Intervals, Thresholds, and Long Slow Distance: the Role of Intensity and Duration in Endurance Training Stephen Seiler1 and Espen Tønnessen2 Sportscience 13, 32-53, 2009 (sportsci.org/2009/ss.htm) |
|
Interval Training: a Long History Training
Plans and Cellular Signaling Training
Intensities of Elite Endurance Athletes Training
Volume of Elite Athletes Intensity
for Recreational Athletes Case
Studies of Training Manipulation Case
1–From Soccer Pro to Elite Cyclist Case
2–From Modern Pentathlete to Runner Valid Comparisons
of Training Interventions The evening before the start of the 2009 European College of Sport Science Congress in Oslo, the two of us were sitting at a doctoral dissertation defense dinner that is part of the time honored tradition of the “doctoral disputas” in Scandinavia. One of us was the relieved disputant (Tønnessen) who had successfully defended his dissertation. The other had played the adversarial role of “førsteopponent.” Tønnessen’s research on the talent development process included extensive empirical analyses of the training characteristics of selected world champion female endurance athletes. His career case-study series systematized training diary logs of over 15,000 training sessions from three World and/or Olympic champions in three sports: distance running, cross-country skiing, and orienteering. Common for all three champions was that over their long, successful careers, about 85 % of their training sessions were performed as continuous efforts at low to moderate intensity (blood lactate £2 mM). Among the 40 guests sat coaches, scientists, and former athletes who had been directly or indirectly involved in winning more endurance sport Olympic gold medals and world championships than we could count. One guest, Dag Kaas, had coached 12 individual world champions in four different sports. In his toast to the candidate he remarked, ”My experience as a coach tells me that to become world champion in endurance disciplines, you have to train SMART, AND you have to train a LOT. One without the other is insufficient.” So what is smart endurance training? The question is timely: research and popular interest in interval training for fitness, rehabilitation, and performance has skyrocketed in recent years on the back of new research studies and even more marketing by various players in the health and fitness industry. Some recent investigations on untrained or moderately trained subjects have suggested that 2-8 wk of 2-3 times weekly intense interval training can induce rapid and substantial metabolic and cardiovascular performance improvements (Daussin et al., 2007; Helgerud et al., 2007; Talanian et al., 2007). Some popular media articles have interpreted these findings to mean that long, steady distance sessions are a waste of time. Whether well founded or not, this interpretation raises reasonable questions about the importance and quantity of high- (and low-) intensity training in the overall training process of the endurance athlete. Our goal with this article is to discuss this issue in a way that integrates research and practice. In view of the recent hype and the explosion in the number of studies investigating interval training in various health, rehabilitation, and performance settings, one could be forgiven for assuming that this training form was some magic training pill scientists had devised comparatively recently. The reality is that athletes have been using interval training for at least 60 years. So, some discussion of interval training research is in order before we address the broader question of training intensity distribution in competitive endurance athletes. Interval Training: a Long HistoryInternational running coach Peter Thompson wrote in Athletics Weekly that clear references to “repetition training” were seen already by the early 1900s (Thompson, 2005). Nobel Prize winning physiologist AV Hill incorporated intermittent exercise into his studies of exercising humans already in the 1920s (Hill et al., 1924a; Hill et al., 1924b). About this time, Swede Gosta Holmer introduced Fartlek to distance running (fart= speed and lek= play in Swedish). The specific term interval training is attributed to German coach Waldemer Gerschler. Influenced by work physiologist Hans Reindell in the late 1930s, he was convinced that alternating periods of hard work and recovery was an effective adaptive stimulus for the heart. They apparently adopted the term because they both believed that it was the recovery interval that was vital to the training effect. Since then, the terms intermittent exercise, repetition training, and interval training have all been used to describe a broad range of training prescriptions involving alternating work and rest periods (Daniels and Scardina, 1984). In the 1960s, Swedish physiologists, led by Per Åstrand, performed groundbreaking research demonstrating how manipulation of work duration and rest duration could dramatically impact physiological responses to intermittent exercise (Åstrand et al., 1960; Åstrand I, 1960; Christensen, 1960; Christensen et al., 1960). As Daniels and Scardina (1984) concluded 25 years ago, their work laid the foundation for all interval training research to follow. In their classic chapter Physical Training in Textbook of Work Physiology, Åstrand and Rodahl (1986) wrote, “it is an important but