The Rise of Elite Short-Course Triathlon Re-Emphasises the Necessity to Transition Efficiently from Cycling to Running

Transitioning e ciently between cycling and running is considered an indication of overall performance, and as a result the cycle–run (C–R) transition is one of the most researched areas of triathlon. Previous studies have thoroughly investigated the impact of prior cycling on running performance. However, with the increasing number of short-course events and the inclusion of the mixed relay at the 2020 Tokyo Olympics, e ciently transitioning from cycle–run has been re-emphasised and with it, any potential limitations to running performance among elite triathletes. This short communication provides coaches and sports scientists a review of the literature detailing the negative e ects of prior variable-cycling on running performance experienced among elite, short-course and Olympic distance triathletes; as well as discussing practical methods to minimise any negative impact of cycling on running performance. The current literature suggests that variable-cycling negatively e ects running ability in at least some elite triathletes and that improving swimming performance, drafting during cycling and C–R training at race intensity could improve an athlete's triathlon running performance. It is recommended that future research clearly de ne the performance level, competitive format of the experimental population and use protocols that are speci c to the experimental population in order to improve the training and practical application of the research ndings. Keywords:cycle–run; transition; triathlon; performance; elite; training 1. Introduction Triathlon comprises several di erent racing formats (Table.) that can be generally categorised as short-course (super sprint/sprint); Olympic distance (short-distance/standard); or long-course (70.3/Ironman). Each category of triathlon places substantially di erent physical demands on the athletes [1], such as short-course triathlon (super-sprint/sprint distance) involving producing repetitive, high-intensity e orts due to the technical courses [2] changing the physiological demands of this type of triathlon [3], compared to the consistent, steady-state paced e orts required during long-course triathlon. However, all formats

or long-course (70.3/Ironman). Each category of triathlon places substantially di erent physical demands on the athletes [1], such as short-course triathlon (super-sprint/sprint distance) involving producing repetitive, high-intensity e orts due to the technical courses [2] changing the physiological demands of this type of triathlon [3], compared to the consistent, steady-state paced e orts required during long-course triathlon. However, all formats of triathlon require an athlete to transition from cycling-to-running. Subjective descriptions of perceived incoordination are commonly reported among triathletes of all levels during the cycle–run (C–R) transition [4], leading to a potential competitive advantage to athletes that can minimise the presence of impaired movement coordination during the C–R transition. Indeed, successful performance in triathlon is considered to be largely dependent on the ability of an athlete to overcome the speci c physiological [5], neuromuscular [6] and biomechanical [7] complications associated with transitioning from cycling to running. Furthermore, recent evidence also suggests that speci c e ects of the C–R exist between elite male and female triathletes [8]. As a result of the recent increase in the number of short-course events held throughout the 2018/19 International Triathlon Sports2019,7, 99; doi:10.3390 /sports7050099 /journal/sports

