# A Biophysical Analysis on the Arm Stroke Efficiency in Front Crawl Swimming: Comparing Methods and Determining the Main Performance Predictors

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

^{2}= 0.98) and physiological (R

^{2}= 0.98). Conclusions: our results suggest that the speed-based method provides the closest values to the “actual ${\eta}_{F}$” and confirm that swimming performance depends on the balance of biomechanical and bioenergetic parameters

## 1. Introduction

## 2. Materials and Methods

#### 2.1. Participants

#### 2.2. Experimental Procedures

#### 2.3. Physiological Assessments

^{−1}·min

^{−1}) were used for analysis and calculations.

^{−}]) analysis were collected from the earlobe at rest, at the end of each step and in the recovery periods (after 1, 3, and 5 min) and analyzed using a portable lactate analyzer (Lactate Pro 2, Arkay, Inc., Kyoto, Japan). The net [La

^{−}], in mmol·L

^{−1}, was then transformed into $\dot{V}{O}_{2}$ equivalents using a 2.7 mL·kg

^{−1}·mmol

^{−1}constant [14,15]:

^{−1}·min

^{−1}) consumed over the duration of each step if the anaerobically produced energy had instead been produced via aerobic pathways and ${t}_{step}$ is the step duration (min).

_{2}($\alpha $), as previously described [16,17]:

^{−1}), ${\dot{E}}_{tot}$ was converted to kJ·s

^{−1}and divided by the swimming speed, as follows:

#### 2.4. Biomechanical Assessments in Free-Swimming

#### 2.5. Biomechanical Assessments on the MAD System

#### 2.6. Speed-Based Efficiency

#### 2.7. Paddle-Wheel Efficiency

#### 2.8. Power-Based Efficiency

#### 2.9. Statistical Analysis

^{2}to describe the proportion of the total variance made up by the variance of the means. The ratio of variance explained of the sample was calculated for each effect and parameter estimate. Interpretation of η

^{2}indicates small (η

^{2}≥ 0.02), medium (η

^{2}≥ 0.13), or large effect sizes (η

^{2}≥ 0.26) for a two-way ANOVA and small (η

^{2}≥ 0.01), medium (η

^{2}≥ 0.06), or large effect sizes (η

^{2}≥ 0.14) for a one-way ANOVA according to the general rules of thumb on magnitudes of effect sizes [18]. In addition, Bland–Altman plots [19] were used to establish an agreement between the ${\eta}_{F}$ estimated from the different methods.

## 3. Results

^{2}= 1; p < 0.001), as indicated in the linear regression equations of each Bland–Altman plot.

^{−}]

_{net}, and $C$ increase with speed (p < 0.001). Moreover, swimming on the MAD System promoted a reduction in $\dot{V}{O}_{2}$ (p < 0.001), [La

^{−}]

_{net}(p = 0.001), and $C$ (p < 0.001) for equivalent speeds. The interaction between swimming speed and swimming condition allowed the individual comparisons between each step and each condition for the $\dot{V}{O}_{2}$ (p = 0.006), [La

^{−}]

_{net}(p < 0.001) and $C$ (p < 0.001). The mean (±SD) values of the metabolic parameters, as well as the individual differences between the free-swimming and MAD System conditions, are presented in Figure 8. Values of $\dot{V}{O}_{2}$ ranged from 31.5 ± 7.4 to 44.9 ± 7.2 mL·kg

^{−1}·min

^{−1}in free-swimming and from 27.4 ± 5.8 to 36.8 ± 5.0 mL·kg

^{−1}·min

^{−1}in the MAD System condition; [La

^{−1}]

_{net}ranged from 0.7 ± 0.5 to 4.9 ± 2.7 mmol·L

^{−1}in free-swimming and from 0.4 ± 0.5 to 1.6 ± 0.6 mmol·L

^{−1}in the MAD System condition; and $C$ ranged from 0.65 ± 0.18 to 0.85 ± 0.20 kj·m

^{−1}in free-swimming and from 0.55 ± 0.13 to 0.64 ± 0.11 kj·m

^{−1}in the MAD System condition.

