POLARIZED TRAINING AND HYPOXIC MUSCLES: Highlights of the ACSM Annual Meeting
Will G Hopkins PhD
Department of Physiology, University of Otago, Dunedin
9001, New Zealand. Email.
The annual meeting of the American College of Sports Medicine is a conference with three largely independent themes: athletic performance, sports injuries, and physical fitness. I was funded by Sport Science NZ to attend this year's meeting in Seattle and to report back on the first of these themes. This article is my report, which SSNZ has kindly allowed me to publish here.
But first, a few words about the program. The ACSM annual meeting is unquestionably the top annual international conference for sport scientists and others interested in factors that affect athletic performance. This year there were so many sessions running in parallel that I missed more than half the presentations I was interested in. For example, on Saturday afternoon I had to choose between a symposium on blood lactate, a symposium on muscle fatigue, a symposium on Olympic issues, and a free-communication slide session on ergogenic aids. I left the meeting with a feeling of frustration over all that I had missed rather than a feeling of satisfaction with all I had seen. I hope to persuade the program committee to make changes to reduce this problem of too much of a good thing. Please contact them or me with any suggestions.
I summarize here only those presentations concerned with performance of competitive athletes. My apologies to those whose relevant presentations I have omitted: either I missed your presentation, or I failed to appreciate your abstract, or there was no abstract. In what follows, numbers in parentheses (e.g., #1234) refer to abstracts in the May supplement of Medicine and Science in Sports and Exercise.
A high point was Stephen Seiler's contribution to the symposium on practical aspects of lactate measurement (#2041). On the basis of his experience with elite cross-country skiers and rowers, he argued that top endurance athletes do comparatively little training at or near lactate-threshold intensity (blood lactate concentrations of ~4 mmol/L, corresponding to intensities of ~85% of maximum oxygen consumption). Instead, their training is "polarized" around this intensity, in the sense that they do a few sessions per week at intensities well above 4 mmol/L and the rest at <2 mmol/L. He described the lactate threshold as the "lactate black hole", to emphasize his idea that too much training at this intensity tends to reduce the quality of higher intensity work-outs and ultimately leads to training monotony and overtraining. Carl Foster, who chaired the symposium, then outlined running programs of the famous Jack Daniels. Most of the training in Daniels' programs is below threshold intensity and the rest is at or above, which could be considered polarized training. Axel Urhausen, one of the speakers at the symposium, commented afterwards that he and his co-workers induce an overtrained state for research purposes by prescribing daily training at the lactate threshold.
Periodization--the structuring of a training program throughout a season--was the subject of only two free communications. One of the perennial problems of periodization is whether you should do base (low-intensity) training before quality (high-intensity) training in the months leading up to a competition, or whether you should do both together. In a 12-week training study in which the event was a simulated 80-s cycling time trial, sub-elite cyclists were randomized to sequential or concurrent training programs with the same total volume. The slowest cyclists gained more from the sequential approach, whereas the fastest cyclists did better with concurrent training (#789, Reid and Sleivert). Incidentally, world-class cross-country skiers and rowers use concurrent training for 10 months of the year, according to Seiler. Another periodization problem is how best to taper the training load in the days immediately before a competition. Sub-elite cyclists training exclusively at 85% VO2max tended to get a bigger improvement in a 20-km time trial following a 7-day taper in which volume was reduced to 50% rather than 30% or 80% (#387, Neary et al.).
Explosive or ballistic training featured in two presentations. In one (#790, Olsen and Hopkins) competitive martial artists attempted such movements with the leg tethered. After 10 weeks their kicks and palm strikes were 4-19% faster than those of a control group who performed extra kicks. In the other (#1554, Nemoto et al.) elite rowers performed ballistic resistance training, first as a warm-up (in a crossover study), then for 6 months as part of training (in an uncontrolled study). The ballistic warm-up enhanced 2000-m time by 0.8% (equivalent to an increase in mean power output of 2.4%) compared with a standard warm-up. Six months of ballistic training reduced 2000-m time by 1.8%, an amazing gain for elite athletes.
Plyometrics is another form of training with rapid excitation of muscles. Sub-elite runners who performed 6 weeks of plyometrics ended up with a 6% better running economy than a control group (#1556, Turner et al.). Effects on performance were not reported.
Sub-elite distance runners who added resistance training to their usual training for 21 weeks improved their 10-km time by 1.3% (not statistically significant) relative to a control group (#1559, Nicholson and Sleivert).
