029
Sign. (2-sided) 0.329 0.872 0.872 0.957 0.208 0.329 0.872 0.544 0.872 0.872 0.111 0.208 0.957
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔCRP (mg‎/dL) Correlation Coefficient (𝝆) −0.029 −0.257 0.543 0.200 0.143 −0.029 0.600 −0.943** 0.600 0.314 −0.086 −0.143 0.714
Sign. (2-sided) 0.957 0.623 0.266 0.704 0.787 0.957 0.208 0.005 0.208 0.544 0.872 0.787 0.111
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔFibrinogen (mg‎/dL) Correlation Coefficient (𝝆) −0.314 −0.143 −0.657 0.029 0.143 −0.314 −0.543 0.600 −0.943** −0.543 −0.257 −0.143 −0.714
Sign. (2-sided) 0.544 0.787 0.156 0.957 0.787 0.544 0.266 0.208 0.005 0.266 0.623 0.787 0.111
n 6 6 6 6 6 6 6 6 6 6 6 6 6
Diabetes ΔInsulin (pmol‎/L) Correlation Coefficient (𝝆) 0.429 0.200 0.771 −0.143 0.029 0.429 0.257 −0.829* 0.829* 0.429 0.371 −0.029 0.600
Sign. (2-sided) 0.397 0.704 0.072 0.787 0.957 0.397 0.623 0.042 0.042 0.397 0.468 0.957 0.208
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔHbA1c (%) Correlation Coefficient (𝝆) −0.334 −0.395 0.030 0.152 0.516 −0.334 0.213 −0.516 −0.273 −0.030 −0.395 −0.516 0.152
Sign. (2-sided) 0.518 0.439 0.954 0.774 0.295 0.518 0.686 0.295 0.600 0.954 0.439 0.295 0.774
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔFasting glucose (mg‎/dL) Correlation Coefficient (𝝆) 0.143 0.086 0.257 −0.371 0.771 0.143 −0.086 −0.371 −0.257 0.257 0.029 −0.771 0.029
Sign. (2-sided) 0.787 0.872 0.623 0.468 0.072 0.787 0.872 0.468 0.623 0.623 0.957 0.072 0.957
n 6 6 6 6 6 6 6 6 6 6 6 6 6
Lipoprotein / Fatty acid metabolism ΔCholesterol (mg‎/dL) Correlation Coefficient (𝝆) −0.486 −0.314 −0.714 0.257 −0.086 −0.486 0.029 0.714 −0.657 −0.314 −0.429 0.086 −0.371
Sign. (2-sided) 0.329 0.544 0.111 0.623 0.872 0.329 0.957 0.111 0.156 0.544 0.397 0.872 0.468
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔTriglycerides (mg‎/dL) Correlation Coefficient (𝝆) −0.086 −0.600 0.143 0.371 0.029 −0.086 0.600 −0.714 0.086 0.029 0.029 −0.029 0.600
Sign. (2-sided) 0.872 0.208 0.787 0.468 0.957 0.872 0.208 0.111 0.872 0.957 0.957 0.957 0.208
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔLDL (mg‎/dL) Correlation Coefficient (𝝆) 0.029 −0.086 −0.314 −0.086 0.086 0.029 0.200 0.486 −0.371 0.143 0.086 −0.086 −0.029
Sign. (2-sided) 0.957 0.872 0.544 0.872 0.872 0.957 0.704 0.329 0.468 0.787 0.872 0.872 0.957
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔHDL (mg‎/dL) Correlation Coefficient (𝝆) 0.371 −0.143 0.029 0.200 −0.714 0.371 0.486 0.086 0.429 0.143 0.600 0.714 0.486
Sign. (2-sided) 0.468 0.787 0.957 0.704 0.111 0.468 0.329 0.872 0.397 0.787 0.208 0.111 0.329
n 6 6 6 6 6 6 6 6 6 6 6 6 6
Cardiovascular system ΔdiastRR (mmHg) Correlation Coefficient (𝝆) 0.754 0.319 0.348 −0.493 0.116 0.754 0.232 0.029 0.232 0.609 0.812* −0.116 0.377
Sign. (2-sided) 0.084 0.538 0.499 0.321 0.827 0.084 0.658 0.957 0.658 0.2 0.05 0.827 0.461
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔsysRR (mmHg) Correlation Coefficient (𝝆) 0.257 −0.486 0.143 0.314 −0.371 0.257 0.771 −0.429 0.314 0.200 0.486 0.371 0.771
Sign. (2-sided) 0.623 0.329 0.787 0.544 0.468 0.623 0.072 0.397 0.544 0.704 0.329 0.468 0.072
n 6 6 6 6 6 6 6 6 6 6 6 6 6
ΔResting HR (bpm) Correlation Coefficient (𝝆) 0.319 0.493 0.551 −0.696 0.986** 0.319 −0.174 −0.377 0.029 0.580 0.058 −0.986** 0.029
Sign. (2-sided) 0.538 0.321 0.257 0.125 0 0.538 0.742 0.461 0.957 0.228 0.913 0 0.957
n 6 6 6 6 6 6 6 6 6 6 6 6 6
Table A4.
Download as
CSV
XLSX