^{2}= 0.98; p < 0.001) and physiological (R

^{2}= 0.98; p < 0.001) models could significantly predict the variances in ${v}_{200}$ and are presented in Equations (23) and (24):

## 4. Discussion

#### 4.1. Arm Stroke Efficiency in the MAD System and Free-Swimming Conditions

#### 4.2. Biophysical Adaptations to Enhance Efficiency

#### 4.3. Biophysical Predictors of Maximal Swimming Speed

## 5. Conclusions

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## References

- Toussaint, H.M.; Janssen, T.; Kluft, M. The influence of paddles on propulsion. Swim. Tech.
**1989**, 26, 28–32. [Google Scholar] - Zamparo, P.; Turri, E.; Peterson Silveira, R.; Poli, A. The interplay between arms-only propelling efficiency, power output and speed in mater swimmers. Eur. J. Appl. Physiol.
**2014**, 114, 1259–1268. [Google Scholar] [CrossRef] [PubMed] - Toussaint, H.M.; Beelen, A.; Rodenburg, A.; Sargeant, A.J.; De Groot, G.; Hollander, A.P.; Van Ingen Schenau, G.J. Propelling efficiency on front-crawl swimming. J. Appl. Physiol.
**1988**, 65, 2506–2512. [Google Scholar] [CrossRef] [PubMed] - Martin, R.B.; Yeater, R.A.; White, M.K. A simple analytical model for the crawl stroke. J. Biomech.
**1981**, 14, 539–548. [Google Scholar] [CrossRef] - Zamparo, P.; Pendergast, D.R.; Mollendorf, J.; Termin, A.; Minetti, A.E. An energy balance of front crawl. Eur. J. Appl. Physiol.
**2005**, 94, 134–144. [Google Scholar] [CrossRef] - Toussaint, H.M.; Carol, A.; Kranenborg, H.; Truijens, M.J. Effect of fatigue on stroking characteristics in an arms-only 100-m front-crawl race. Med. Sci. Sports Exerc.
**2006**, 38, 1635–1642. [Google Scholar] [CrossRef] - Figueiredo, P.; Zamparo, P.; Sousa, A.; Vilas-Boas, J.P.; Fernandes, R.J. An energy balance of the 200 m front crawl race. Eur. J. Appl. Physiol.
**2011**, 111, 767–777. [Google Scholar] [CrossRef] - Hollander, A.P.; Toussaint, H.M.; De Groot, G.; Van Ingen Schenau, G.J. Active drag and swimming performance. NZJ Sports Med.
**1985**, 13, 110–113. [Google Scholar] - Seifert, L.; Schnitzler, C.; Bideault, G.; Alberty, M.; Chollet, D.; Toussaint, H.M. Relationships between coordination, active drag and propelling efficiency in crawl. Hum. Mov. Sci.
**2015**, 39, 55–64. [Google Scholar] [CrossRef] - Toussaint, H.M. Differences in propelling efficiency between competitive and triathlon swimmers. Med. Sci. Sports Exerc.
**1990**, 22, 409–415. [Google Scholar] [CrossRef] - Toussaint, H.M.; Knops, W.; de Groot, G.; Hollander, A.P. The mechanical efficiency of front crawl swimming. Med. Sci. Sport Exerc.
**1990**, 22, 402–408. [Google Scholar] [CrossRef] - Toussaint, H.M.; Vervoorn, K. Effects of specific high resistance training in the water on competitive swimmers. Int. J. Sports Med.
**1990**, 11, 228–233. [Google Scholar] [CrossRef] [PubMed] - Ribeiro, J.; Toubekis, A.G.; Figueiredo, P.; de Jesus, K.; Toussaint, H.M.; Alves, F.; Vilas-Boas, J.P.; Fernandes, R.J. Biophysical Determinants of Front Crawl Swimming at Moderate and Severe Intensities. Int. J. Sports Physiol. Perform.
**2017**, 12, 241–246. [Google Scholar] [CrossRef] [PubMed] - Di Prampero, P.E.; Pendergast, D.R.; Wilson, D.R.; Rennie, D.W. Blood lactic acid concentrations in high velocity swimming. In Swimming Medicine IV; Eriksson, B., Furberg, B., Eds.; University Park Press: Baltimore, MD, USA, 1978; pp. 249–261. [Google Scholar]
- Thelevein, X.; Daily, D.; Persyn, U. Measurement of total energy use in the evaluation of competitive swimmers. In Current Topics in Sports Medicine; Bachl, N., Prakup, L., Suckert, R., Eds.; Urban & Schawarzenerg: Wien, Austria, 1984; pp. 668–676. [Google Scholar]
- Capelli, C.; Pendergast, D.; Termin, B. Energetics of swimming at maximal speeds in humans. Eur. J. Appl. Physiol. Occup. Physiol.