Training low while using a nitrogen house to live high (~2500 m) for 12-16 hours daily for 25 days produced a 5.7% increase in maximum oxygen consumption (VO2max) and a 3.5% increase in red-cell mass relative to live-low train-low controls in a study of triathletes and cross-country skiers (#277, Rusko et al.). Changes in erythropoetin concentration and arterial oxygen saturation while living high did not correlate significantly with changes in VO2max. Even the changes in red-cell mass did not correlate with changes in VO2max, either because errors in measurement obscured the correlation or because a placebo effect or some other mechanism mediated at least part of the increase in VO2max. One important effect of living high and training low appears to be an increase in exercise efficiency, in a controlled study of highly trained triathletes who spent 23 nights in a nitrogen house at 3000 m (#856, Gore et al.). That could account for any improved endurance performance, but it would not by itself produce an increase in VO2max. The physiological mechanism behind an altitude-induced improvement in efficiency is unclear.
The Dallas researchers (#811, Stray-Gundersen et al.) have now analyzed muscle biopsies of the subjects in their three training groups (live and train high, live and train low, live high and train low). Buffering capacity unexpectedly decreased in the high-high group, and there were no meaningful changes in muscle enzymes, fiber types, or capillary density in any groups. So an increase in oxygen-carrying capacity mediated by an increase in red-cell mass remains the most likely explanation for improvements in endurance performance produced by living high and training low. (For a review of the Dallas group's other data, see Baker and Hopkins, 1998.)
Living low and training high in an altitude chamber for 2 hours per day for 10 days at 2500 m increased maximum oxygen uptake (3.5%), Wingate mean power (3.8%), and Wingate peak power (4.9%) relative to low-low controls in a study of elite male triathletes (#787, Meeuwsen et al.). The athletes were in a pre-season phase of low-intensity training, so the benefits of living low and training high during a competitive season are uncertain.
Previous studies of carbohydrate loading have shown a clear positive effect on endurance performance long and hard enough to be limited by muscle glycogen stores. But these days athletes consume carbohydrate during such events, so is carbo loading still worthwhile? Probably. Although the result was not statistically significant, there was a tendency for carbohydrate loading to enhance mean power (by 2.4%) in a 100-km cycle test that included high-intensity bouts and carbohydrate feedings (#296, Schabort et al.). Carbohydrate loading also produced a 1.4% enhancement in pace during the last stage (~13 min) of a 100-min endurance cycling test in well-trained women, although the authors misinterpreted lack of statistical significance as "no evidence of an effect" (#880, Paul et al.). Athletes in both these studies were not blind to the carbohydrate loading, so the observed enhancements could be due at least in part to a placebo effect.
Relative to a moderate-carbohydrate meal, a high-carbohydrate meal 3 hours before a simulated 93-min mountain-bike trial produced a tendency towards a 3% enhancement of performance time (#882, Cramp et al.). Self-selection of different paces at the beginning of the trial may have been responsible for the difference in performance.
In a simulation of high-intensity team sports, physical and mental performance were substantially higher when players consumed a carbohydrate drink before and during the simulation relative to a drink containing no carbohydrate (#484, Welsh et al.). But were the players fasted, and do players normally go onto the field fasted?
A high-fat diet for 11 days followed by 2 days of carbohydrate loading and a high-fat pre-exercise meal probably enhanced performance in a cycle test lasting ~5 hours, relative to the control high-carbohydrate diet and meal (#295, Rowlands and Hopkins).
Five days of a high-fat diet produced a tendency for enhancement of cycling performance relative to a high-carbohydrate diet in a test lasting 2.5 hours, but the subjects performed the test under the unrealistic conditions of an overnight fast and no feeding during the test (#297, Burke et al.).
In a double-blind crossover, a supplement containing medium-chain triglycerides and carbohydrate produced no substantial change in performance of sub-elite female cyclists in an endurance test lasting ~140 min compared with carbohydrate alone (#459, Eimer et al.). A couple of previous studies had suggested that feedings of medium-chain triglycerides can increase fat oxidation and decrease carbohydrate utilization, but doses that the gut can tolerate don't seem to work.
Bench-press strength decreased by 0.7% and squat strength increased by 3.5% (not statistically significant) when resistance trained males increased training volume and protein intake, relative to a matched control group who did not change their diet (#485, Marcum et al.).
A pre-exercise drink of a nearly isotonic salt solution produced a 9% increase in work output relative to a hypotonic no-sodium placebo in a 15-min time trial following 45 min of submaximal work on a cycle ergometer (#1481, Coles et al.)..