qPCR data (n=18/20). Correlations between miR baseline concentrations and changes in physiological and clinical data between T0 and T1

Δ clinical parameters between T0 and T1 Statistics miR baseline concentration at t0
16-5p 486-5p 210-3p 146a-5p 150a-5p 190a-5p 223-3p 143-3p 130b-3p 505-3p 21-5p 126-3p 103a-3p
Physical performance ΔrelV˙O2max (ml/kg·min) Correlation Coefficient (𝝆) 0.059 0.229 0.289 −0.372 −0.28 0.121 −0.316 −0.136 0.269 −0.227 −0.049 −0.091 0.33
Sign. (2-sided) 0.815 0.361 0.245 0.128 0.261 0.633 0.201 0.590 0.280 0.366 0.847 0.718 0.182
n 18 18 18 18 18 18 18 18 18 18 18 18 18
ΔPPO (W/kg) Correlation Coefficient (𝝆) 0.215 0.076 0.165 −0.244 −0.255 0.285 −0.363 −0.251 0.272 −0.291 −0.163 −0.184 0.348
Sign. (2-sided) 0.391 0.765 0.513 0.329 0.308 0.252 0.138 0.316 0.275 0.242 0.518 0.464 0.156
n 18 18 18 18 18 18 18 18 18 18 18 18 18
Body weight ΔBMI (kg/m2) Correlation Coefficient (𝝆) 0.023 0.043 0.112 0.111 0.223 0.194 0.051 0.053 −0.014 0.102 0.199 0.117 0.557*
Sign. (2-sided) 0.925 0.858 0.638 0.640 0.34 4 0.412 0.830 0.823 0.955 0.670 0.400 0.624 0.011
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Inflammation ΔIL6 (ng‎/L) Correlation Coefficient (𝝆) −0.086 −0.178 −0.093 0.095 −0.084 0.212 0.107 0.227 0.008 −0.029 −0.027 0.172 −0.17
Sign. (2-sided) 0.718 0.454 0.696 0.691 0.724 0.369 0.654 0.336 0.975 0.902 0.910 0.470 0.474
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔCRP (mg‎/dL) Correlation Coefficient (𝝆) −0.142 0.035 −0.164 −0.026 −0.032 −0.082 −0.287 −0.09 −0.445* −0.295 −0.282 −0.153 −0.179
Sign. (2-sided) 0.552 0.882 0.490 0.913 0.892 0.731 0.221 0.706 0.049 0.207 0.228 0.518 0.450
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔFibrinogen (mg‎/dL) Correlation Coefficient (𝝆) −0.401 −0.463 −0.438 0.127 −0.009 −0.377 0.205 0.206 −0.158 0.124 0.11 0.045 −0.071
Sign. (2-sided) 0.099 0.053 0.069 0.617 0.971 0.123 0.416 0.412 0.531 0.624 0.665 0.858 0.779
n 18 18 18 18 18 18 18 18 18 18 18 18 18
Diabetes ΔInsulin (pmol‎/L) Correlation Coefficient (𝝆) −0.058 −0.111 0.123 −0.167 −0.114 0.072 −0.202 −0.028 0.276 −0.284 0.021 0.114 0.188
Sign. (2-sided) 0.808 0.642 0.604 0.481 0.631 0.762 0.392 0.906 0.239 0.224 0.930 0.631 0.427
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔHbA1c (%) Correlation Coefficient (𝝆) −0.065 −0.158 −0.041 0.105 −0.016 −0.044 0.067 0.038 0.139 0.12 0.206 0.324 −0.023
Sign. (2-sided) 0.786 0.507 0.863 0.660 0.947 0.855 0.779 0.872 0.560 0.613 0.383 0.163 0.924
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔFasting glucose (mg‎/dL) Correlation Coefficient (𝝆) 0.115 0.127 0.176 −0.242 0.317 −0.074 −0.244 −0.138 0.028 −0.227 −0.136 0.003 −0.232
Sign. (2-sided) 0.628 0.595 0.459 0.304 0.174 0.756 0.299 0.561 0.907 0.336 0.568 0.990 0.325
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Lipoprotein / Fatty acid metabolism ΔCholesterol (mg‎/dL) Correlation Coefficient (𝝆) −0.239 −0.186 −0.134 0.126 0.088 −0.16 0.087 0.314 0.138 0.062 0.128 0.226 −0.161
Sign. (2-sided) 0.309 0.433 0.574 0.595 0.712 0.502 0.715 0.177 0.561 0.794 0.591 0.337 0.498
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔTriglycerides (mg‎/dL) Correlation Coefficient (𝝆) −0.377 −0.463* −0.007 0.277 −0.363 −0.117 0.102 0.314 0.113 0.031 0.277 0.283 0.12
Sign. (2-sided) 0.102 0.04 0.977 0.236 0.116 0.622 0.668 0.178 0.636 0.897 0.237 0.227 0.616
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔLDL (mg‎/dL) Correlation Coefficient (𝝆) −0.149 −0.074 −0.222 −0.203 −0.021 −0.349 −0.183 0.011 −0.044 −0.193 −0.168 −0.06 −0.393
Sign. (2-sided) 0.531 0.757 0.347 0.39 0.93 0.131 0.441 0.962 0.855 0.416 0.478 0.801 0.086
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔHDL (mg‎/dL) Correlation Coefficient (𝝆) −0.207 −0.035 −0.341 −0.373 −0.319 −0.066 −0.334 −0.13 −0.255 −0.359 −0.398 −0.298 −0.131
Sign. (2-sided) 0.381 0.882 0.141 0.105 0.17 0.782 0.149 0.584 0.277 0.12 0.082 0.201 0.582
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Cardiovascular System ΔdiastRR (mmHg) Correlation Coefficient (𝝆) 0.164 0.048 0.108 −0.242 −0.169 −0.178 −0.305 −0.229 0.114 −0.243 −0.201 −0.215 −0.042
Sign. (2-sided) 0.491 0.84 0.649 0.305 0.477 0.452 0.191 0.332 0.633 0.301 0.395 0.362 0.86
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔsysRR (mmHg) Correlation Coefficient (𝝆) 0.022 0.017 −0.06 −0.244 −0.386 −0.031 −0.341 −0.433 −0.378 −0.302 −0.205 −0.434 0.029
Sign. (2-sided) 0.926 0.945 0.801 0.3 0.092 0.896 0.141 0.057 0.1 0.196 0.386 0.056 0.905
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Δ Resting Heart Rate (bpm) Correlation Coefficient (𝝆) 0.005 0.023 −0.164 0.086 0.071 0.175 0.236 0.089 0.044 0.178 0.118 0.114 −0.096
Sign. (2-sided) 0.985 0.928 0.516 0.734 0.778 0.488 0.345 0.725 0.864 0.479 0.641 0.653 0.704
n 18 18 18 18 18 18 18 18 18 18 18 18 18
Table A5.
Download as
CSV
XLSX