**1988**, 78, 385–393. [Google Scholar] [CrossRef] - Di Prampero, P.E. The energy cost of human locomotion on land and Water. Int. J. Sports Med.
**1986**, 7, 55–72. [Google Scholar] [CrossRef] - Cohen, J. Statistical Power Analysis for the Behavioral Sciences, 2nd ed.; Lawrence Erlbaum Associates: Hillsdale, NJ, USA, 1988. [Google Scholar]
- Bland, J.M.; Altman, D.G. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet
**1986**, 327, 307–310. [Google Scholar] [CrossRef] - Zamparo, P. Effects of age and gender on the propelling efficiency of the arm stroke. Eur. J. Appl. Physiol.
**2006**, 94, 134–144. [Google Scholar] [CrossRef] - Zamparo, P.; Lazzer, S.; Antoniazzi, C.; Cedolin, S.; Avon, R.; Lesa, C. The interplay between propelling efficiency, hydrodynamic position and energy cost of front crawl in 8 to 19-year-old swimmers. Eur. J. Appl. Physiol.
**2008**, 104, 689–699. [Google Scholar] [CrossRef] - Peterson Silveira, R.; Castro, F.A.S.; Figueiredo, P.; Vilas-Boas, J.P.; Zamparo, P. The effects of leg kick on the swimming speed and on arm stroke efficiency in front crawl. Int. J. Sports Physiol. Perform.
**2017**, 12, 728–735. [Google Scholar] [CrossRef] - Chollet, D.; Chalies, S.; Chatard, J.C. A new index of coordination for the crawl: Description and usefulness. Int. J. Sports Med.
**2000**, 21, 54–59. [Google Scholar] [CrossRef] - Gatta, G.; Cortesi, M.; Swaine, I.; Zamparo, P. Mechanical power, thrust power and propelling efficiency: Relationships with elite sprint swimming performance. J. Sports Sci.
**2018**, 36, 506–512. [Google Scholar] [CrossRef] [PubMed] - Zamparo, P.; Swaine, I.L. Mechanical and propelling efficiency in swimming derived from exercise using a laboratory-based whole-body swimming ergometer. J. Appl. Physiol.
**2012**, 113, 584–594. [Google Scholar] [CrossRef] [PubMed] - Gastin, P.B. Energy system interaction and relative contribution during maximal exercise. Sports Med.
**2001**, 31, 725–741. [Google Scholar] [CrossRef] [PubMed] - Alexander, R.M. Swimming. In Mechanics and Energetics of Animal Locomotion; Alexander, R.M., Goldspink, G., Eds.; Chapman and Hall: London, UK, 1977; pp. 222–248. [Google Scholar]
- Daniel, T.L. Efficiency in aquatic locomotion: Limitations from single cells to animals. In Efficiency and Economy in Animal Physiology; Blake, R.W., Ed.; Cambridge University Press: Cambridge, UK, 1991; pp. 83–96. [Google Scholar]
- Zamparo, P.; Gatta, G.; Pendergast, D.; Capelli, C. Active and passive drag: The role of trunk incline. Eur. J. Appl. Physiol.
**2009**, 106, 195–205. [Google Scholar] [CrossRef] - Zamparo, P.; Capelli, C.; Pendergast, D. Energetics of swimming: A historical perspective. Eur. J. Appl. Physiol.
**2010**, 111, 367–378. [Google Scholar] [CrossRef] - Pendergast, D.; Zamparo, P.; di Prampero, P.E.; Capelli, C.; Cerretelli, P.; Termin, A.; Craig, A., Jr.; Bushnell, D.; Paschke, D.; Mollendorf, J. Energy balance of human locomotion in water. Eur. J. Appl. Physiol.
**2003**, 90, 377–386. [Google Scholar] [CrossRef] - Di Prampero, P.E.; Pendergast, D.R.; Zamparo, P. Swimming economy (energy cost) and efficiency. In World Book of Swimming: From Science to Performance; Seifert, L., Chollet, D., Mujika, I., Eds.; Nova Science Publishers: New York, NY, USA, 2011; pp. 297–312. [Google Scholar]
- Pendergast, D.; Zamparo, P. Balance of biomechanical and physiological contributions to swimming performance. Port. J. Sport Sci.
**2011**, 11, (Suppl.3), 51–59. [Google Scholar] - Figueiredo, P.; Pendergast, D.R.; Vilas-Boas, J.P.; Fernandes, R.J. Interplay of biomechanical, energetic, coordinative, and muscular factors in a 200 m front crawl swim. BioMed Res. Int.
**2013**, 2013, 897232. [Google Scholar] [CrossRef] - Zamparo, P.; Gatta, G.; di Prampero, P.E. The determinants of performance in master swimmers: An analysis of master world records. Eur. J. Appl. Physiol.
**2012**, 12, 3511–3518. [Google Scholar] [CrossRef]