unsolved question which type of training is most effective: to maintain a level representing 90 % of the maximal oxygen uptake for 40 min, or to tax 100 % of the oxygen uptake capacity for about 16 min.” (The same chapter from the 4th edition, published in 2003, can be read here.) This quote serves as an appropriate background for defining high intensity aerobic interval training (HIT) as we will use it in this article: repeated bouts of exercise lasting ~1 to 8 min and eliciting an oxygen demand equal to ~90 to 100 % of VO2max, separated by rest periods of 1 to 5 min (Seiler and Sjursen, 2004; Seiler and Hetlelid, 2005). Controlled studies comparing the physiological and performance impact of continuous training (CT) below the lactate turnpoint (typically 60-75 % of VO2max for 30 min or more) and HIT began to emerge in the 1970s. Sample sizes were small and the results were mixed, with superior results for HIT (Henriksson and Reitman, 1976; Wenger and Macnab, 1975), superior results for CT (Saltin et al., 1976), and little difference (Cunningham et al., 1979; Eddy et al., 1977; Gregory, 1979). Like most published studies comparing the two types of training, the CT and HIT interventions compared in these studies were matched for total work (iso-energetic). In the context of how athletes actually train and perceive training stress, this situation is artificial, and one we will come back to later. McDougall and Sale (1981) published one of the earliest reviews comparing the effects of continuous and interval training, directed at coaches and athletes. They concluded that both forms of training were important, but for different reasons. Two physiological assumptions that are now largely disproven influenced their interpretation. First, they concluded that HIT was superior for inducing peripheral changes, because the higher work intensity induced a greater degree of skeletal muscle hypoxia. We now know that in healthy subjects, increased lactate accumulation in the blood during exercise need not be due to increased muscle hypoxia (Gladden, 2004). Second, they concluded that since stroke volume already plateaus at 40-50 %VO2max, higher exercise intensities would not enhance ventricular filling. We now know that stroke volume continues to rise at higher intensities, perhaps even to VO2max, in well trained athletes (Gledhill et al., 1994; Zhou et al., 2001). Assuming a stroke volume plateau at low exercise intensity, they concluded that the benefit of exercise on cardiac performance was derived via stimulation of high cardiac contractility, which they argued peaked at about 75 %VO2max. Thus, moderate-intensity continuous exercise over longer durations and therefore more heart beats was deemed most beneficial for enhancing cardiac performance. While newer research no longer supports their specific conclusions, they did raise the important point that there are underlying characteristics of the physiological response to HIT and CT that should help explain any differential impact on adaptive responses. Poole and Gaesser (1985) published a citation classic comparing 8 wk of 3 × weekly training of untrained subjects for either 55 min at 50 %VO2max, 35 min at 75 %VO2max, or 10 × 2 min at 105 %VO2max with 2-min recoveries. They observed no differences in the magnitude of the increase in either VO2max or power at lactate threshold among the three groups. Their findings were corroborated by Bhambini and Singh (1985) in a study of similar design published the same year. Gorostiaga et al. (1991) reported findings that challenged McDougall and Sale's conclusions regarding the adaptive specificity of interval and continuous training. They had untrained subjects exercise for 30 min, three days a week either as CT at 50 % of the lowest power eliciting VO2max, or as HIT, alternating 30 s at 100 % of power at VO2max and 30 s rest, such that total work was matched. Directly counter to McDougall and Sales conclusions, they found HIT to induce greater changes in VO2max, while CT was more effective in improving peripheral oxidative capacity and the lactate profile. At the beginning of the 1990s, the available data did not support a consensus regarding the relative efficacy of CT vs HIT in inducing peripheral or central changes related to endurance performance. Twenty years on, research continues regarding the extent to which VO2max, fractional utilization of VO2max, and work efficiency/economy are differentially impacted by CT and HIT in healthy, initially untrained individuals. Study results continue to be mixed, with some studies showing no differences in peripheral and central adaptations to CT vs HIT (Berger et al., 2006; Edge et al., 2006; Overend et al., 1992) and others greater improvements with HIT (Daussin et al., 2008a; Daussin et al., 2008b; Helgerud et al., 2007). When differences are seen, they lean in the direction that continuous work at sub-maximal intensities promotes greater peripheral adaptations and HIT promotes greater central adaptations (Helgerud et al., 2007). Controlled studies directly comparing CT and HIT in already well-trained subjects were essentially absent from the literature until recently. However, a few single-group design studies involving endurance athletes did emerge in the 1990s. Acevedo and Goldfarb (1989) reported improved 10-km performance and treadmill time to exhaustion at