The Starting Strength Training Registry

by John Petrizzo, DPT, SSC | January 30, 2019

john petrizzo teaches at a camp

In October of 2018, an article entitled “A Feasibility Study Investigating the Sustainability and Safety of a Non-Periodized Protocol with Linear Load Progression during the Initial 12 Weeks of Strength Training” was published in the Journal of Exercise Physiology online. I know, it is a mouthful. That article is actually the culmination of approximately four years of work that began with the idea for the “Starting Strength Training Registry” (SSTR) which was born at the 2014 Starting Strength Coaches Association Conference.

Initially, we all had lofty aspirations for this project with ideas floating around in our heads that we would collect data on hundreds, if not thousands, of subjects and get an article (or articles) published in a highly regarded peer-reviewed journal like JAMA (Journal of the American Medical Association). If you are at all familiar with the article referenced above, then you already know that we fell well short on both of those goals.

The stated purpose of the SSTR was to investigate the effects of a free weight resistance training program on healthy participants between the ages of 18 and 80 by creating an anonymous registry of individuals who were voluntarily partaking in a formalized version of the “Starting Strength Novice Linear Progression” (SSNLP).

The data that we collected included the following: age, sex, height, initial and final body weight, initial and final waist and hip measurements, initial and final training loads for the squat, press, deadlift, and bench press, as well as whether or not the subject was able to complete 12 weeks of the LP. Additionally, for subjects who did not complete the full 12 week protocol, we wanted to know why they stopped. It is important to note that subjects were never asked for any identifying personal information such as their names, email addresses, etc.

To be clear, while our idea was to create an anonymous online registry for people interested in performing the SSNLP, the protocol that we used to carry out our study had some modifications from the original version of the program as detailed in Starting Strength: Basic Barbell Training and Practical Programming for Strength Training. Some of the differences between our protocol and the original program are as follows: our protocol was limited to 12 weeks, chin-ups and power cleans were both excluded in order to lower the barrier for inclusion, subjects were allowed to train either 2 or 3 days per week, and there was no inclusion of a “light day” for the squat as subjects progressed through the program. In terms of the actual layout of the program itself, similar to the SSNLP, the squat was performed at every session, and the press and the bench press were alternated at each session. All three exercises were performed for 3 sets of 5 reps following several progressively heavier warm-up sets. The deadlift was still only performed for a single set of 5 reps following warm-ups. However, unlike the SSNLP which initially has trainees deadlifting at every session and then transitioning to an alternating schedule with power cleans and then eventually with chin-ups, subjects who participated in our study were only required to perform the deadlift a minimum of once per week.

Ultimately, over the course of a couple of years, we ended up collecting date on a total of 79 subjects, 45 males and 34 females. The average age of our subjects was about 26 years old (25.9 + 9.6 years old) with our youngest participants being 18 and our oldest being 60 years old at the time of data collection.

We found that a total of 52.6% of all subjects completed the full 12 weeks of the protocol with the average duration of participation from subjects who did not complete the entire 12 weeks being a little over 9 weeks (9.1 + 1.8 weeks). During follow-up, the most common reasons cited for discontinuing the program were an inability to continue to make linear progress on the squat as well as an inability to train the minimally required twice per week. No subjects reported suffering an acute injury during the training period, but we did have one subject drop out at the 5 week mark due to complaints of elbow tendinitis.

Table 1 contains the anthropometric data of the entire subject pool (male and female):

Table 1: Aggregate Anthropometric Data:


Height (inches)

BW (pounds)

Waist (inches)

Hip (inches)


67.8 + 4.0

166.0 + 40.0

30.9 + 5.3

38.5 + 3.4



171.3 + 41.8

30.7 + 5.9

38.6 + 3.5

* Please note that height, waist, and hip measurements are reported in inches and body weight is reported in pounds

When analyzing the results, we found a statistically significant increase in body weight across all subjects. However, when we looked at the male and female data individually, we saw that the increase was due entirely to an increase in the body weight of our male subjects.