**Figure 1.**Stroke parameters assessed in the central 10 m of the swimming pool, as well as from a frontal camera recording the frontal plane of the swimmer. $v$: average swimming speed; SF: average stroke frequency; SL: average stroke length; θ: elbow angle at the end of the in-sweep phase; l: shoulder to hand distance.

**Figure 2.**The structure of the Measurement of Active Drag (MAD) System. Forces were applied on the push-off pads and assessed for each arm stroke by a force transducer.

**Figure 3.**Values of stroke efficiency assessed in the MAD System condition by different methods in a range of speeds, from 80 to 100% of ${v}_{200}$; * All methods were different for each swimming speed (p < 0.001).

**Figure 4.**Bland–Altman plots testing the agreement between the speed-based efficiency and the MAD System assumption (

**a**), paddle-wheel and MAD System assumption (

**b**), and paddle-wheel and speed-based efficiency (

**c**). SD: standard deviation.

**Figure 5.**Froude efficiency assessed by the power-based (

**a**), paddle-wheel (

**b**), and speed-based (

**c**) methods at different speeds, during the incremental protocol; *** Different from arm stroke efficiency values at 95 and 100% of ${v}_{200m}$ (p < 0.05); ** Different from arm stroke efficiency values at 85%, 95%, and 100% of ${v}_{200m}$ (p < 0.05); * Different from arm stroke efficiency values at all swimming speeds (p < 0.05).

**Figure 6.**Froude efficiency assessed by the power-based, paddle-wheel, and speed-based methods at different speeds during the incremental protocol; a. All methods are different; b. Difference between the power-based method and the paddle-wheel model; c. Difference between the speed-based method and the paddle-wheel model.

**Figure 7.**Bland–Altman plots testing the agreement between the speed-based and power-based efficiencies (

**a**), paddle-wheel and power-based efficiency (

**b**), and paddle-wheel and speed-based efficiency, and (

**c**) in the free-swimming condition.

**Figure 8.**Individual differences in oxygen uptake (

**a**), blood lactate concentration, (

**b**) and energy cost (

**c**) between free-swimming and the MAD System condition for each imposed speed (* p < 0.05).

**Table 1.**Biomechanical parameters in free-swimming and Measurement of Active Drag (MAD) System conditions. Values of ${\dot{W}}_{ext}$ and ${\dot{W}}_{k}$ in free-swimming were obtained from the speed-based method to assess the arm stroke efficiency.

Step | Swimming Speed (%${\mathit{v}}_{200}$ and m·s ^{−1})
| Active Drag (N) | Speed-Specific Drag | Stroke Frequency (Hz) | Stroke Length (m) | ${\dot{\mathit{W}}}_{\mathit{e}\mathit{x}\mathit{t}}$ (W) | ${\dot{\mathit{W}}}_{\mathit{d}}$ (W) | ${\dot{\mathit{W}}}_{\mathit{k}}$ (W) | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|

Both Conditions | Both Conditions | Both Conditions | Free-Swimming | MAD System | Free-Swimming | MAD System | Free-Swimming | MAD System | Free-Swimming | MAD System | Free-Swimming | MAD System | ||

