Experimental approach of the problem

This systematic review was carried out according to the criteria present in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and was registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the identification CDR42022308290. The search was performed using Web of Science, PubMed and Scopus databases through September 2022. The following string was used in the search ("HIIT" OR "interval training" OR "sprint" OR "strength training" OR "plyometric training" OR "running" OR "warm up" OR "cool down" OR "training" OR "recovery") AND ("performance" OR "running economy" OR "lactate threshold" OR "VO2max" OR "vVO2max" OR "oxygen uptake" OR " oxygen consumption" OR "time trial" OR "time-trial") AND ("5-km" OR "5km" OR "5 km" OR "5000m" OR "5 kilometers") AND (athlete OR runner) NOT (" basketball" OR "rugby" OR "soccer") - example for PubMed search.

The PICOS strategy (population, intervention, comparison, result and study design) was used to define the eligibility criteria. Inclusion criteria for this analysis were (1) athletes and well-trained individuals, adults, both male and female. To ensure this characteristic, the study sample should have a better 5,000 m running time below 22 minutes (male sex) and 25 minutes (female sex), or, when time data was not available, VO₂max values above 53 ml.kg-1.min-1 (male sex) and 40 ml.kg-1.min-1 (female sex), which are classified by ^{Herdy and Caixeta (2016)} as excellent for the Brazilian population; (2) specific running training protocols or not as intervention. Some examples are continuous running protocols, interval running, uphill running, neuromuscular training, altitude training, ischemic training, hypoxia training, and polishing and acclimatisation strategies were accepted; (3) comparisons pre vs post-intervention, including or not more than one group (groups submitted to different interventions during the same period or control groups, which were not submitted to any interventions). (4) presence of at least one of the following outcomes: 5,000 m running time, or physiological variables associated to endurance running, specifically the 5,000 m race, such as VO₂max, vVO₂max, anaerobic threshold and running economy; and (5) study design types: observational, experimental or quasi-experimental studies. Exclusion criteria were (1) a study sample with athletes of team sports like soccer, basketball, volleyball and others, (2) nutritional strategies as intervention, (3) paper language other than English or Portuguese, (4) recreational or amateur runners, and (5) child, adolescent or elderly as a sample.

Procedures

The initial list was examined to check and exclude duplicates. After that, the titles and abstracts were read, excluding those not meeting the eligibility criteria. The next step was to read the full articles and exclude those that no longer met the inclusion criteria or that addressed some exclusion criteria. In both stages, the eligibility judgment was made independently by two researchers (TB and AG), and, in case of disagreement, a third evaluator (MM) was activated. In addition to the search, others were identified in secondary sources that underwent the eligibility analysis using the same criteria.

Risk of bias assessment

The methodological quality and risk of bias were assessed based on the Cochrane Handbook for Systematic Reviews of Interventions (^{Higgins & Green, 2008}). Two researchers (AG and TB) independently performed this step, with disagreements resolved by a third one (MM). They analysed the risks of: (1) selection bias, (2) performance, (3) detection, (4) attrition, and (5) reporting.

Data extraction and outcomes

The reading of the articles was carried out in order to extract the information for analysis of the risk of bias and the main information of this review: authors, sample size, age, gender, VO₂max, 5,000 m time trial (TT), type of intervention, duration of the protocol, structuring of the intervention and results (specifically the variables that are the object of research in this work). The findings were categorised according to the type of training protocol: running, neuromuscular, taper and a category called "others".

Risk of bias

The risk of bias results showed that the included studies have, in most cases, a high risk for the types of bias "selection" and "performance", and a low risk for "attrition" and "report". It should be noted that the high risk of performance bias is probably due to the characteristics of the intervention: carrying out a certain type of physical training is an intervention that is difficult to be "blinded", especially for well-trained athletes and practitioners, who usually have good knowledge about physical training methods. The risk of bias results are presented in Table 1.

Effects of running protocols

Seventeen articles (Table 2) observed the results of 35 running protocols. Ten protocols applied High-Intensity Interval Training (HIIT), seven protocols involved both interval and continuous training, eight protocols proposed uphill HIIT, five protocols employed continuous running, two protocols used increased running training, two protocols involved cross-training interventions (combination of exercises from different modalities in a training session or increase the number of running sessions), and one protocol involved deep water running.

