What you’re getting yourself into:
~3,300 words, 11-22 minute read time
Key Points:
Many people think steroids make a massive difference in terms of competitiveness in strength sports. They are wrong.
Other people (somehow) think that steroids don’t make much of a difference. They are also wrong.
Steroids DO help people gain muscle mass and absolute strength at a much faster rate, but the increase in muscle mass generally means you’re forced to move into a heavier weight class where you need to lift more in order to be equally competitive.
When looking at world records, experimental evidence, and cross-sectional studies, the resulting picture shows that steroids increase your competitiveness in strength sports by about 10%.
This is the second part in a 7-part series about steroids. Make sure you check out the preceding parts before diving in, so you don’t feel like you’re stepping into the middle of a conversation.
Part 3 – Why the benefits of steroids are blown out of proportion
Part 4 – Genetics vs. steroids
Part 5 – How steroids affect training
Part 6 – Fat Free Mass Index, and steroids in other sports (team sports, endurance sports, and bodybuilding). Dose and compound considerations.
Part 7 – Why people use steroids, and how they avoid getting caught, ethical issues with steroid accusations, and closing thoughts
This is a follow-up I’ve been planning ever since the
Science of Steroids was published. The outline has been collecting dust as a draft in the post editor for almost a year, and a recent round of discussions spurred by
this article reminded me that this outline existed.
The section of that article people vehemently disagreed with:
When both drug-free AND drug-using lifters reach close to their body’s physiological peak (something like 10-15 years after they start training), they are pushing as hard as they can, which means that their training will actually be very similar in structure. The only big difference is that drug-using lifters will have their performance around 10% higher than drug-free lifters, if not a bit more.
It seems the prevailing notion is that steroids are literally a game changer in every aspect of training and performance; they allow (or require) you to train in a drastically different manner, and they improve performance dramatically.
That notion is dead wrong. This article deals with why a 10% increase in performance is pretty much dead-on (or perhaps a bit high), and part 5 of this series will tackle why drugs don’t impact training all that much, at least in the context of strength sports.
Performance
To start with, let’s peruse the
all-time world records in powerlifting. For this analysis, I’m using squats and totals without knee wraps (because most of the biggest drug-tested organizations don’t allow knee wraps), and bench presses performed in a full meet; this doesn’t change the outcome very much anyways. And, for posterity’s sake, I’m using the numbers as they stand at the time of writing on August 27th, 2015. All weights are in pounds.
| Powerlifting World Records |
| Squat |
| Weight Class | Drug-Tested | Untested | Difference |
| 123 | 639 | 639 | 0% |
| 132 | 551 | 551 | 0% |
| 148 | 556 | 556 | 0% |
| 165 | 553 | 610 | 9.34% |
| 181 | 617 | 744 | 17.07% |
| 198 | 750 | 766 | 2.09% |
| 220 | 667 | 785 | 15.03% |
| 242 | 727 | 826 | 11.99% |
| 275 | 850 | 854 | 0.47% |
| 308 | 859 | 914 | 6.02% |
| 350 | 938 | 938 | 0% |
| | | 5.64% overall |
| Bench Press |
| Weight Class | Drug-Tested | Untested | Difference |
| 123 | 391 | 391 | 0% |
| 132 | 380 | 424 | 10.38% |
| 148 | 402 | 440 | 8.64% |
| 165 | 462 | 485 | 4.74% |
| 181 | 468 | 556 | 15.83% |
| 198 | 496 | 565 | 12.21% |
| 220 | 518 | 582 | 11% |
| 242 | 529 | 603 | 12.27% |