1 | 80% | 1.09 ± 0.09 ^{b} | 43.0 ± 11.1 ^{b} | 36.6 ± 9.4 | 0.49 ± 0.04 ^{a,b} | 0.40 ± 0.03 ^{a,b} | 2.22 ± 0.23 ^{a c} | 2.70 ± 0.00 ^{a} | 72 ± 23 ^{a,b} | 47 ± 14 ^{a,b} | 47 ± 14 ^{b} | 47 ± 14 ^{b} | 25 ± 11 ^{a,b} | 0 ± 0 ^{a} |

2 | 85% | 1.15 ± 0.09 ^{b} | 47.7 ± 11.7 ^{b} | 35.9 ± 8.3 | 0.53 ± 0.04 ^{a,b} | 0.42 ± 0.04 ^{a,b} | 2.20 ± 0.17 ^{a c} | 2.70 ± 0.00 ^{a} | 85 ± 28 ^{a,b} | 55 ± 16 ^{a,b} | 55 ± 16 ^{b} | 55 ± 16 ^{b} | 30 ± 13 ^{a,b} | 0 ± 0 ^{a} |

3 | 90% | 1.22 ± 0.10 ^{b} | 52.6 ± 12.3 ^{b} | 35.4 ± 7.5 | 0.58 ± 0.06 ^{a,b} | 0.45 ± 0.04 ^{a,b} | 2.12 ± 0.15 ^{a c} | 2.70 ± 0.00 ^{a} | 104 ± 33 ^{a,b} | 65 ± 18 ^{a,b} | 65 ± 18 ^{b} | 65 ± 18 ^{b} | 39 ± 16 ^{a,b} | 0 ± 0 ^{a} |

4 | 95% | 1.29 ± 0.10 ^{b} | 57.7 ± 13.3 ^{b} | 34.8 ± 6.7 | 0.65 ± 0.07 ^{a,b} | 0.47 ± 0.04 ^{a,b} | 1.97 ± 0.14 ^{a,b} | 2.70 ± 0.00 ^{a} | 130 ± 42 ^{a,b} | 75 ± 21 ^{a,b} | 75 ± 21 ^{b} | 75 ± 21 ^{b} | 55 ± 22 ^{a,b} | 0 ± 0 ^{a} |

5 | 100% | 1.35 ± 0.10 ^{b} | 63.3 ± 14.4 ^{b} | 34.4 ± 6.2 | 0.76 ± 0.08 ^{a,b} | 0.51 ± 0.04 ^{a,b} | 1.79 ± 0.11 ^{a,b} | 2.70 ± 0.00 ^{a} | 165 ± 52 ^{a,b} | 87 ± 24 ^{a,b} | 87 ± 24 ^{b} | 87 ± 24 ^{b} | 78 ± 29 ^{a,b} | 0 ± 0 ^{a} |

^{a}Different from the other condition (p < 0.05);

^{b}Different from all steps (p < 0.05).

Step | ${\dot{\mathit{E}}}_{\mathit{a}\mathit{e}\mathit{r}}$ (W) | ${\dot{\mathit{E}}}_{\mathit{a}\mathit{n}\mathit{a}\mathit{e}\mathit{r}}$ (W) | ${\dot{\mathit{E}}}_{\mathit{t}\mathit{o}\mathit{t}}$ (W) | Aerobic Contribution (%) | Anaerobic Contribution (%) | |||||
---|---|---|---|---|---|---|---|---|---|---|

Free Swimming | MAD System | Free Swimming | MAD System | Free Swimming | MAD System | Free Swimming | MAD System | Free Swimming | MAD System | |

1 | 702 ± 222 ^{a c} | 598 ± 166 ^{a c} | 15 ± 15 ^{c} | 9 ± 10 ^{e} | 716 ± 230 ^{a c} | 600 ± 173 ^{a b} | 98 ± 1 ^{c} | 99 ± 1 ^{d} | 2 ± 1 ^{c} | 1 ± 1 ^{d} |

