The conventional wisdom is that fast-releasing carbohydrate is better at promoting post-exercise recovery than slower-releasing carbohydrate. But is this really the case? Andrew Hamilton looks at the scientific evidence... MORE
Are protein drinks effective in boosting sports performance?
New research suggests that there’s less evidence for adding protein to carbohydrate than many would have us believe
In recent years, the addition of protein to carbohydrate to enhance endurance performance has become a popular practice for endurance athletes. However, according to Kevin Tipton and Asker Jeukendrup, there’s less evidence for this practice than many would have us believe.
Evaluation of nutritional interventions to improve athletic performance such as adding protein to carbohydrate drinks is a notoriously tricky matter. Given that a difference in performance of less than 1% may be the difference between a gold medal and not even making the team, measurement of performance must be very rigid. Unfortunately, that sort of precision is not easy to achieve in a laboratory. There are three factors that must be considered when evaluating a measure of performance: validity, reliability and sensitivity(1). Let’s take a look at what these mean:
Validity – A valid protocol must measure performance as closely as possible to the real thing. The primary methods used to assess endurance performance are the time to exhaustion protocol (TTE), in which the subject exercises for as long as possible at a set intensity, and simulated time trials (STT). Clearly, the validity of a TTE must be questioned. We are unaware of any events in which a gold medal is awarded for exercising as long as possible at a set intensity! Most endurance events are won by the competitor who finishes the fastest – as in a STT. Furthermore, the physiological responses to STT trials are more similar to those of competition than TTE trials (1). For these reasons, we consider TTE as an evaluation of endurance capacity, not performance.
Reliability – The ability to reproduce findings across studies. The reliability of TTE has been determined to be somewhere in the range of 25%, whereas STT is closer to 3-4%(2). These numbers suggest that on a day-to-day basis, with no other factors changing, changes in TTE will be roughly 6 times greater than STT. Certainly, improvements that large with protein ingestion would be remarkable and would revolutionise sport, making champions out of also-rans!
Key: TTE – time to exhaustion test; STT – simulated time trial; Ss – subjects; Cho – carbohydrate.
Overall – refers to the applicability of the conclusions for a recommendation to top-level endurance athletes.
Ratings: (based a four star system)
- **** Best available – solid conclusions regarding performance possible.
- *** Very good – conclusions regarding endurance performance based on these methods may be made with only minimal equivocation.
- ** Fair – conclusions regarding endurance performance must be tempered and carefully evaluated prior to making recommendations.
- *Poor – methodology used is inappropriate for making conclusions regarding endurance performance and should not be used to make recommendations.
These numbers lead to another important point. Reliability is much better with well trained subjects(1). It is not much of a stretch to imagine that the responses of highly trained athletes are different than those of less well trained athletes. So, if assessment of the importance of adding protein to carbohydrate for performance of elite athletes is the goal, then for both validity and reliability reasons, STT in well trained athletes should be the test used.
Sensitivity – The signal to noise ratio of a protocol. If there is a lot of noise (poor reliability), the signal will not be detected unless it is very large. If performance changes are small compared to the noise of the measurement, the sensitivity is low. Sensitivity is more difficult to determine than reliability, since few studies report sensitivity or provide the data that make it possible to retrospectively calculate sensitivity.
There is a great deal of interaction among these factors. Taken together, it is clear that when the impact of a nutritional intervention, such as the addition of protein to carbohydrate, is the goal, STT protocols are preferable to TTE. Furthermore, the performance level of the subjects tested should be close to that of the athletes for whom recommendations will be made.
In Table 1, we have summarised our evaluations of nine studies (see references 3-11 overleaf) that have attempted to determine the impact of adding protein to carbohydrate. We have put that table together from the standpoint of using the studies to make recommendations for elite athletes.
Protein plus carbohydrate ingestion during exercise
Hopefully, you can see that if we are to properly evaluate the importance of adding protein to carbohydrate for endurance performance, the design of the studies must be carefully considered. There have now been many studies that have examined the impact of protein ingestion during exercise on endurance capacity and performance (see figure 1). Several have demonstrated a dramatic (13-26%) improvement in endurance capacity, ie TTE, when protein is ingested during exercise.