| 275 | 585 | 650 | 10% |
| 308 | 565 | 666 | 15.17% |
| 350 | 710 | 710 | 0% |
| | | 9.11% overall |
| Deadlift |
| Weight Class | Drug-Tested | Untested | Difference |
| 123 | 562 | 634 | 11.36% |
| 132 | 600 | 628 | 4.46% |
| 148 | 697 | 697 | 0% |
| 165 | 684 | 717 | 4.6% |
| 181 | 766 | 791 | 3.16% |
| 198 | 825 | 870 | 5.17% |
| 220 | 859 | 901 | 4.66% |
| 242 | 804 | 893 | 9.97% |
| 275 | 906 | 906 | 0% |
| 308 | 939 | 939 | 0% |
| 350 | 903 | 1015 | 11.03% |
| | | 4.95% overall |
| Total |
| Weight Class | Drug-Tested | Untested | Difference |
| 123 | 1306 | 1339 | 2.46% |
| 132 | 1457 | 1457 | 0% |
| 148 | 1432 | 1482 | 3.37% |
| 165 | 1560 | 1650 | 5.45% |
| 181 | 1727 | 1840 | 6.14% |
| 198 | 2015 | 2015 | 0% |
| 220 | 1868 | 2099 | 11.01% |
| 242 | 1912 | 2080 | 8.08% |
| 275 | 2171 | 2226 | 2.47% |
| 308 | 2102 | 2353 | 10.67% |
| 350 | 2205 | 2298 | 4.05% |
| | | 4.88% overall |
Notice that the untested records are an average of 5.64% higher for the squat, 9.11% higher for the bench press, 4.95% higher for the deadlift, and 4.88% higher for the total. Also notice that there are 11 instances where the difference between the tested and untested records is 0%. In those instances, someone lifted more in a drug-tested meet than anyone’s ever lifted in an untested meet.
There are some obvious issues with this data set I’ll address in a bit.
Turning to weightlifting records, I’ll be comparing the
world records (it’s acknowledged that rampant steroid use takes place in most of the countries that regularly produce the world’s best weightlifters) to the
American records. All numbers are in kilograms.
| Weightlifting Records |
| Snatch |
| Weight Class | World Record | American Record | Difference |
| 56 | 138 | 112 | 18.84% |
| 62 | 154 | 123 | 20.13% |
| 69 | 166 | 135 | 18.67% |
| 77 | 176 | 157.5 | 10.51% |
| 85 | 187 | 166 | 11.23% |
| 94 | 188 | 169 | 10.11% |
| 105 | 200 | 173 | 13.5% |
| 160 | 214 | 197.5 | 7.71% |
| | | 13.84% overall |
| Clean and Jerk |
| Weight Class | World Record | American Record | Difference |
| 56 | 170 | 135 | 20.59% |
| 62 | 182 | 153 | 15.93% |
| 69 | 198 | 174 | 12.12% |
| 77 | 210 | 190 | 9.52% |
| 85 | 218 | 203 | 6.88% |
| 94 | 233 | 211 | 9.44% |
| 105 | 242 | 220 | 9.09% |
| 160 | 263 | 237.5 | 9.7% |
| | | 11.66% overall |
| Total |
| Weight Class | World Record | American Record | Difference |
| 56 | 305 | 245 | 19.67% |
| 62 | 332 | 272 | 18.07% |
| 69 | 359 | 305 | 15.04% |
| 77 | 380 | 342.5 | 9.87% |
| 85 | 394 | 362 | 8.12% |
| 94 | 418 | 372 | 11% |
| 105 | 436 | 390 | 10.55% |
| 160 | 472 | 430 | 8.9% |
| | | 12.65% overall |
Here, the world records are, on average, 13.84% higher in the snatch, 11.66% higher in the clean and jerk, and 12.65% higher for the total.
Now, there are some problems with both of these data sets. Starting with powerlifting:
- There are people with drug-tested records who have failed drug tests and were almost certainly using steroids when they set their records; Sergey Fedosienko and Ed Coan come to mind.
- There are likely others who were on drugs when they set drug-tested records, because the drug tests are (or at least used to be) easy to beat. Organizations affiliated with the IPF have started drug testing out of contest in recent years, which makes it more difficult to get away with steroid use (though it’s still far from impossible), but many of the records are from an era when drug testing only happened on the day of the meet, or were set in federations that still only do in-meet drug testing. In that context, it’s just a matter of cycling off steroids far enough out from the meet that you pee clean – a simple matter of knowing the half lives of the compounds you’re taking. Hardly rocket science.