The analysis of the studies showed quite comprehensive results. Among the 21 protocols that evaluated the 5,000 m time trial, 16 mentioned a reduction in time (71.43%), and six (28.57%) showed no change in this measure. Of the 35 that applied running protocols, 23 evaluated running economy (RE): 13 (56.52%) protocols showed improvements in this aspect (VO₂ reduction at a given intensity), while four (17.39%) got worse and six (26.09%) showed no changes. VO₂max was evaluated in 31 protocols, showing significant improvement in 15 (48.39%) protocols but did not change in the other 16 (51.61%). The anaerobic threshold point was analysed in 21 protocols, increasing in 16 (71.43%) and not changing in eight (28.57%). Finally, the speed at VO₂max (vVO2max) was determined in 15 protocols, not changing in four (26.67%) protocols and increasing in 11 (73.33%)

Among all protocols that improved the 5,000 m time trial or any of the predicted variables studied in this review, 33.33% used both interval and continuous training, 33.33% used uphill HIIT, 13.33 used HIIT, 6.66% applied cross training, and 6.66% applied increased running training. It was possible to observe that both continuous and HIIT training (including uphill training) are the most effective strategies to improve 5kmTT or physiological markers recognised as predictors of performance in this distance.

Effects of neuromuscular protocols

Nine articles (Table 3) totalling 12 neuromuscular intervention protocols were observed in this review. Four protocols were based on strength resistance training, four protocols dealt with complex training, two with plyometric training, one with hypertrophy training and, finally, a protocol with explosive training.

Of the 12 neuromuscular intervention protocols, eleven analysed running economy, with eight (72.73%) showing improvement and three maintaining this variable. Four protocols quantified the 5,000 m trial time, all (100%) indicating reduction. Among the 10 protocols that investigated VO₂max, nine (90%) did not show significant changes, while only one protocol showed improvement. Finally, only three protocols studied the vVO₂max, with improvement found in only one protocol (33.33%).

Among all protocols that improved the 5,000 m time trial or any of the predicted variables studied in this review, 50% applied complex training, 25% used plyometric as an intervention and 25% used strength resistance training. It was possible to observe that complex training, strength resistance training and plyometric training are effective strategies to improve 5kmTT or physiological markers recognised as predictors of performance in this distance.

Effects of taper protocols

Fourteen protocols (seven articles) (Table 4) dealt with training taper. Taper can occur in the days or weeks before the competition by decreasing the distance covered, training frequency and/or intensity. All articles selected in this review reduced the distance covered in a week, 21% decreased training frequency (days a week), and 14% decreased training intensity.

The taper interventions had a considerable diversity of results. Of the 14 protocols evaluated, four analysed the running economy, with three (75%) showing improvement (reduction) and one showing no significant changes. Eight (57.14%) protocols showed reductions in the time needed to complete the 5,000 m time trial, four showed no changes in this measure, and two showed an increase in the time needed to complete this test. VO₂max increased in eight of the 10 protocols that investigated this variable (80%), while two protocols with no significant changes. Finally, the anaerobic threshold increased in all eight protocols that evaluated this variable vVO₂max was not analysed in the taper protocols.

It was possible to observe that reducing the distance covered is an effective strategy to improve 5kmTT or physiological markers recognized as predictors of performance in this distance.

Effects of "others" protocols

It was possible to verify two types of protocols that in this study were classified as "others" in order to maintain the organisation of the study (Table 5). One of the protocols investigated the effect of immersion in high-temperature water after several exercise sessions (^{Zurawlew et al., 2016}) and was observed as a result of an improvement in 5,000 m TT when exercising in high-temperature environments. The other protocol consisted of a training protocol in a hypoxic environment (^{Pedlar et al., 2008}), which had no modification for 5,000 m TT.

Running

After analysing our results, it was possible to affirm that a training plan must include continuous and interval training. The continuous training protocols involved running at the intensity of 65-92%HRmax and 100%vMLSS (maximal lactate steady state) (^{Helgerud et al., 2007}; ^{Philp et al., 2008}; ^{Enoksen et al., 2011}).

The improvements observed in running protocols based on HIIT were possibly linked to the fact that the anaerobic training presented in these HIIT sessions can influence aerobic capacity. That occurs due to the improvement of the body's capacity to remove hydrogen ions from the muscle, postponing fatigue. Additionally, HIIT improves the uptake, transport and use of oxygen in the body since high intensities, despite predominantly using anaerobic metabolism, also use aerobic metabolism and, thus, improve the individual's VO₂max (^{Wen et al., 2019}). Thus, it enables the individual to better perform aerobic activities due to the fact that the body generates physiological adaptations such as an improvement in the quality of mitochondria (^{MacInnis & Gibala, 2017}). The importance of HIIT in VO₂max gains is already well established in the literature, with more significant increases when compared to continuous training in moderate intensity (^{Neves et al., 2022}). Among HIIT protocols, those that involved running at the intensity of 95-100%vVO2max and the duration of 60%Tlim presented improvements in 5,000 m TT (^{Denadai et al., 2006}). Other protocols were able to improve the physiologic variables, using the intensity of 95-100%vVO2max or 90-95% da HRmax, using a duration of 50-75%Tlim (time to exhaustion at vVO₂max) (^{Helgerud et al., 2007}; ^{Silva et al., 2017}; ^{Smith et al., 1999}).