qPCR data (n=20). Changes in miR concentrations between T0 and T1

log2-fold change
miR N Min Max Mean SD
16-5p 20 −2.598 1.100 −1.022 1.107
486-5p 20 −3.478 1.050 −1.551 1.365
210-3p 20 −2.187 1.600 −0.405 1.028
146a-5p 20 −1.627 2.652 0.590 1.073
150a-5p 20 −2.197 2.015 0.106 0.948
190a-5p 20 −2.162 1.495 −0.645 1.060
223-3p 20 −1.732 2.062 0.389 0.954
143-3p 20 −1.452 2.308 0.354 0.907
130b-3p 20 −1.632 2.207 0.090 0.939
505-3p 20 −2.002 1.447 0.225 0.899
21-5p 20 −2.067 1.810 0.250 1.103
126-3p 20 −2.307 1.792 0.223 1.039
103a-3p 20 −2.222 1.887 −0.044 1.025
Table A6.
Download as
CSV
XLSX

qPCR data (n=18/20). Correlations between changes in miR concentrations and changes in physiological and clinical data, both between T0 and T1

Δ clinical parameters Statistics miR fold change (log2) after 12 weeks of training intervention
between T0 and T1 16-5p 486-5p 210-3p 146a-5p 150a-5p 190a-5p 223-3p 143-3p 130b-3p 505-3p 21-5p 126-3p 103a-3p
Physical performance ΔrelV˙O2max (ml/kg·min) Correlation Coefficient (𝝆) −0.307 −0.390 −0.336 0.214 −0.048 −0.514* 0.261 −0.023 −0.201 0.259 0.166 −0.014 −0.180
Sign. (2-sided) 0.216 0.110 0.173 0.394 0.851 0.029 0.296 0.928 0.425 0.300 0.510 0.958 0.475
n 18 18 18 18 18 18 18 18 18 18 18 18 18
ΔPPO (W‎/kg) Correlation Coefficient (𝝆) −0.406 −0.517* −0.504* 0.227 −0.093 −0.355 0.278 0.370 −0.289 0.406 0.223 0.129 0.044
Sign. (2-sided) 0.095 0.028 0.033 0.365 0.714 0.149 0.264 0.131 0.245 0.095 0.374 0.610 0.863
n 18 18 18 18 18 18 18 18 18 18 18 18 18
Body weight ΔBMI (kg/m2) Correlation Coefficient (𝝆) −0.167 −0.178 −0.335 −0.083 −0.159 −0.310 −0.058 −0.165 −0.233 −0.150 −0.285 −0.256 −0.508*
Sign. (2-sided) 0.482 0.452 0.149 0.726 0.504 0.184 0.808 0.488 0.322 0.529 0.223 0.276 0.022
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Inflammation ΔIL6 (ng‎/L) Correlation Coefficient (𝝆) −0.078 0.011 −0.081 −0.357 −0.345 0.072 −0.233 −0.143 −0.254 −0.360 −0.270 −0.433 −0.096
Sign. (2-sided) 0.745 0.965 0.733 0.122 0.137 0.762 0.322 0.548 0.279 0.119 0.249 0.057 0.689
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔCRP (mg‎/dL) Correlation Coefficient (𝝆) 0.244 0.247 0.344 −0.053 0.065 0.143 0.017 0.023 0.317 0.055 0.343 0.396 −0.148
Sign. (2-sided) 0.299 0.294 0.138 0.823 0.784 0.548 0.945 0.922 0.174 0.818 0.139 0.084 0.533
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔFibrinogen (mg‎/dL) Correlation Coefficient (𝝆) 0.215 0.086 0.243 −0.077 −0.144 0.298 −0.132 0.020 0.153 −0.071 −0.186 −0.018 0.379
Sign. (2-sided) 0.392 0.735 0.332 0.760 0.570 0.231 0.601 0.938 0.545 0.779 0.460 0.945 0.121
n 18 18 18 18 18 18 18 18 18 18 18 18 18
Diabetes ΔInsulin (pmol‎/L) Correlation Coefficient (𝝆) 0.061 0.059 −0.138 0.021 −0.189 −0.156 0.084 0.456* −0.279 0.287 0.002 −0.158 0.130