However, before we get too excited about additional protein for endurance performance, it is important to point out that up till now, no study has demonstrated improved STT time, ie endurance performance, with additional protein ingested during exercise. So, a strict interpretation of the research to date would be that if the goal is to ride a bike as long as possible at a fixed intensity, then ingesting protein along with carbohydrate during exercise is advisable. On the other hand, if going faster is the goal then there is no evidence that adding protein to carbohydrate is worthwhile.
Another factor to consider is that the results of many TTE studies are questionable because the time was actually shown to the subjects in those studies. It has been demonstrated quite elegantly that providing feedback about the time will not result in an ‘honest’ measurement of performance(2) This problem is even more pronounced when carbohydrate and carbohydrate plus protein drinks are provided; it’s very difficult to mask the taste of the protein, making trials very difficult to perform in truly a ‘double blind’ fashion.
A further aspect of performance that has recently been suggested to benefit from protein ingestion during exercise is ‘late-exercise performance’. A recent study from a laboratory in Virginia concluded that whereas protein did not impact overall STT time, completion of the latter stages was faster(3). But does this result mean protein should be recommended? If we were to tell an athlete that adding protein allowed them to go faster at the end of a race, they would certainly get very excited. However, the excitement would almost certainly be limited once we explained that overall time was not improved. Moreover, if an athlete is faster at the end of a race, but not overall, does that not mean the early stages must have been slower? So, the results from this single study cannot be considered conclusive. More research needs to be performed to determine if, in fact, protein ingestion helps late-stage exercise performance.
In the light of the above, we have recently tested that very question in our laboratory. We used a well established STT protocol that has been tested to be very reliable and as sensitive as possible. Moreover, this STT is ‘front-loaded’ – that is, the cyclists ride for two hours at a submaximal intensity prior to the STT itself. In this way, we can specifically test late-stage exercise performance.
The cyclists participated in two pre-loaded STT protocols in which they consumed carbohydrate alone or carbohydrate plus protein. We found no significant difference in time to complete the time trial. In fact, the average time was slightly faster for the trial without protein. Moreover, there was no difference in average power output during the rides, which is often considered to be a much more sensitive indicator of performance(1). Finally, power output declined in the last half of the time trial, but there was no difference in power output between carbohydrate-only and protein-carbohydrate at any time during the STT, including the last two stages of the ride. Thus, in a well controlled assessment of late-stage cycling performance, there was no impact on performance from consuming extra protein. Finally, it’s worth adding that although many have been suggested, there’s no sensible mechanism identified by which protein may improve endurance performance.
Despite some rationale and even some evidence, we don’t feel that a solid recommendation for endurance athletes to ingest protein in addition to carbohydrate can be supported by the science. Certainly, the solid and specific recommendations that are often made are not supported by the research mainly because there’s simply no evidence that ingesting protein during exercise leads to better performance in that event. Although it’s a well worn cliché, more studies are needed – specifically with highly trained athletes, using appropriate techniques and controlled for diet before and after exercise (a notoriously difficult task)!
Kevin Tipton is a senior lecturer in exercise metabolism in the School of Sport and Exercise Sciences at the University of Birmingham UK.
Asker Jeukendrup is professor of exercise metabolism at the University of Birmingham and has published over 150 research papers and books on exercise metabolism and nutrition, as well as being consultant to many elite athletes.
1. Sports Med 2008;38:297-316
2. Med Sci Sports Exerc 1996;28:266-270
3. Int J Sport Nutr Exerc Metab 2009;19:136-149
4. Int J Sport Nutr Exerc Metab 2003;13:382-395
5. Med Sci Sports Exerc 2004;36:1233-1238
6. Med Sci Sports Exerc 2006;38:1476-1483
7. Med Sci Sports Exerc 2006;38:1608-1616
8. J Strength Cond Res 2007;21:678-684
9. J Sports Sci 2008;26:227-233
10. Int J Sport Nutr Exerc Metab 2008;18:363-378
11. Med Sci Sports Exerc 2009; In Press