the same pace up a 2 % grade in well-trained runners who increased their training intensity to 90-95 %VO2max on three of their weekly training days. In these already well-trained athletes, VO2max was unchanged after 8 wk of training intensification, but a right shift in the blood lactate profile was observed. In 1996 -97, South African sport scientists published the results of a single group intervention involving competitive cyclists (Lindsay et al., 1996; Weston et al., 1997). They trained regionally competitive cyclists who were specifically selected for study based on the criteria that they had not undertaken any interval training in the 3-4 months prior to study initiation. When 15 % of their normal training volume was replaced with 2 d.wk-1 interval training for 3-4 wk (six training sessions of six 5-min high intensity work bouts), 40-km time trial performance, peak sustained power output (PPO), and time to fatigue at 150 %PPO were all modestly improved. Physiological measurements such as VO2max and lactate profile changes were not reported. Stepto and colleagues then addressed the question of interval-training optimization in a similar sample of non-interval trained, regional cyclists (Stepto et al., 1999). They compared interval bouts ranging from 80 to 175 % of peak aerobic power (30 s to 8 min duration, 6-32 min total work). Group sizes were small (n=3-4), but the one group that consistently improved endurance test performance (~3 %) had used 4-min intervals at 85 % PPO. These controlled training intensification studies essentially confirmed what athletes and coaches seemed to have known for decades: some high-intensity interval training should be integrated into the training program for optimal performance gains. These studies also seemed to trigger a surge in interest in the role of HIT in athlete performance development that has further grown in recent years. If doing
some HIT (1-2 bouts per week) gives a performance boost, is more even better?
Billat and colleagues explored this question in a group of middle distance
runners initially training six sessions per week of CT only. They found that
a training intensification to four CT sessions, one HIT session, and one
lactate threshold (LT) session resulted in improved running speed at VO2max (but not VO2max
itself) and running economy. Further intensification to two CT sessions,
three HIT sessions and one LT session each week gave no additional adaptive
benefit, but did increase subjective training stress and indicators of
impending overtraining (Billat et al., 1999). In
fact, training intensification over periods of 2-8 wk with frequent
high-intensity bouts (3-4 sessions per week) is an effective means of
temporarily compromising performance and inducing overreaching and possibly
overtraining symptoms in athletes (Halson and Jeukendrup, 2004). There is likely an appropriate balance
between high- and low-intensity training in the day-to-day intensity
distribution of the endurance athlete. These findings bring us to two related
questions: how do really good endurance athletes actually train, and is there
an optimal training intensity distribution for long-term performance
development? While arguments can be made that tradition, resistance to change and even superstition may negatively influence training methods of elite endurance athletes, sports history tells us that athletes are experimental and innovative. Observing the training methods of the world's best endurance athletes represent a more valid picture of “best practice” than we can develop from short-term laboratory studies of untrained or moderately trained subjects. In today’s performance environment, where promising athletes have essentially unlimited time to train, all athletes train a lot and are highly motivated to optimize the training process. Training ideas that sound good but don't work in practice will fade away. Given these conditions, we argue that any consistent pattern of training intensity distribution emerging across sport disciplines is likely to be a result of a successful self-organization (evolution) towards a “population optimum.” High performance training is an individualized process for sure, but by population optimum, we mean an approach to training organization that results in most athletes staying healthy, making good progress, and performing well in their most important events. Exercise Intensity ZonesTo describe intensity distribution in endurance athletes we have to first agree on an intensity scale. There are different intensity zone schemes to choose from. Most national sport governing bodies employ an intensity scale based on ranges of heart rate relative to maximum and associated typical blood lactate concentration range. Research approaches vary, but a number of recent research studies have identified intensity zones based on ventilatory thresholds. Here we will examine an example of each of these scales. Table 1 shows the intensity scale used by all endurance sports in Norway. A valid criticism of such a scale is that it does not account for individual variation in the relationship between heart rate and blood lactate, or activity specific variation, such as the tendency for maximal steady state concentrations for blood lactate to be higher in activities activating less muscle mass (Beneke and von Duvillard, 1996; Beneke et al., 2001).