Sign. (2-sided) 0.799 0.803 0.563 0.930 0.425 0.512 0.726 0.043 0.233 0.220 0.992 0.506 0.584
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔHbA1c (%) Correlation Coefficient (𝝆) 0.186 0.217 0.148 −0.140 0.282 −0.007 −0.065 0.062 0.049 −0.011 −0.139 −0.032 −0.107
Sign. (2-sided) 0.433 0.358 0.534 0.556 0.228 0.977 0.786 0.796 0.838 0.964 0.558 0.894 0.655
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔFasting glucose (mg‎/dL) Correlation Coefficient (𝝆) −0.029 0.027 −0.166 0.231 0.014 0.229 0.246 0.403 −0.151 0.187 0.182 −0.060 0.371
Sign. (2-sided) 0.905 0.910 0.485 0.326 0.952 0.331 0.296 0.078 0.526 0.430 0.443 0.801 0.107
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Lipoprotein / Fatty acid metabolism ΔCholesterol (mg‎/dL) Correlation Coefficient (𝝆) 0.235 0.171 0.306 −0.171 0.075 0.141 −0.058 −0.337 −0.034 −0.096 −0.160 −0.241 −0.034
Sign. (2-sided) 0.318 0.472 0.189 0.472 0.753 0.554 0.808 0.146 0.887 0.686 0.500 0.305 0.887
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔTriglycerides (mg‎/dL) Correlation Coefficient (𝝆) 0.249 0.314 0.089 −0.360 0.274 0.013 −0.147 −0.252 −0.142 −0.030 −0.226 −0.092 −0.182
Sign. (2-sided) 0.290 0.177 0.710 0.119 0.243 0.957 0.535 0.284 0.550 0.900 0.337 0.700 0.442
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔLDL (mg‎/dL) Correlation Coefficient (𝝆) 0.254 0.140 0.378 0.131 0.281 0.238 0.236 −0.083 0.196 0.188 0.059 0.008 0.245
Sign. (2-sided) 0.279 0.556 0.100 0.582 0.230 0.311 0.316 0.729 0.407 0.427 0.803 0.972 0.297
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔHDL (mg‎/dL) Correlation Coefficient (𝝆) 0.081 −0.066 0.149 0.014 −0.009 0.030 0.102 −0.114 0.227 0.011 0.158 0.014 −0.081
Sign. (2-sided) 0.733 0.784 0.530 0.952 0.970 0.900 0.667 0.633 0.335 0.962 0.505 0.952 0.733
n 20 20 20 20 20 20 20 20 20 20 20 20 20
Cardiovascular system ΔdiastRR (mmHg) Correlation Coefficient (𝝆) −0.050 −0.064 −0.169 0.299 0.491* −0.013 0.411 0.244 0.092 0.371 0.167 0.073 0.329
Sign. (2-sided) 0.835 0.789 0.477 0.200 0.028 0.957 0.072 0.300 0.700 0.107 0.481 0.759 0.156
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔsysRR (mmHg) Correlation Coefficient (𝝆) 0.006 −0.030 0.067 0.375 0.657** −0.038 0.453* 0.188 0.514* 0.366 0.275 0.514* −0.092
Sign. (2-sided) 0.980 0.900 0.779 0.103 0.002 0.875 0.045 0.427 0.021 0.113 0.241 0.020 0.700
n 20 20 20 20 20 20 20 20 20 20 20 20 20
ΔResting HR (bpm) Correlation Coefficient (𝝆) 0.082 0.010 0.174 −0.025 −0.070 0.085 −0.075 −0.059 −0.065 −0.138 0.012 −0.023 −0.078
Sign. (2-sided) 0.747 0.967 0.490 0.922 0.781 0.738 0.769 0.816 0.797 0.586 0.961 0.928 0.759
n 18 18 18 18 18 18 18 18 18 18 18 18 18

