Vespa Gel

Vespa claims to improve endurance performance by improving fat burning. However the science does not seem to back this up, with no human studies that show improved endurance. The few animal studies that exist generally use vastly higher doses, the equivalent of 100-500 packets of Vespa. So why does Vespa seem so popular with top ultra-endurance athletes?

Contents

1 Background

At 2 inches (50mm) long, the Asian giant hornet is the largest of the hornets and is able to fly long distances. These giant hornets consume only a liquid diet of 43% carbohydrate and 57% amino acids, with the carbohydrate as the primary fuel for flying. The hornets get this mixture from the saliva of their young. There are two companies that are marketing products based on this saliva:

• Vespa contains Giant Hornet Saliva (GHS) extracted from the insect larvae and each standard packet contains 100mg of the saliva. This would provide about 57mg of amino acids. Note that while Vespa is a gel like product, it is not an energy gel and does not provide a meaningful source of fuel.
• Another company produces Hornet Juice which contains 17 amino acids in the ratio found in the Giant Hornet Saliva. This mixture is often referred to as VAAM (Vespa Amino Acid Mixture).
  • Hornet Juice is made in New Zealand, and a similar product, VAAM is produced in Japan. The two differ only in the artificial sweeteners used.
  • Having a product and the mixture both called VAAM does cause some confusion, so this article will use 'Hornet Juice' for the products and VAAM for the amino acid mixture.

The two types of product are similar, but not identical.

• GHS contains not just the 17 amino acids, but also small amounts of Taurine, beta-alanine and GABA.
• The amino acids in GHS may be in the form of peptides, not isolated amino acids.
• GHS contains carbohydrates, including glucose and Trehalose.
• The amount of the active ingredient differs, with VAAM/Hornet Juice containing 3.7g of amino acids and Vespa containing only 100mg (about 57mg of amino acids).

2 The proposed Mechanism of Action

The proposed Mechanism of Action behind GHS and Hornet Juice is that it triggers a switch to burning fat for exercise rather than carbohydrate. This is not how the mixture works in the Giant Hornet, which primarily uses carbohydrate for fuel. The Giant Hornet gets over half its calories from amino acids, not a tiny amount that acts as some type of catalyst.

3 Scientific Studies

The science behind VAAM is rather limited, with no studies that show an improvement in human endurance performance, and one study showed a reduction in fat burning. There is only one study that used GHS. Here is a brief summary of the studies that are available.

• When untrained, fasted mice were put in a pool and forced to swim to avoid drowning, VAAM improved swimming times over water, glucose or casein (a milk protein). This is the only study that shows improved endurance rather than other blood markers. The VAAM given was equivalent to about 900 packets of Vespa (700mg/Kg or 52.5g for a 150Lb/75Kg human).
  • This study also checked the blood lactate and glucose level in the mice after 30 minutes of swimming. This test used GHS in addition to VAAM, and GHS showed a greater level of blood glucose than VAAM, with a similar (low) level of lactate.
• A second similar study again used untrained, fasted mice that swam to avoid drowning. Here VAAM showed increased fat metabolism and increased ketogenesis compared with casein.
• In untrained subjects, 5.7g VAAM slightly reduced the RPE during 60 minutes of cycling compared with a non-calorie control. However, the rate of fat burning was actually higher in the placebo case than with VAAM. (The study does some ugly analysis to show that because the fat burning started even lower for the placebo, it went up relatively more than the VAAM.) The study used the equivalent of about 100 packets of Vespa.
• Taking 10.8g of VAAM with 38g of sugar increased markers of ketogenesis during 45 minutes of exercise in humans compared with 48.8g sugar. Note that the study did not show that fat burning was increased, just markers of ketogenesis. It's possible that the differences in ketone levels are due to the extra carbohydrate intake rather than the VAAM. This experiment used the equivalent of about 190 packs of Vespa.
• Taking VAAM (3g/day) & exercise or placebo & exercise for 12 weeks in previously sedentary elderly women produced a slightly greater improvement in fitness and body fat. This is the equivalent of about 50 packs of Vespa per day.
• An unpublished study (available as an abstract only) showed that 10g/70Kg of VAAM produces a slight reduction in heart rate during 30 minutes of exercise compared with casein protein. This is the equivalent of about 175 packs of Vespa.
• VAAM given to rats prevented a rise in blood lactate and increased markers of ketosis compared with a milk protein control.