- The IPF recently restructured its weight classes, so about a dozen of the drug-tested records are actually a bit lower than they should be. For example, Krzysztof Wierzbicki’s world record total at 220 was actually set at a bodyweight of 207 – those extra 13lbs can make a big difference.
- Equipment and judging differences play a role as well. In general, the major drug-tested organizations (the IPF and its affiliates) require deeper squats and aren’t as lenient on shaky deadlift lockouts. There are exceptions (the Raw Unity meet, USPA meets, and major GPC meets don’t drug test and generally have very strict judging, and other organizations with a drug-tested division are a bit more lenient), but drug-tested records are generally set in meets with stricter judging standards. Equipment plays a role as well. Stiffer squat bars, thinner deadlift bars, and squatting out of a monolift help people lift a bit more weight, and are allowed in many untested meets, but aren’t allowed in the IPF and its affiliates.
- Weight cuts factor in. Most of the untested records and a few of the drug-tested records were set in organizations that allow the lifter to weigh in the day before the meet. This allows the lifter to make a lighter weight class by shedding water (often 5-10% of someone’s body weight), and gives them time to rehydrate before the meet. The bulk of the drug-tested records were set in organizations that have weigh ins on the morning of the meet. Lifters will often still cut 1-3kg (2-6lbs) of water weight, but they’re much closer to the actual weight class limit when they step on the platform. Combine this with the restructured IPF weight classes, and you’ve got the potential for huge differences in body weights. For example, Krzysztof Wierzbicki’s drug-tested 220 record of 1868 was set at a bodyweight around 207, and Dan Green’s 220 total record of 2099 was set at a bodyweight around 235 or 240 – a massive difference.
- Powerlifting isn’t a hugely competitive sport. If you can make the assumption that the world records represent the people with the most elite genetics for a particular sport, then the majority of the variation between drug-tested and untested records should be explained by the effects of drugs (or at least the difference between the amount of drugs someone can take and still pass a drug test, and the “everything and the kitchen sink” approach to steroid use). You can’t win an Olympic medal in powerlifting, and there are no strong financial incentives, so the sport just doesn’t attract the cream-of-the-crop talent you’d need for an accurate comparison.
On the whole, it appears that the drug-tested records in powerlifting are probably a little bit higher than they “should” be, if we assumed they were, in fact, all set without steroids, and that the records were all set under identical circumstances. Some of the drug-tested records were doubtlessly set by people using steroids (meaning those records are inflated), but many were also set with stricter judging, less helpful equipment, at lighter body weights, and without the aid of massive water cuts (meaning those records are lower than they could/should be). My supposition is that the advantage of drug use more than counterbalances those other disadvantages, but it’s impossible to say how large of an effect those opposing factors make.
Now, let’s turn our attention to the weightlifting records. I’m assuming that almost every American record was set without drugs. That’s a pretty safe assumption. USADA (the American arm of the World Anti-Doping Agency) is pretty rabid about catching steroid users in weightlifting, and the culture of American weightlifting is militantly anti-steroid. They test very frequently out of contest, meaning that American weightlifters on steroids would have to be very careful about microdosing (more on that in a later article), and generally resort to drugs and dosages that are 1) more expensive and 2) less helpful. USADA is also very good at what it does. They got
two samples from
Pat Mendes that were positive for human growth hormone, which is remarkable because hGH has a half life of
only 10-20 minutes. If USADA suspects someone is using performance-enhancing drugs (essentially any high-level weightlifter), they pursue that person with a passion.
I’m also assuming that almost every world record was set with the use of drugs. Many other countries’ testing agencies aren’t as rabid about catching drug users as USADA, and there are strong incentives to use (Olympic glory, and financial incentives for winning medals and setting records in many countries).
There are fewer issues with this data set, but there are several notable ones:
- There are almost certainly lifts that exceed the American records that have been done without steroids. Other countries (Canada, Australia, much of Northern and Western Europe) test their athletes just as rigorously as USADA tests American lifters. Quite frankly, I’m too lazy to research every country’s testing procedures and compare all the records across all those countries.