Table 2: Male and Female Body Weight:


Male BW

Female BW


187.4 + 37.3

138.3 + 24.0


196.3 + 35.2

138.9 + 24.7

* Please note that body weight is reported in pounds

As a group, there were no statistically significant changes in either waist or hip measurements, but when we analyzed the male and female data individually, we again saw a difference between the sexes.

Table 3: Male and Female Waist and Hip Measurements:


Male Waist

Female Waist

Male Hip

Female Hip


33.5 + 5.1

27.7 + 3.6

39.4 + 3.3

37.6 + 3.4


33.9 + 4.7

27.0 + 3.4

39.9 + 3.2

37.1 + 3.3

* Please note that waist and hip measurements are reported in inches

From the table above, we can see that male waist and hip measurements both increased approximately half an inch on average while female waist and hip measurements both decreased half an inch or so on average. When looking at the changes seen in the subject’s anthropometric data, it is important to note that subjects were not given any sort of nutritional counseling and were asked not to make any significant changes to their normal diet while carrying out the resistance training protocol.

Not surprisingly, we saw statistically significant increases in strength across all lifts both when analyzing the data in aggregate and by sex.

Table 4: Aggregate Strength Data:





Bench Press


127.0 + 71.0

65.7 + 27.5

153.5 + 78.5

106.7 + 58.3


219.7 + 97.3

101.7 + 46.1

242.5 + 113.8

147.9 + 72.8

*Please note that all strength data is reported in pounds

Table 5: Male Strength Data:


M Squat

M Press

M Deadlift

M Bench Press


167.2 + 65.4

82.2 + 24.2

197.0 + 75.4

142.8 + 52.3


281.7 + 78.3

132.0 + 37.5

314.1 + 94.4

199.4 + 52.1

*Please note that all strength data is reported in pounds

Table 6: Female Strength Data:


F Squat

F Press

F Deadlift

F Bench Press


73.7 + 32.7

43.8 + 11.9

96.0 + 31.9

58.8 + 15.9


137.5 + 44.8

61.5 + 15.4

147.9 + 50.2

79.9 + 23.3

*Please note that all strength data is reported in pounds

In aggregate, we saw a 72.9% increase in squat strength, a 54.7% increase in press strength, a 58% increase in deadlift strength, and a 38.6% increase in bench press strength.

Overall, we were pretty pleased with these numbers. Now, all that was left to be done was to write up a paper for submission to a peer-reviewed journal. My original idea for the manuscript was to do a literature review and examine the data of previous studies that used a “non-periodized” resistance training program in hopes of highlighting that no previous studies used a non-periodized resistance training protocol that was comprised entirely of compound barbell exercises as well as the fact that no previous studies utilized a loading strategy that required the subjects to increase the load used across all exercises at every session.

In conducting my review of the literature, I found about a half dozen studies from the mid-1990s to the early 2000s that utilized a non-periodized resistance training protocol, for approximately 12 weeks in length, and incorporated at least one of the basic barbell exercises used in our study [1–6]. At the time that these studies were conducted, they were used to help support the notion that periodized resistance training programs are superior to non-periodized programs across a broad spectrum of trainees. With that being said, most of these studies also have very small sample sizes, do not clearly identify their subject population (i.e. what constitutes a “trained” subject), and do not clearly identify the loading scheme used in their “non-periodized” resistance training protocols. Additionally, when comparing the results of our data versus that of the previously performed studies, we saw pretty significant differences in terms of changes in strength.

For example, Baker, Herrick and Stone published a study in 1996 in the Journal of Strength and Conditioning Research that utilized untrained women for their subject pool. They had a total of 10 untrained women in their non-periodized group training 2x/week for 15 weeks. 1RMs were tested every three weeks throughout the data collection period. Their untrained female subjects increased their squat and bench press 1RM by 46.3% and 25.2% respectively. However, when compared to our data set that included 34 untrained females, we saw an 86.0% and 36.6% improvement in their 5 rep squat and bench press over the course of their participation in our program.