3.1 Publication Bias?

Publication bias is where the studies that support an idea are published and those that are unsupportive are ignored. It is hard to know when publication bias occurs, though there are some techniques such as the funnel plot that can help when there is plenty of comparable research available. However, looking at the research that has been performed on VAAM, it seems strange that the obvious test of a potentially endurance performing supplement has not been performed. That obvious test is to check if endurance is actually improved, such as a treadmill test to exhaustion or a simulated time trial. This type of test would be easier to perform than some of the existing research, and would give a more definitive answer. That, combined with the egregious bias shown in the cycling study above, suggests to me some level of publication bias.

4 Explanations for Anecdotal Support

There are a large number of highly publicized endorsements of Vespa, including a number of top athletes. There are a number of explanations for this.

• The anecdotal evidence may be misleading. Some of the anecdotes are based around athletes completing ultramarathons on fewer calories than is normally expected. However, for my 100 mile PR (16 hours) I only consumed about 2,700 calories, which is also far less than is normally expected.
• There's an old saying that "absence of evidence is not evidence of absence." It is possible that Vespa does improve endurance performance, but the supporting scientific studies have not yet been performed. While this is possible, this seems unlikely to me given the studies that have been published and the tiny amounts of the saliva that are included in each packet.
• The benefit of Vespa may be due to the placebo effect, especially given the high cost which generally improves the effectiveness of a placebo. Some of the reports of improved performance with Vespa are when it's taken with large amounts of protein, which should negate any benefit, strongly suggesting a placebo effect. To understand how widespread the anecdotal evidence could be for a placebo, you can look at Homeopathy which is generally accepted as quackery.
• The way Vespa is used may result in indirect performance improvements. Vespa recommends a much lower level of calorie intake when using the Vespa than is common in ultra-endurance events. It may be that athletes that use Vespa may consume a more effective level of fueling that promotes better performance.

5 Testing Vespa

I did a simple trial of Vespa using two runs, one with Vespa and one with a placebo. I had my son make up two electrolyte drinks, each with 3/4 teaspoon of salt and two packets of sugar free flavoring in 64 ounces of water. In one he added 2 packets of Vespa and the other 10g honey. Each packet of Vespa contains 5g of Honey and I wanted to eliminate any impact that might come from the carbohydrates. This proved to be important as there was slight taste of honey detectable in both cases even with the flavoring. At the time of the run I'd been on a ketogenic diet, with blood ketones at or above 1.0 mmol/liter for more than four weeks. On each run I took half the drink 45 minutes before running, then the remainder at the half way point of the marathon distance. I drank other fluids to thirst, but had no other calories. This protocol matches the recommendations on the Vespa web site that states "For most marathoners it is recommended that one VESPA is taken 45 minutes prior to the race and another at the half marathon mark". Each run was performed in the morning on a treadmill set to 8:00 min/mile pace and a 4% decline. The treadmill is in a garage, so temperature and humidity was high, but similar in both cases.

5.1 Blood metrics

The samples below were taken using a Precision Xtra meter and ketone strips, a TRUEresult meter and glucose strips, and a ReliOn 741CREL blood pressure meter. I stopped running to obtain the measurements at the half way point, which took a similar amount of time on each test. Two blood glucose samples were taken at each point to counter inaccuracy of glucose testing. The variation between the two tests is within what I would consider random variation. The blood ketone level is slightly lower in the Vespa test, which might indicate slightly less fat burning, but both tests indicate ongoing ketosis.