- At least in the US, it’s almost certain that the people most gifted for weightlifting aren’t participating in the sport. They’re playing in the NFL, making millions of dollars per year. Weightlifting is a niche sport that not many Americans know about (though CrossFit is rapidly changing that), and it just doesn’t have as many incentives as other sports. There are exceptions, of course. The most obvious is CJ Cummings, who was already breaking senior national records at the ripe old age of 14. It’s likely that the weightlifters who hold the world records represent a pool of athletes more suited for the sport, due to 1) increased interest/visibility/incentives increasing the athlete pool 2) national talent identification programs 3) lack of American football (the major sport in the US is largely power-dependent. Soccer, the dominant sport in the rest of the world, is to some degree, but not nearly to the same degree as American football).
- Again, any test can be beaten. There’s no guarantee that none of the American records were set by people on steroids. There’s also no guarantee that none of the world records were set by people who don’t use drugs. However, this problem does not plague this data set to nearly the same degree as the data set in powerlifting.
So, on the whole, the world records in weightlifting likely approximate the true limits of performance with the aid of drugs, and the American records are somewhat below the true limits of performance without the aid of drugs.
Put all of that together, and you can assume the effect of drugs on powerlifting performance is likely a bit larger than the 4.88-9.11% observed when comparing drug-tested and untested world records, and the effect of drugs on weightlifting performance is likely a bit smaller than the 11.66-13.85% observed when comparing the world records to the American records.
Or, in other words, a 10% difference is pretty reasonable.
In fact, that’s about what you should expect. The reasons why are the topic of the next article.
If a comparison of world records doesn’t cut it for you, we can look at experimental evidence instead.
| Bashin, 1996 |
| Drug-free | 600mg Test per week |
| Before | After | Increase | Before | After | Increase |
| Body Mass (kg) | 85.5 | 86.4 | 1.05% | 76 | 82 | 7.89% |
| Bench (kg) | 109 | 119 | 9.17% | 97 | 119 | 22.68% |
| Squat (kg) | 126 | 151 | 19.84% | 102 | 140 | 37.25% |
| Triceps area (mm^2) | 4052 | 4105 | 1.31% | 3483 | 3984 | 14.38% |
| Quadriceps area (mm^2) | 9920 | 10454 | 5.38% | 8550 | 9724 | 13.73% |
| KG fat free mass per cm | 0.407 | 0.418 | 2.7% | 0.372 | 0.407 | 9.41% |
| Bench Wilks | 71.53 | 77.62 | 8.51% | 68.49 | 80.02 | 16.83% |
| Squat Wilks | 82.68 | 98.5 | 19.13% | 72.2 | 94.14 | 30.39% |
| BP+SQ Wilks | 154.21 | 176.12 | 14.21% | 140.69 | 174.16 | 23.79% |
Here, again, steroids helped the participants gain more muscle, faster, but Wilks Score only improved by 9.58% more for the steroid users. The
Wilks Score is based on a formula that was developed to accurately compare powerlifting performances, since neither absolute strength nor a simple strength/bodyweight ratio give an entirely fair comparison (more on that in the next installment of this series). Put another way, the drug-free lifters were 8.77% better lifters at the start of the study (154.21 vs. 140.69), and only 1.11% better lifters by the end (176.12 vs. 174.16).
| Bashin, 2001 |
| 125mg Test per week | 300 or 600mg Test per week |
| Before | After | Increase | Before | After | Increase |
| Body Mass (kg) | 75.3 | 78.3 | 3.98% | 73.45 | 79.01 | 7.56% |
| Leg press (kg) | 419.2 | 444.6 | 6.06% | 435.7 | 516.8 | 18.61% |
| KG fat free mass per cm | 0.338 | 0.354 | 4.73% | 0.3305 | 0.371 | 12.25% |
| Thigh muscle volume (cm^3) | 890 | 966 | 8.54% | 825.5 | 930.5 | 12.72% |
| Quadriceps volume (cm^3) | 508 | 546 | 7.48% | 484.5 | 540 | 11.46% |
| Leg Press Wilks | 297.88 | 307.75 | 3.31% | 315.24 | 355.655 | 12.82% |
Here I compared the group getting 125mg of Testosterone per week (keeping them with basically the same Test levels they started with) to the two groups receiving supraphysiological doses (the 125mg group’s test levels were about 10% lower than they were at the start of the study, while the 300mg group’s were elevated roughly two-fold, and the 600mg group’s were elevated roughly four-fold).