In a symposium on weight control for athletes (#1445), Craig Horswill reviewed the evidence that modest weight loss has little effect on brief high-intensity performance, possibly negative effects on sustained short-term exercise, and probably negative effects on endurance.
There were too many free communications on creatine for me to evaluate fully. Short-term supplementation (4-7 days) enhanced single or repeated sprint performance in most (#355, #358, #364, #365, #1243, #1283, #1286) but not all (#356, #1276, #1285) studies. Longer-term supplementation (1-3 months) maintained (#354), increased (#362, #363), or had no effect (#357, #367, #1289) on strength or sprint performance. There were clear increases in muscle size in two studies (#362, #1274). One small-scale study found it was safe (#361), but another found a greater incidence of muscle cramps or tightness in a creatine supplementation group (3 out of 7 subjects) compared with a placebo group (1 out of 9 subjects) following prolonged dehydrating exercise (#1285, Webster et al.).
In a symposium on creatine (#1736) Ron Meyer opted for reduced decline in ATP concentration as the most likely mechanism for the benefit of acute supplementation, and he discounted any substantial effects of creatine's ability to shuttle energy from mitochondria, buffer hydrogen ions, or modulate enzymatic activity. Larry Spriet pointed out that the contribution of energy from (phospho)creatine increases while that from glycolysis decreases with repeated sprints, so that's presumably why it works better for repeated sprints than for single efforts. Bill Kraemer presented evidence of positive acute and chronic effects on strength. So all in all it looks like creatine really is a useful ergogenic supplement for some kinds of short-term performance, and it has anabolic effects with prolonged use in training.
Hydroxymethylbutyrate (HMB) is a potentially anabolic supplement, because it appears to reduce breakdown of protein. In a study of previously untrained males, HMB supplementation combined with resistance training did show signs of inhibiting muscle breakdown, and there was a tendency for greater gains in strength compared with placebo (#2069, Gallagher et al.). Supplementation with creatine, hydroxymethylbutyrate (HMB), or both for 28 days reduced the loss of strength following downhill running compared with placebo (#1275, Byrd et al.). HMB also reduced soreness.
Seven days of consuming Siberian ginseng had no substantial effect on prolonged cycling performance of moderately trained men (#444, Eschbach et al.). On the other hand, six weeks of CordyMax, a supplement based on a Chinese herb, produced a 7.0% increase in VO2max relative to placebo in a group of elderly (sedentary?) subjects (#774, Xiao et al.). Athletes will be next, presumably.
In a double-blind crossover trial, daily supplementation with beta-carotene for an unstated period enhanced 5000-m pace by 3.2% in sub-elite runners (#453, LeBlanc and Nelson).
Eight weeks of a zinc-magnesium supplement increased isokinetic strength at 180°/s and 300°/s by 8.9% and 7.9% relative to a placebo in football players (#483, Brilla and Conte).
Dehydroepiandrosterone (DHEA) and another potentially anabolic commercial supplement (AN6) did not increase muscle strength or alter body composition more than placebo when taken daily during eight weeks of resistance training (#1292, Reifenrath et al.).
Caffeine seemed to work in a sprint to exhaustion following a series of submaximal sprints (#2067, Whelan and Drinkwater).
Pyruvate is a partial breakdown product of glucose, but supplementing with it for seven days didn't have any substantial effect on performance of repeated sprints by young adult "subjects" (#2068, Hulver et al.).
A meta-analysis shows that moderate doses of anabolic-androgenic steroids have moderate effects on muscle size and strength (#2071, Spence). Any meta-analysts reading this article, please note: show outcomes for effects on performance as percent changes, not as Cohen-style effect sizes. Make sure you factor in the competitive ability or level of the athlete.
I could stay only for Hugh Welch's contribution to the mini-symposium on metabolic carts (#247, Yates et al.). Main points: the Naflon drying line in modern carts needs to be washed in distilled water and dried after each test or results will be spurious; most of the old carts aren't as good as bags; and there aren't enough published validation studies of the current generation of carts.
A resistance-exercise test (six sets of 80% of 10RM, each to exhaustion) showed modest reliability for the last three sets (typical within-subject variation 3-7%) (#227, Lambert et al.).
Validated against DEXA scans, skinfolds appear to be better than underwater weighing or bioelectric impedance as a way of measuring percent body fat in male marathon runners (#938, Lillystone et al.).