Additional information

Ethical statements

  • The MultiPill-Exercise pilot study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Medical Faculty, University of Tübingen (Ref. No. 298/2019BO2, 04.05.2019).
  • Written informed consent was obtained from all subjects included in the study.
  • The authors declared that no experiments on animals were performed for the present study.
  • The authors declared that no commercially available immortalized human and animal cell lines were used in the present study.

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.

Artificial Intelligence (AI) use

The authors accept full responsibility for the content of the manuscript, including the disclosure of any use of AI. No AI tools were used in the preparation of this manuscript.

Funding

This research was funded by University Medicine Tübingen AKF grant 445-0-0 (to IK and BM) and a grant from the Dr. Karl Kuhn-Stiftung (to BM). RB was supported by the “Promotionskolleg Medizin” of University Medicine Tübingen. The lifestyle intervention study “MultiPill-Exercise Pilot Study” was funded by the AOK Baden-Württemberg.

Author contributions

Conceptualization: IK, AMN and BM; methodology: RB, AS, AF, SS, MW, AMN, and BM; software: RB, AS, and BM; validation: RB, AS, AF, and BM; formal analysis: RB, AS, AF, and BM; investigation: RB, AS, AF, and MW; resources: SS, MW, AMN, IK, and BM; data curation: RB, AF, and AS; writing—original draft preparation: RB and BM; writing—review and editing: RB, AS, SS, AMN, IK, and BM; visualization: RB; supervision: AMN, IK, and BM; project administration: AMN, IK, and BM; funding acquisition: AMN, IK, and BM. All authors have read and agreed to the published version of the manuscript.

Author ORCIDs

Simone Schweda https://orcid.org/0000-0001-5827-2664

Inga Krauß https://orcid.org/0000-0001-6818-6276

Barbara Munz https://orcid.org/0000-0002-4582-6119

Data availability

All data are available from the authors upon reasonable request.

login to comment