Sports2019,7, 99 2 of 18 Union (ITU) World Triathlon Series (WTS), as well as the advent of the Super League Triathlon series (variable short-course distances), and the inclusion of the mixed relay event at the 2020 Tokyo Olympics, the relative importance of e ciently transitioning from C–R during triathlon, particularly during short-course formats, is re-emphasised and with it, any potential limitations to running performance. Table 1.Commonly raced formats of triathlon. Event Swim Bike Run Course Structure Event Characteristics Super sprint y 400 m (0.25 mi) 10 km (6.2 mi) 2.5 km (1.6 mi) Short circuit racing, highly technical Repetitive, high-intensity accelerations, high power/speed, technical courses, highly tactical, drafting/non-drafting, emphasis on C–R transition. Sprint y 750 m (0.47 mi) 20 km (12 mi) 5 km (3.1 mi) Circuit racing, criterium-style bike leg, relatively technical Repetitive, high-intensity accelerations, high power/speed, technical courses, highly tactical, drafting/non-drafting, emphasis on C–R transition. Olympic * 1.5 km (0.93 mi) 40 km (25 mi) 10 km (6.2 mi) Often circuit racing, draft/non drafting bike leg, some technical aspects Repetitive, high-intensity accelerations, high power/speed, technical courses, highly tactical, drafting/non-drafting, emphasis on C–R transition energy conservation/minimising physical e ort. 70.3 § 1.9 km (1.2 mi) 90 km (56 mi) 21.1 km (12 mi) Long course, non-drafting bike leg, out-and-back courses, non-technical Prolonged, submaximal steady-state e orts, management of energy consumption and e ort, non-drafting cycle leg, non-technical course. Ironman § 3.9 km (2.4 mi) 180 km (112 mi) 42.2 km (26.2 mi) Long course, non-drafting bike leg, out-and-back courses, non-technical Prolonged, submaximal steady-state e orts, management of energy consumption and e ort, non-drafting cycle leg, non-technical course. Mixed relay * ,y 300 m (0.19 mi) 8 km (5.0 mi) 2 km (1.2 mi) Short circuit racing, highly technical, similar to super sprint events Repetitive, high-intensity accelerations, high power/speed, technical courses, highly tactical, drafting/non-drafting, emphasis on C–R transition. * denotes Tokyo 2020 event; y denotes short-course event; § denotes long-course event. 2. The In uence of Cycling on Running Performance in Elite Triathletes 2.1. The Disparity Between Cycle–Run Testing Protocols and Race Demands Previous research has indicated that among highly-trained

to super sprint events Repetitive, high-intensity accelerations, high power/speed, technical courses, highly tactical, drafting/non-drafting, emphasis on C–R transition. * denotes Tokyo 2020 event; y denotes short-course event; § denotes long-course event. 2. The In uence of Cycling on Running Performance in Elite Triathletes 2.1. The Disparity Between Cycle–Run Testing Protocols and Race Demands Previous research has indicated that among highly-trained triathletes competing in short-course and Olympic-distance triathlon, any negative impact of prior cycling on running performance is minimal, compared to e ects experienced by lesser trained, recreational triathletes [1,9–11]. However, a majority of research investigating the C–R use constant/steady state or incremental cycling protocols, prior to running. As a result, these ndings may lack practical and training speci city for triathletes competing in draft-legal short-course and Olympic distance triathlon where cycling is highly variable with respect to both power output and cadence ranges [3] that would therefore, have a substantially di erent impact on running performance [2]. Subsequent studies have aimed to replicate the metabolic demand of cycling experienced during short-course triathlon, by prescribing constant cycling intensities based on a percentage of maximal aerobic power (~72% MAP) [12] or above the ventilatory threshold (~80% VO2max) [13]. Exercising at such intensities may re ect the average metabolic cost of the cycle leg of short-course and Olympic distance triathlon however, considering the variable nature of cycling during these formats of triathlon such testing protocols lack speci city, at least concerning elite draft-legal short-course and Olympic distance triathlon. As a result, it could be argued that

Sports2019,7, 99 3 of 18 non-speci c testing protocols contribute to the lack of clarity regarding the understanding of the e ects of prior cycling on running performance speci c to elite short-course and Olympic distance triathlon. Alternatively, others [11] have used a run-cycle–run protocol to determine the di erences between elite short-course versus elite long-course [12] and elite junior (male and female) versus elite senior triathletes [10,14]. While the ndings of these experiments provide valuable information relating to identifying the physiological characteristics of select cohorts of elite triathletes, potentially aiding talent identi cation and monitoring training progression, they do not speci cally outline potential changes among elite short-course and Olympic distance triathletes transitioning from C–R; that is a primary aim of this short communication. Therefore, discussions of past research in Section in Table and used a variable-cycling protocol [15] (i.e., variable power output and/or cadence) prior to running or a protocol that re ected the workload of a short-course triathlon [11] (Table). 2.2. The E ects of Variable-Cadence Cycling Protocols on Running Performance in Elite Triathletes In order to speci cally understand the impact of prior cycling on running performance previous researchers developed a protocol to determine the e ects of cycling on the movement pattern of subsequent running (Table) [ 15]. These authors designed a moderate-intensity protocol aimed at minimising the impact of fatigue. In particular, they identi ed typical cadence ranges from data collected from elite triathletes competing at an international level to create their cycling protocol. Using this variable-cadence protocol (i.e., individually preferred cadence, 55–60, 75–80, and 95–100 rpm), on-going studies involving elite triathletes aimed to identify changes to the neuromuscular control [16], muscle recruitment patterns [17,18], kinematics [17,19], biomechanics [11] and economy [6,20] of subsequent running performance (Table). Alternatively, others [21] have used moderate-intensity (variable-cadence) and high-intensity (power pro le test) protocols to quantify changes to the neuromuscular control between an isolated/control run (IR) and C–R using electromyography (EMG) and joint angle waveforms (kinematics) sampled from the muscles of the lower limb (i.e., quadriceps, hamstrings and gastrocnemius and tibialis anterior) in elite triathletes (Table). Using