In the symposium on practical aspects of lactate measurement (#2041), Ralph Beneke claimed that the maximum lactate steady state is the gold standard. He admitted that measuring it is impractical, because the athlete has to come back to the lab on four or more separate days for a 30-min test. You have to keep testing until you find the maximum intensity that evokes a steady lactate (that is, the concentration varies by no more than 1 mmol/L during the last 20 min of the test). He didn't justify why he regarded this measure as a gold standard. As he was speaking, I realized that the words "maximum steady state" could easily trick people into thinking that this intensity is somehow special for an endurance athlete. In fact, no intensity is steady state as far as the fatigue process is concerned: whatever the intensity, the athlete will eventually fatigue. So what lactate intensity would qualify as the gold standard? If you're using a test to define a reference intensity for training, I presume it's the intensity that correlates best with pace in, say, a 30-min race. But if you're using the test to track performance, you want the best correlation between longitudinal changes in intensity and race pace. Whether the maximum lactate steady state comes out on top for either of these applications was not addressed.
The next speaker at the lactate symposium, Axel Urhausen, dealt with lactate measures derived from incremental tests. He did not identify which one was best, but he did point out that depletion of muscle glycogen has little effect on the so-called individual anaerobic threshold. This measure may therefore be good for tracking endurance performance when athletes vary their diet and training before each test. Unfortunately the protocol for the individual anaerobic threshold requires the athlete to go to maximum effort in the test, whereas most other lactate measures have the advantage that the test can be stopped before maximum effort.
Biomechanical studies relevant to athletic performance were almost entirely descriptive: a comparison of grab and track starts in swimming (#616), scoring of the Hecht vault (#619), and analysis of pole vault (#618), tennis serve (#621), glide kip (#622), windmill-style fastpitch (#623), and shot-put by wheelchair athletes (#615). One experimental study dealt with modifying slap skates (#611, Houdijk et al.), and another looked at the effects of verbal cueing on novices learning a golf swing (#613, Baldwin et al.).
Two studies provided evidence that an initial slow pace in an endurance event could enhance overall performance. In the first study (#383, Mattern et al.), cyclists performed a 20-km time trial at self-selected pace. They then performed the test on two further occasions, in which for the first 4 min they were forced to go either 15% slower or 15% faster than their mean pace in the first test. Overall their mean pace for the 20 km was faster when the first 4 min was slower, but the analysis needs to account for learning effects. In the other study (#617, Dutto and Smith), the best speed skaters at the winter Olympics and world championships showed the least decrement in lap speed. A slower start seems to be called for here, too.
Breathing 100% oxygen at 2.5 atmospheres for 100 min either 2 or 24 hours after eccentric exercise did not appear to make any difference to recovery of muscle damage (#206, Harrison et al.).
Muscle soreness following eccentric exercise was reduced by local icing (#212, Hiruma et al.) and local heating (#213, Weingand et al.), but not by consumption of a protease supplement (#214, Bailey).
Reduction in plasma glutamine has been suggested as a potentially useful marker of overtraining, but in a longitudinal study of swimmers the decline in glutamine with increased training did not recover after tapering, when the swimmers all performed well (#278, Koziris et al.). Sure, these swimmers weren't overtrained, but if you were monitoring glutamine you might think they were. Plasma malondialdehyde, a marker of oxidative stress and therefore another potentially useful marker of overtraining, increased in elite 4-km pursuit cyclists during an overload training camp (#1242, Johnson et al.). None got overtrained, though.
Acclimating to training in the heat for 7 days did not result in substantially enhanced simulated 40-km cycle time-trial performance at normal temperature in highly trained competitive cyclists (#1389, Morrison et al.).
Pre-cooling produced a 6% increase in distance traveled in a 30-min cycling time trial in the heat (#1514, Kay et al.), but a 5.6% reduction in time for a ~7 min run to exhaustion (#1517, Dugas et al.).
Oh no! The I form of the ACE gene (the gene for angiotensin converting enzyme) may not confer superior endurance and trainability after all. In a far larger study than those reported earlier (see Hopkins, 1998), non-athletes with two Ds increased VO2max by 23-25% more than those with two Is after 20 weeks of training (#1907, Rankinen et al.). If I'd crossed paths with the authors I'd have asked them to explain away the earlier findings.