Again, the two groups with elevated Testosterone gained a lot more muscle and fat free mass than the group with normal Test levels, but the Wilks Score tells the same story as the previous study (I’m sure the powerlifting gods are frowning on me for using Wilks to compare leg press performance, but YOLO). The two groups with elevated Test increased their Wilks Scores by 40.4 points (12.82%), vs. 9.87 points (3.31%) for the group with normal levels, meaning the groups with elevated Test got a 9.51% advantage. That sounds eerily similar to the ~10% spread in the world records, and the 9.58% advantage from the other Bashin study, now doesn’t it?
Now, you may be saying to yourself, “But aren’t those pretty low doses? People in the real world use more Test than that (250-500mg/week is a pretty standard ‘first cycle,’ generally adding more Test and other anabolic compounds for subsequent cycles), and these studies were only 10 and 20 weeks long. Maybe higher doses for longer periods of time would yield a larger advantage.”
I understand that line of thought, but there’s not currently any reason to think it does. In fact, the opposite may be true.
For starters, grouping the 300mg and the 600mg groups together in the
2001 Bashin study actually did a favor for the group taking the higher dose of 600mg per week. Their Wilks Score only improved by about 10%, versus about 15% for the 300mg group. The 600mg group DID gain more muscle and overall mass, but they only gained about the same amount of strength (a bit less, actually), so their performance relative to their size – which is what matters for Wilks Score – didn’t increase as much as the 300mg group’s did.
We also have another
study to take into consideration:
| Yu, 2014 |
| Steroids | Drug-Free | Difference (positive means higher for Drug-Free) |
| Body Mass (kg) | 108 | 110 | 1.85% |
| Fat Free Mass (kg) | 89.8 | 74.6 | -16.93% |
| Squat PR (kg) | 254 | 265 | 4.33% |
| Squat Wilks | 150.34 | 155.95 | 3.73% |
| Squat/FFM | 2.83 | 3.55 | 25.59% |
| Bench PR (kg) | 205 | 190 | -7.32% |
| Bench Wilks | 121.34 | 111.81 | -7.85% |
| Bench/FFM | 2.28 | 2.55 | 11.57% |
| Deadlift PR (kg) | 257 | 269 | 4.67% |
| Deadlift Wilks | 152.12 | 158.31 | 4.07% |
| Deadlift/FFM | 2.86 | 3.61 | 26% |
| Total | 716 | 724 | 1.12% |
| Total Wilks | 423.8 | 426.07 | 0.54% |
| Total/FFM | 7.97 | 9.71 | 21.72% |
This study didn’t involve an intervention, so we can’t infer cause and effect, but it still supplies some interesting data.
The researchers compared folks who had been using steroids for 5+ years to people who had never used. Of course, we can’t be sure the drug-free group had, in fact, never used steroids but, from the study: “Clean subjects had signed a contract with their local clubs and the Swedish Power Lifting Federation, committing them to never use any drugs, under severe monetary punishment. The subjects have been continuously doping-tested with negative results.”
Another key limitation was that all the drug-free subjects were powerlifters, and the subjects on steroids were a mix of powerlifters, bodybuilders, and strongmen, so assessing strength via the squat, bench press, and deadlift favored the drug-free lifters (and probably hints that training specificity matters more than drugs do). However, they also assessed strength via maximal isometric force in the squat with the knees at a 105-degree angle (above parallel); presumably the strongmen were squatting heavy in their training, and even if the bodybuilders squatted high in training, the position they measured from was basically a half squat, so that should be a pretty fair comparison. In that test, the drug-free lifters clobbered the guys on steroids. Maximum squat force was 36.7% higher in absolute terms (3302N vs 2416N), maximum squat force per kg of lean body mass was 68.8% higher (49.8N/kg vs. 29.5N/kg), and maximum squat force per kg of lean leg mass was 47.7% higher (130N/kg vs. 88N/kg).