running performance (Table). Alternatively, others [21] have used moderate-intensity (variable-cadence) and high-intensity (power pro le test) protocols to quantify changes to the neuromuscular control between an isolated/control run (IR) and C–R using electromyography (EMG) and joint angle waveforms (kinematics) sampled from the muscles of the lower limb (i.e., quadriceps, hamstrings and gastrocnemius and tibialis anterior) in elite triathletes (Table). Using the high-intensity, power pro le test [ 21], these authors reported that overall, cycling minimally a ects the stride-to-stride reproducibility of joint angles (<1.9 ) or muscle recruitment patterns (<5.1%) during the early phase of subsequent running. Although, one participant did demonstrate altered muscle recruitment of biceps femoris following moderate-intensity, variable-cadence cycling, suggesting that adaptation to the C–R may be individual-speci c. These results are in agreement with similar ndings [6,17,19] indicating that most elite triathletes are able to e ectively replicate pre-cycling running patterns when transitioning from moderate-intensity (variable-cadence) cycling. Moreover, using the same moderate-intensity, variable-cadence cycling protocol, no signi cant change in average muscle recruitment patterns or kinematics during the C–R, compared with an IR are evident [17,18]. However, among some elite triathletes (5/14 participants), muscle recruitment patterns recorded from the tibialis anterior muscle during the C–R closely re ect those recorded during cycling, and not the IR [17]. Alterations seen in some elite triathletes could be re ective of previous evidence that associates elite triathletes with a history of exercise-related leg pain (ERLP—5/10) with substantial ( 10%) alterations in EMG muscle patterns during C–R [19]. These authors suggested that those elite triathletes with a history of ERLP are ~2.4 more likely to have di culty replicating neuromuscular control during running after variable-cadence cycling. Similarly, increased variability in muscle recruitment patterns recorded in the lower limb (coe cient of variation range=18–37% has been observed during the early phase of C–R, i.e., 0–180 s), compared to IR [18]. These ndings may suggest the presence of neuromuscular interference or the crossover of a generalised movement pattern that a ects the control of running due to the prior, repetitive performance of the cycling movement pattern [15,16]. Furthermore, the changes in muscle

of variation range=18–37% has been observed during the early phase of C–R, i.e., 0–180 s), compared to IR [18]. These ndings may suggest the presence of neuromuscular interference or the crossover of a generalised movement pattern that a ects the control of running due to the prior, repetitive performance of the cycling movement pattern [15,16]. Furthermore, the changes in muscle recruitment patterns observed among some elite triathletes during the C–R, compared to an IR, have been associated with a reduction in running economy (3.7 0.9%) [6]. Similarly, signi cant