Eric Hoffman, in a President's Lecture on the molecular biology of muscle (no abstract), made it clear that other performance genes are in the pipeline. He entertained with his research on the discovery of the gene that makes "Schwartznegger" quarter horses (a point mutation in a gene for a sodium channel in muscle). He then explained how mutations in genes for membrane structural proteins give rise to greater sensitivity to muscle damage, which is why some muscular dystrophies begin with a phase of hypertrophy before the muscle is damaged beyond repair. Abnormal dystrophin proteins, which can give rise to Becker muscular dystrophy, appear to be responsible for the success of some athletes. Performance enhancement using adenoviruses to insert genes into muscle may be just around the corner.
I was dismayed by the widespread misinterpretation of p>0.05 in the abstracts, posters, and talks at this meeting. What's more, few people in our discipline understand the need to state the magnitude and precision of estimates of their outcomes. So I was disappointed that the turnout at my tutorial lecture on confidence limits was not that great. Please read the relevant section [http://www.sportsci.org/resource/stats/generalize.html] in the stats pages at this site. You can also download the slides I showed in the lecture.
Do muscles run short of oxygen in maximal endurance exercise? In one form or another this question has been the focus of several review articles in Medicine and Science in Sports and Exercise over the last few years, and it featured in a brief item [http://www.sportsci.org/news/ferret/ferret9803.html#fatigue] at this site. The outgoing editor of MSSE, Peter Raven, brought together two mavens of exercise science to debate the issue in what for me was a high point of the conference. Tim Noakes of the University of Cape Town argued that muscles don't become hypoxic, because the body has a protective mechanism that limits drive to the muscles. Brian Whipp of the University of London argued that skeletal muscles do become hypoxic, because they try to consume more oxygen than the heart can transport. The rhetoric was wonderful. Noakes, citing a personal communication from firstname.lastname@example.org: "I believe Tim Noakes: he fits his theories to My facts." Whipp, referring to Noakes by citing Voltaire on Descartes: "He seems intent on excising the errors and replacing them with his own."
Tim Noakes Brian Whipp
Brian Whipp first attacked Noakes’ linking of heart failure to muscle hypoxia. He pointed out that cardiac output is likely to be limited not by impaired coronary function but simply by limits to heart rate and stroke volume. He cited evidence of "coronary reserve": infusion of adenosine increases coronary flow at maximum exercise. So if heart muscle needs more blood, coronary resistance will simply decrease and divert blood from the periphery to the heart muscle. He questioned the earlier evidence against muscle hypoxia (based on biopsies) but he did not address the evidence against hypoxia in Richardson's recent work (based on NMR spectroscopy of active muscles in vivo). He presented an elegant model whereby regions of slight mismatch in rates of metabolism and perfusion in muscle can produce patchy regions of hypoxia. This model, with other data, predicts that there is indeed no plateau at VO2max in a single test, either with incremental or supramaximal constant-load exercise. Nevertheless, a plateau is obvious in plots of the VO2 at the endpoint of repeated constant-load tests of increasing intensity. He pointed out that Noakes' arguments about the effects of altitude are based on chronic exposure to altitude, whereas the effects of various forms of acute exposure are consistent with the classic model. He made several other minor points before finishing with this quote, by St Paul in 1 Timothy 6:20: "O Timothy, keep that which is committed to thy trust, avoid profane and vane babblings and oppositions of science."
So, are muscles hypoxic in intense exercise? I can't see that Whipp's plateau is pivotal, because it could be a consequence of the way Noakes' governor cuts in. Whipp's idea of patchy hypoxia is compelling, but it seems to be incompatible with Richardson's data. To the extent that there is no hypoxia, Noakes is right. Whether his governor exists and can explain how fatigue limits endurance performance is another matter.
Muscles may not become hypoxic, but you can certainly push them to a higher intensity than they can maintain aerobically. For example, at 110% of your maximum oxygen uptake, your muscles have to use anaerobic mechanisms to meet the shortfall in your demand for power. At such a high intensity you fatigue after several minutes. What stops you at the point of maximum effort? Muscle fatigue resulting directly from depletion of anaerobic capacity? Or a reduction in motor drive dictated by a governor that protects the heart? In his response to a draft of this article, Noakes would concede only that more data are needed to settle the issue. He and Zig Gibson are currently looking for that reduction in drive.
Baker A, Hopkins WG (1998). Altitude training for sea-level competition. In: Sportscience Training & Technology. Internet Society for Sport Science: sportsci.org/traintech/altitude/wgh.html
Hopkins WG (1998). Performance gene discovered. Sportscience 2(4), sportsci.org/jour/9804/inbrief.html#gene (475 words).
Edited by Stephen Seiler.