Time	Measurement	Test 1 (placebo)	Test 2 (Vespa)
rowspan="3"| 45 minutes before the start.	Blood glucose	80 (78, 81)	76 (75, 76)
	Blood ketones	1.2	1.0
	Blood pressure	111/80, 67	128/81, 57
rowspan="3"| Half way	Blood glucose	83 (82,83)	81 (81,80)
	Blood ketones	0.8	0.6
	Blood pressure	97/63, 96	101/72, 94
rowspan="3"| Finish	Blood glucose	82 (80,83)	84 (84, 83)
	Blood ketones	1.1	0.9
	Blood pressure	93/68, 102	96/65, 102

5.2 Heart Rate metrics

The table below shows the heart rate data for the two runs. The distances are close to miles, but calibration errors on my Footpod resulted in each distance lap being about 1.07 miles. There does not appear any significant difference in heart rate values except for the last couple of miles which are probably due to slight differences in hydration.

Distance	Test 1 (placebo) Avg. HR	Test 2 (Vespa) Avg. HR
1	108 (56%)	114 (58%)
2	121 (62%)	119 (61%)
3	122 (63%)	121 (62%)
4	122 (63%)	122 (62%)
5	121 (62%)	120 (62%)
6	127 (65%)	124 (64%)
7	127 (65%)	125 (64%)
8	128 (66%)	125 (64%)
9	123 (63%)	118 (60%)
10	126 (65%)	120 (61%)
11	129 (66%)	123 (63%)
12	129 (66%)	124 (64%)
13	122 (63%)	119 (61%)
14	121 (62%)	119 (61%)
15	124 (64%)	125 (64%)
16	128 (65%)	125 (64%)
17	130 (67%)	129 (66%)
18	123 (63%)	134 (69%)
19	124 (63%)	126 (65%)
20	131 (67%)	129 (66%)
21	132 (68%)	131 (67%)
22	123 (63%)	134 (69%)
23	119 (61%)	137 (70%)
24	134 (69%)	138 (71%)
25	135 (69%)	141 (72%)
Overall	125	126

5.3 Perceived Exertion

Both runs had similar levels of perceived exertion through the duration.

5.4 Initial Fatigue

One concern with this test is if the first run creates fatigue that changes the second run. Looking at the metrics I track for my training, the two tests have similar levels of training stress and this length of run is not unusual for me; I've run the marathon distance or longer 48 times in the last 12 months.

Training Metric	Test 1 (placebo)	Test 2 (Vespa)
Total miles in previous 28 days	320	329
Training Monotony	1.06	1.25
CTL	326	321
ATL	330	321
Banister Fitness	7002	6905
Banister Fatigue	4243	4131
Busso Fitness	9952	9866
Busso Fatigue	5738	5634

5.5 Test Conclusion

I learned after both tests that the Vespa condition was the second test. I was not able to detect a difference between the Vespa condition and the placebo based on any metric. Obviously this test has only a single sample, and while I tried to standardize as many of the conditions as possible, there are still variations in temperature, hydration, nutrition on the prior day, sleep patterns, or other minor factors.

5.6 Test Limitations

There are obvious limitations on this test, but it is the best I can do given the limited resources I was able to provide. To be of scientific value, the test would need to be repeated with a larger sample size. However, I hope that this will inspire people to perform similar blinded tests of products.

6 Recommendations

• It's hard to recommend taking Vespa given it costs $6.75 per packet ($69 for 12) and has little scientific support. To get the amount that improved performance in untrained mice would require eating 900 gels and cost $5,175. On the other hand, there is little in the Vespa that could cause a problem, so other than the cost it should be low risk.
• Using Hornet Juice is cheaper at $3 per serving and each serving provides 37 times more VAAM than Vespa.
• I tried Hornet Juice a number of times on long (24-30 mile) training runs and noticed no difference, but as noted above, Hornet Juice and Vespa are not identical in their ingredients.
• My simple test of Vespa showed no benefit at the marathon distance.
• It's quite possible that Vespa acts only as a placebo, but it's been shown that a good placebo can improve performance dramatically.