Sports2019,7, 99 4 of 18 increases to the cost of running, respiratory exchange ratio, breathing frequency and heart rate have been reported among elite triathletes during the early phase of C–R, compared to an IR using the same moderate-intensity, variable-cadence cycling protocol (Table) [ 20]. These authors also reported an overall decrease in mean stride length (IR=2.64 0.18 v. C–R=2.53 0.17 m) and increase in mean stride frequency at the mean response time, ~63% of time to steady-state (IR=85.7 2.7 v. C–R=87.0 2.6 strides min 1 ) and at 180 s (IR=85.4 2.6 v. C–R=87.3 2.7 strides min 1 ) when running at the same self-selected velocity (13.5 0.9 km h 1 ) during the IR and C–R condition. Furthermore, changes in running mechanics suggest that leg sti ness increases in elite triathletes during running after a bout of fatiguing cycling [11]. Increased leg sti ness during C–R may indicate superior elastic energy storage and improved e ciency of repetitive stretch-shortening cycle movements during C–R [11]. Such biomechanical changes observed in elite triathletes, coupled with decreased stride length and increased stride rate, may act to counter-balance the potential negative changes to other physiological and neuromuscular variables in order to meet the demands of C–R. 2.3. The E ects of Variable-Power Cycling Protocols on Running Performance The aforementioned research provides evidence to suggest that even at moderate-intensity variable-cadence cycling can have a negative e ect on subsequent running performance in at least some elite triathletes. To date, no research has investigated the e ects of variable-power cycling, at any intensity, on subsequent running in elite triathletes. However, results of a previous study suggest that variable-power cycling (i.e., 10–90 s intermittent e orts between 40–140% maximal aerobic power) had a greater negative impact on C–R performance, compared to constant-power cycling (i.e., 65% maximal aerobic power) among well-trained triathletes [22]. These authors reported signi cantly higher levels of blood lactate at the start of the C–R after variable-power cycling (64 61%) compared to that after constant power cycling. The elevated blood lactate levels re ected a reduced running velocity at lactate threshold (~4

impact on C–R performance, compared to constant-power cycling (i.e., 65% maximal aerobic power) among well-trained triathletes [22]. These authors reported signi cantly higher levels of blood lactate at the start of the C–R after variable-power cycling (64 61%) compared to that after constant power cycling. The elevated blood lactate levels re ected a reduced running velocity at lactate threshold (~4 mM) of 0.6 0.9 km h 1 during the C–R following variable-power cycling. Furthermore, increased central and peripheral fatigue of knee extensor muscles after variable-power, compared to constant power cycling, have also been reported in well-trained triathletes [23]. The participants in this study did not complete any subsequent running after cycling however, the increased neuromuscular fatigue that re ected a reduced strength output of the knee extensors of 12.8 6.1% that would likely contribute to a decrement in running performance. Overall, variable-cycling does not appear to heavily a ect muscle recruitment patterns and joint kinematics in most elite triathletes competing in draft-legal short-course and Olympic distance triathlon, during the C–R period. Despite this, it is well acknowledged that the physiological cost of running during triathlon is substantially greater compared to IR [24], particularly during the early phase of C–R (0–180 s) [18]. Moreover, among those elite triathletes whose running performance is impacted by prior cycling, the e ects are likely to have a substantially negative in uence on overall performance; while increased variability of muscle recruitment patterns and altered stride mechanics during the early phase of C–R is evident among some elite triathletes and is more likely to be present in athletes with a history of ERLP.

Sports2019,7, 99 5 of 18 Table 2. Physiological, neuromuscular and biomechanical e ects of prior cycling on running performance speci c to elite short-course and Olympic distance triathletes. Participants Protocol E ects Conclusions Reference 8 elite triathletes (1 male) —international level (top 50 world ranking) Run–cycle–run - 7-min run at sprint distance race-pace (18 and 15.1 0.6 km h 1 ) - maximal incremental cycle (70 W increments/3-min from 70–280 W, 35 W increments/2-min to volitional exhaustion) - 7-min run at sprint distance race-pace (18 and 15.1 0.6 km h 1 ) -"[La-] between 1st and 2nd 7-min run - mean 3.7%#C Rduring 2nd 7-min run v. 1st 7-min run - mean 4.2%#DH STRIKE during 2nd 7-min run v. 1st 7-min run. - Small mean (4.3%)DC Mduring 2nd 7-min run v. 1st 7-min run. - Cost of running is not signi cantly a ected by a fatiguing bout of cycling in elite triathletes, despite changes in [La-] between 7-min run bouts. - reduced mechanical changes during the 2nd 7-min run suggest that leg sti ness is better preserved in elite triathletes. Millet, Millet, Hofmann and Candau (2000) 8 elite triathletes (1 male) —international level (top 50 world ranking) see Millet et al. (2000) - No signi cantDthe mechanical or kinetic cost of running pre- and post-fatiguing cycling - A prior bout of high- intensity, fatiguing cycling does not a ect the subsequent running mechanics in elite triathletes. Millet, Millet and Candau (2001) 16 elite triathletes —national/international level see Chapman et al. (2009) - 10-min CR - 20-min variable-cadence cycling followed by a 30-min TR - NoDTA EMG patterns CR v TR - NoDSL, SD or kinematic joint angles CR v TR - 5/14 did show#EMG amplitude of TA during TR - Short periods of variable-cadence, moderate-intensity cycling does not a ect running kinematics or SL among elite triathletes. - However, cycling may in uence muscle activation patterns during TR in some elite triathletes. Chapman, Vicenzino, Blanch, Dowlan and Hodges (2008) 34 elite/highly-trained triathletes —national/international level —World championship quali ed —Olympic distance specialisation see Chapman et al. (2009) - 10-min CR