2 | 759 ± 241 ^{a c} | 629 ± 171 ^{a c} | 23 ± 25 ^{c} | 10 ± 10 ^{e} | 782 ± 255 ^{a c} | 648 ± 181 ^{a b} | 97 ± 2 ^{c} | 99 ± 1 ^{d} | 3 ± 2 ^{c} | 1 ± 1 ^{d} |

3 | 833 ± 260 ^{a b} | 699 ± 160 ^{a b} | 39 ± 31 ^{a b} | 16 ± 11 ^{a d} | 873 ± 283 ^{a b} | 707 ± 186 ^{a b} | 96 ± 2 ^{a b} | 98 ± 1 ^{a d} | 4 ± 2 ^{a b} | 2 ± 1 ^{d} |

4 | 948 ± 276 ^{a b} | 763 ± 186 ^{a b} | 71 ± 52 ^{a b} | 23 ± 21 ^{a d} | 1018 ± 316 ^{a b} | 784 ± 192 ^{a b} | 94 ± 3 ^{a b} | 97 ± 2 ^{a d} | 6 ± 3 ^{a b} | 3 ± 2 ^{d} |

5 | 1032 ± 270 ^{a b} | 824 ± 166 ^{a b} | 130 ± 88 ^{a b} | 40 ± 20 ^{a b} | 1162 ± 337 ^{a b} | 877 ± 193 ^{a b} | 90 ± 5 ^{a b} | 96 ± 2 ^{a b} | 10 ± 5 ^{a b} | 4 ± 2 ^{d} |

^{a}Different from the other conditions (p < 0.05);

^{b}Different from all steps (p < 0.05);

^{c}Different from steps 3, 4, and 5 (p < 0.05);

^{d}Different from step 5 (p < 0.05).

**Table 3.**Loading values of the studied parameters in the first two principal components of the set of observations.

Parameter | Loading Values | |
---|---|---|

Principal Component 1 | Principal Component 2 | |

Shoulder to hand distance | 0.77 | 0.44 |

Stroke frequency | 0.67 | −0.25 |

Stroke length | −0.13 | 0.44 |

Active drag | 0.84 * | 0.45 |

Power to overcome drag | 0.88 * | 0.37 |

Speed-specific drag | 0.41 | 0.80 ** |

Arm stroke efficiency | −0.87 ** | −0.12 |

External mechanical power | 0.92 ** | 0.33 |

Power wasted to the water | 0.94 * | 0.30 |

Aerobic metabolic power | 0.94 * | 0.07 |

Anaerobic metabolic power | 0.77 | −0.59 |

Total metabolic power | 0.96 ** | −0.09 |

Oxygen uptake | 0.88 * | 0.15 |

Blood lactate concentration | 0.70 | −0.64 |

Energy cost | 0.93 ** | −0.10 |

Aerobic contribution | −0.61 | 0.71 |

Anaerobic contribution | 0.61 | −0.71 |

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Peterson Silveira, R.; Soares, S.M.; Zacca, R.; Alves, F.B.; Fernandes, R.J.; Castro, F.A.d.S.; Vilas-Boas, J.P.
A Biophysical Analysis on the Arm Stroke Efficiency in Front Crawl Swimming: Comparing Methods and Determining the Main Performance Predictors. *Int. J. Environ. Res. Public Health* **2019**, *16*, 4715.
https://doi.org/10.3390/ijerph16234715

**AMA Style**

Peterson Silveira R, Soares SM, Zacca R, Alves FB, Fernandes RJ, Castro FAdS, Vilas-Boas JP.
A Biophysical Analysis on the Arm Stroke Efficiency in Front Crawl Swimming: Comparing Methods and Determining the Main Performance Predictors. *International Journal of Environmental Research and Public Health*. 2019; 16(23):4715.
https://doi.org/10.3390/ijerph16234715

**Chicago/Turabian Style**

Peterson Silveira, Ricardo, Susana Maria Soares, Rodrigo Zacca, Francisco B. Alves, Ricardo J. Fernandes, Flávio Antônio de Souza Castro, and João Paulo Vilas-Boas.
2019. "A Biophysical Analysis on the Arm Stroke Efficiency in Front Crawl Swimming: Comparing Methods and Determining the Main Performance Predictors" *International Journal of Environmental Research and Public Health* 16, no. 23: 4715.
https://doi.org/10.3390/ijerph16234715