Several recent studies examining training intensity distribution (Esteve-Lanao et al., 2005; Seiler and Kjerland, 2006; Zapico et al., 2007) or performance intensity distribution in multi-day events (Lucia et al., 1999; Lucia et al., 2003) have employed the first and second ventilatory turnpoints to demarcate three intensity zones (Figure 1). The 5-zone scale in the table above and the 3-zone scale below are reasonably super-imposable in that intensity Zone 3 in the 5-zone system coincides well with Zone 2 in the 3-zone model. While defining five “aerobic” intensity zones is likely to be informative in training practice, it is important to note that they are not based on clearly defined physiological markers. Note also that 2-3 additional zones are typically defined to accommodate very high intensity sprint, anaerobic capacity, and strength training. These zones are typically defined as “anaerobic” Zones 6, 7, and 8.
Training Plans and Cellular SignalingAthletes do not train at the same intensity or for the same duration every day. These variables are manipulated from day to day with the implicit goals to maximize physiological capacity over time, and stay healthy. Indeed, the former is quite dependent on the latter. Training frequency is also a critical variable manipulated by the athlete. This is particularly evident when comparing younger (often training 5-8 times per week) and more mature athletes at peak performance level (often training 10-13 sessions per week). Ramping up training frequency (as opposed to training longer durations each session) is responsible for most of the increase in yearly training hours observed as teenage athletes mature. Cycling might be an exception to this general rule, since cycling tradition dictates single daily sessions that often span 4-6 h among professionals. The ultimate targets of the training process are individual cells. Changes in rates of DNA transcription, RNA translation, and ultimately, synthesis of specific proteins or protein constellations are induced via a cascade of intracellular signals induced by the training bout. Molecular exercise biologists are unraveling how manipulation of intensity and duration of exercise specifically modifies intracellular signaling and resulting protein synthetic rates at the cellular or whole muscle/myocardial level (Ahmetov and Rogozkin, 2009; Hoppeler et al., 2007; Joseph et al., 2006; Marcuello et al., 2005; McPhee et al., 2009; Yan, 2009). About 85 % of all publications involving gene expression and exercise are less than 10 y old, so we do not yet know enough to relate results of Western blots to the specific training of an athlete. The signaling impact of a given exercise stress (intensity×duration) almost certainly decays with training (Hoppeler et al., 2007; Nordsborg et al., 2003). For example, AMP activated protein kinase α2 (AMPK) activity jumps 9-fold above resting levels after 120 min of cycling at 66 %VO2max in untrained subjects. However, after only 10 training sessions, almost no increase in AMPK is seen after the same exercise bout (McConell et al., 2005). Manipulating exercise intensity and duration also impacts the systemic stress responses associated with training. Making this connection is further complicated by recent findings suggesting that muscle glycogen depletion can enhance and antioxidant supplementation can inhibit adaptations to training (Brigelius-Flohe, 2009; Gomez-Cabrera et al., 2008; Hansen et al., 2005; Ristow et al., 2009; Yeo et al., 2008). It seems fair to conclude that while we suspect important differences exist, we are not yet able to relate specific training variables (e.g., 60 min vs 120 min at 70 %VO2max) to differences in cell signaling in a detailed way. Our view of the adaptive process remains limited to a larger scale. We can still identify some potential signaling factors that are associated with increased exercise intensity over a given duration (Table 2) or increased exercise duration at a given sub-maximal intensity (Table 3). Some of these are potentially adaptive and others maladaptive. There is likely substantial overlapping of effects between extending exercise duration and increasing exercise intensity. It may be a hard pill to swallow for some exercise physiologists, but athletes and coaches do not need to know very much exercise physiology to train effectively. They do have to be sensitive to how training manipulations impact athlete health, daily training tolerance, and performance, and to make effective adjustments. Over time, a successful athlete will presumably organize their training in a way that maximizes adaptive benefit for a given perceived stress load. That is, we can assume that highly successful athletes integrate this feedback experience over time to maximize training benefit and minimize risk of negative outcomes such as illness, injury, stagnation, or overtraining.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