Then, looking at squat, bench, and deadlift, the drug-free lifters’ overall Wilks Score was slightly higher, and their absolute performance (at roughly the same body weight) was better for both squat and deadlift. The bench PRs for the steroid users were higher, as you’d expect, both from the powerlifting world records (with a larger spread in the bench press than in the squat or deadlift) and because steroids generally affect the muscles of the
shoulder girdle moreso than the muscles of the rest of the body.
And, if you think the problem with Bashin’s study was low doses, check out the cycles of the guys in
Yu’s study:
Here’s the article in a single graph (click the image to enlarge it):
When you look at the data from these five sources (the best I could find), the overall average advantage afforded by steroids is 6.73%. That jumps to 9.22% if you disregard the data from
Yu, due to the training differences between the groups, and 10.97% if you toss out the powerlifting world records due to the heterogeneity of the data set. In other words, steroids give you a ~10% advantage in strength sports. They give you a larger boost in absolute performance, but with the extra muscle mass comes the
expectation of being able to lift more, in order to be equally competitive.
So, here’s a quick recap to wrap things up:
- Steroids help you get bigger and stronger, faster.
- Both observational evidence (comparing records in powerlifters and weightlifters) and experimental evidence suggest that the advantage you get from steroids is quite large in terms of absolute measures (total muscle and strength gained), but that they only confer a ~10% advantage in terms of how competitive you’ll be in strength sports. As you gain more muscle mass, you’re expected to lift more to be equally competitive in your sport. When taking into account both gains in strength and muscle mass, the performance edge they give you is about 10%.
- Higher doses for longer periods of time don’t seem to provide an increasingly larger advantage. The relative performance improvement in Bashin’s 20-week study was identical to the performance improvement in his 10-week study. Additionally, in his 20-week study, the group using 300mg of Test per week actually improved their Wilks Score more than the group using 600mg of Test per week. Furthermore, in Yu’s study comparing drug-free lifters to lifters who had been using much higher doses for 5+ years, the advantage of steroid use for improving relative performance disappeared entirely, though there were some important drawbacks to that study.
- The advantage steroids give you is likely larger for tests of upper body strength, such as the bench press. Though Bashin’s 1996 study actually found a larger advantage in the squat than the bench press, the world records in powerlifting, Yu’s study, and insights into the muscle groups most affected by steroid use all seem to suggest that the bench is the lift that benefits the most from steroid use.
Practical takeaway: If you want to compare the performance of someone who uses steroids to someone who doesn’t, a 10% adjustment will give you a pretty accurate comparison. In other words, if two lifters are in the same weight class, and the one on steroids totals 10% more than the one who’s drug-free, the steroids likely explain the gap in performance, and those two lifters would be pretty evenly matched if they both used steroids or they were both drug-free. If two lifters are in different weight classes, you can make the same comparison using Wilks or Sinclair scores with a 10% adjustment. If the gap is smaller than 10%, the drug-free lifter would likely be the better lifter if both used steroids or if both were drug-free, and if the gap is larger than 10%, the lifter on steroids would likely still perform better if both athletes used steroids or if both athletes were drug-free.
Steroids do help someone become a more competitive lifter, but it’s not a night and day difference. A 10% advantage is huge when you’re talking about elite-level competition (if someone runs 10.7 100m sprint when Usain Bolt is winning the meet at 9.7, they got destroyed). However, drugs aren’t the difference between, say, a 1500 total and a 2000 total in the same weight class in powerlifting. They’ll make a decent lifter an above average lifter, an above average lifter a good lifter, a good lifter a great lifter, and a great lifter one of the best in the world. But they aren’t going to a magically turn a 300 Wilks into a 500 Wilks. The next article will discuss why you should expect the difference to be relatively small. Give this article a share, and stay tuned. The next installment should be out within the next week or two.