Similarly, another study that was published in the Journal of Strength and Conditioning Research in 1997 by Kramer, et al. examined the differences between not only variations in intensity and volume, but also single versus multiple sets. This study included 43 moderately trained college-aged males and had them train 3x/week for 14 weeks with the squat being performed twice weekly. The authors broke the subjects up to three groups. The first performed a single set to momentary muscular fatigue, the second performed a multi-set protocol that called for 3 sets of 10 reps to be performed at each session, and a third group performed a varied set and rep scheme. The three groups improved their 1RM squat strength anywhere from 12.0% - 25.6% over the course of 14 weeks. In comparison, our 45 male subjects improved their squat strength by 68.4% over a fairly similar time period.

Another study that involved the use of untrained female subjects was performed by Marx, et al. and published in Medicine & Science in Sports & Exercise in 2001. In this study, 34 untrained female subjects trained 3-4x/week for 24 weeks with a 1RM bench press being measured at weeks 1, 12, and 24. The subjects were again divided into three groups. The first group performed a single set of 8-12 reps to momentary muscular fatigue 3x/week, the second group performed 2-4 sets of 3-15 reps 4x/week, and the third group was a control group. At the 12 week mark, 1RM bench press improved 12.2% for the single set group, and 23.4% for the multi-set group. However, the 34 untrained subjects who contributed to our data set improved their bench press strength 36.6% over the course of the same time period.

These are just a few of the studies that I found, but I believe they are the most suitable for comparison to our data set since they involved the use of untrained or only moderately-trained subjects. While they do not make for perfect comparisons to our data, all other studies I found that would be suitable for comparison all claimed to use “trained” males [1,5,6]. Regardless, I think these examples are enough to provide the reader with an example of the quality of the research that is out there on this particular topic.

With all that being said, once I had analyzed the data and put together an initial draft of the manuscript, I realized that getting this paper published in the form that I had hoped was going to be a much bigger challenge than I anticipated. Admittedly, most of the blame for that falls squarely on me.

The first issue that arose during the review process was that I had a hard time convincing readers that there was anything novel about our protocol or results. Essentially, the thought process was that all resistance training programs result in improvements in strength, especially in novices, so why should our paper get published? Apparently, the fact that we used a protocol and loading scheme that was unique to the literature and that we obtained greater improvements in strength than any other paper I could find that utilized a non-periodized protocol was not convincing enough.

The other problems that we had during the review process were mainly due to faults in our study design and execution. Some of the other issues that were brought up were the lack of a control group for comparison, the fact that the squat was the only exercise performed at every session, the lack in uniformity in training days per week, and finally, the fact that there were not uniform load increases for all subjects throughout the training period. Our thought process behind these decisions, while well intentioned and an attempt to foster greater subject compliance and applicability to the real world, ultimately caused tremendous difficulty in getting our paper published and lead to the paper having to be significantly re-written, with an entirely new statistical analysis that left out nearly all the data presented above.

In the end, my goal in writing this article was to clarify questions that have come up about the peer-reviewed paper that was published last October and to get the rest of the data out there for the public to see. While I am disappointed that all of our hard work was not initially received the way I had hoped it would be, in my opinion, the results of this data set clearly show that a barbell-based, linearly progressed resistance training program such as the SSNLP is a safe and effective way for a novice to quickly improve their strength and positively impact their anthropometric characteristics during the first several months of their training career. It is my hope that the data presented above as well as the initial paper published last October will be just the beginning of a growing body of research that supports the use of progressively loaded barbell exercises to improve a wide array of fitness and health related characteristics across a broad spectrum of the population.


Discuss in Forums

Starting Strength Weekly Report

Highlights from the StartingStrength Community. Browse archives.

Your subscription could not be saved. Please try again.
Your subscription has been successful.