Bearing in mind the aspects presented about HIIT, the variation of this method on hills had interesting results in some of the studied protocols and can be of great value considering that running on hills can improve 5,000 m TT and vVO₂max (^{Barnes et al., 2013a}); and VO₂max, LT and RE (^{Barnes et al., 2013a}; ^{Ferley et al., 2016}).

Neuromuscular

The influence of neuromuscular protocols on performance predictors in the 5,000 m time trial was evident, mainly on running economy. This fact is probably due to the better synchronization of the recruited musculature, thus resulting in greater muscular strength (^{Cormie et al., 2011}) and, therefore, the effort required to maintain the running activity would be less costly to the body. With this improvement, there is an improvement in the time of the 5,000 m time trial, considering that it will allow the athlete to perform the same level of effort, but at a higher speed.

Strength training was able to improve 5,000 m TT (^{Li et al., 2019}) and RE (^{Guglielmo et al., 2009}; ^{Li et al., 2019}; ^{Støren, 2008}). These protocols were performed in 4-6RM for leg exercises, through 4 to 8 weeks. ^{Barnes et al. (2013b)} used 6RM to 20RM and did not find significant improvement. Possibly this choice led to the use of a comparatively smaller amount of weight, which may have impaired strength development and contributed to the lack of results. This comparison allows us to infer that coaches should prioritize strength training, performing up to 6 maximum repetitions to improve running performance.

Plyometric training provided significant performance improvement in the 5,000 m TT (^{Hamilton et al., 2006}) and RE (^{Saunders et al., 2006}). Its benefits are probably not related to the improvement of cardiopulmonary variables, but to the development of muscle power and better use of stored elastic energy, as stated by ^{Saunders et al. (2006)}.

The protocols that combine the amount of repetitions or weight that can be classified into more than one type of neuromuscular training are suitable for the complex training category. These protocols involved a combination of strength, plyometric and explosive exercises. Complex training provided improved performance in the 5,000 m TT (^{Li et al., 2019}; ^{Paavolainen et al., 1999}), and an increase in VO2max (^{Paavolainen et al., 1999}), in vVO2max (^{Beattie et al., 2016}) and in RE (^{Paavolainen et al., 1999}; ^{Beattie et al., 2016}). The complex training protocols studied in our review were different, but all of them used strength training in the session. Based on the results of these studies, we can infer that the use of strength training as an integral part of complex training is crucial.

Explosive training did not generate significant results in the evaluated variables (^{Guglielmo et al., 2009}), such as hypertrophy training (^{Yamanaka et al., 2021}).

Taper

The taper showed a wide variety of results due to the great diversity of applied protocols. In general, the taper that lasted from 1 to 2 weeks had a positive effect on the 5,000 m time trial, with a reduction of 15 to 65% during that time reduction period. This is related to adequate rest, providing the athlete with the best possible conditions to perform the test (^{Bompa & Haff, 2012}). Longer training reductions may show a worsening of time in the 5,000 m test, possibly due to the athlete starting to reach levels of deconditioning.

Decreasing only the distance improved the 5,000 m performance (^{Banister et al., 1999}; ^{Houmard et al., 1994}; ^{Zarkadas et al., 1995}). The decrease in intensity presented inconsistent results: worse performance for some protocols (^{McConell et al., 1993}; ^{Wittig et al., 1992}) but better performance for the ^{Bellenger et al. (2019)} study.

It is not possible to conclude on the influence of frequency, given that, as a rule, this decrease was accompanied by a decrease in intensity as well. Analysis of the duration of the taper allows us to infer that this phase should last a maximum of 14 days. The articles that exceeded this period caused worsening (^{McConell et al., 1993}; ^{Wittig et al., 1992}) or its maintenance (Houmard et al., 1999), which goes against the objective of this phase of training. We can also infer that the taper should be characterized by the reduction of the distance covered and that the intensity should not be changed.

Others

The intervention protocols in submersion in hot water after the running intervention demonstrated that for high temperatures, there was an adaptation of the organism to higher temperatures resulting in improvements in the time trial at 33°C, this may be related to the adaptation of the organism to higher temperatures allowing the individual to perform better in such situations, but when performing the test at milder temperatures (18°C) there were no significant changes. The intervention seeking to change the atmospheric pressure was ineffective for the performance in the 5,000 m time trial, this can be explained by the individual maintaining daily life at normal atmospheric pressures, making the adaptations that living in hypoxia would generate impossible.

Limitations and future prospects

The choice to use only English-language articles may have left some relevant articles out of the research. However, it is worth highlighting that the number of articles found that are not in English was not significant and would probably not alter the final results.

As for the 5,000 m race, future research should analyse the training protocols to be used by amateur and recreational runners, which were excluded from the scope of this work. Furthermore, this type of review should be done for other Olympic events, such as the 10,000m and the marathon.