- Short periods of variable-cadence, moderate-intensity cycling does not a ect running kinematics or SL among elite triathletes. - However, cycling may in uence muscle activation patterns during TR in some elite triathletes. Chapman, Vicenzino, Blanch, Dowlan and Hodges (2008) 34 elite/highly-trained triathletes —national/international level —World championship quali ed —Olympic distance specialisation see Chapman et al. (2009) - 10-min CR - 20-min variable-cadence cycling followed by a 30-min TR - NoDjoint kinematics or EMG muscle patterns in most triathletes (70%) - 30% of triathletes showedDEMG patterns during TR -DEMG muscle patterns associated with 3.7 0.9%#R E("VO 2) Prior variable-cadence cycling impairs neuromuscular control on some elite triathletes that are associated with reduced TR economy Chapman, Vicenzino, Hodges, Dowlan, Hahn, Alexander and Milner (2009) 34 elite/highly-trained triathletes —national/international level —World championship experience —Olympic distance specialisation see Chapman et al. (2009) - 10-min CR - 20-min variable-cadence cycling followed by a 30-min TR - NoDjoint kinematics - EMG patterns di ered by 10% between CR and TR in 5/24 control triathletes and 5/10 triathletes with a history of ERLP Potential association between ERLP and neuromuscular control during TR in elite triathletes with a history of ERLP Chapman, Hodges, Briggs, Stapley and Vicenzino (2010)

Sports2019,7, 99 6 of 18 Table 2.Cont. Participants Protocol E ects Conclusions Reference 7 elite triathletes (3 female) —international level —national representatives at world level Low-intensity - see Chapman et al. (2009) High-intensity cycling - see Quod et al. (2010) - NoDR Eor neuromuscular control of the left leg during TR following low and high intensity cycling. - NoDlower limb kinematics - NoDEMG patterns following high-intensity cycling - 1/7 triathletes showed altered EMG patterns during TR following low-intensity cycling - Low and high intensity variable cycling does not adversely impact TR neuromuscular control of R Ein elite triathletes Bonacci, Saunders, Alexander, Blanch and Vicenzino (2011) 6 triathletes —National/international level —ITU Olympic distance race experience see Chapman et al. (2009) - No meanDlower limb EMG muscle activity patterns between CR and TR -"variability of EMG activity during TR - lower limb EMG activity patterns are not substantially in uenced by variable-cadence cycling in elite triathletes Walsh, Stamenkovic, Lepers, Peoples and Stapley (2015) 8 triathletes —National/international level —ITU Olympic distance race experience see Chapman et al. (2009) -"C R, RER and HR at MRT and 10th minute of TR v CR -#SL,"SR during TR v CR - moderate-intensity variable-cadence cycling signi cantly a ects physiological and stride pattern variables during TR, compared to CR. Walsh, Dawber, Lepers, Brown and Stapley (2017)

This article reviews the impact of cycling on running performance in elite triathletes.