Cues Gone Wrong-“Shoulders Down and Back”

Cues Gone Wrong-“Shoulders Down and Back”

“Shoulders down and back”. This cue is practically gospel in the world of physical therapy. And it has seeped into the world of fitness as the “safest position” for the shoulders.

Unfortunately, this cue is overused and creates confusion when misapplied.

While shadowing as a PT student, I recall watching a PT cue a patient doing dumbbell overhead presses. They kept telling the patient to keep their shoulders “down and back”. The patient struggled and struggled. Over the next few weeks the patient would lament, “I just can’t seem to get it”. At the time I thought the patient just needed more practice. I’ve since learned that the patient couldn’t keep their shoulders down and back while overhead pressing, because that’s not what the shoulders do in an overhead press.  

In this article I’ll discuss the use (and misuse) of this cue in pressing exercises.

What is this cue intended to do?

This cue is intended to create scapular retraction (squeezing the shoulder blades towards the spine) and depression (bringing the shoulder blades towards the pockets, away from the ears).

When does this cue apply?

The bench press and its variants.

In the bench press creating a back arch allows the user to lift more weight. This back arch creates various mechanical efficiencies: a smaller range of motion (ROM) for the bar, improved leverage in the lift1, and “tightness” or stability that allows the user to generate more force.

Shoulders down and back can help the user to create that arch, thereby allowing them to lift more weight.

When does this cue NOT apply?

The overhead press.

Trying to keep the shoulders down and back when lifting overhead is contradictory to the motion of the scapula.

When the arm lifts overhead, the scapula rotates upwards. It is estimated that 1/3 of the range of motion lifting overhead comes from this scapular upward rotation2.

Further, this scapular upward rotation is created by the synergistic actions of the trapezius muscles and the serratus anterior2 . So using the upper trapezius is actually necessary to lift overhead with full ROM.

Figure 2: Scapular Upward Rotation

As a self-experiment, lift your arm overhead. Then pin your shoulders “down and back” and try lifting again. There will be significantly less ROM because pinning the shoulders down and back limits scapular movement, which limits full shoulder ROM.

The pushup is another exercise where the shoulders should not be held down and back. During the upward, concentric phase of the pushup, the scapula protract (move away from the spine, see figure 1 above). That protraction is accomplished mostly with the serratus anterior. Trying to cue shoulders down and back is contradictory to that “natural motion” of the pushup.

Why do the pushup and bench press have different cues?

First, the bench press is derived from the sport of powerlifting where the goal is to lift the maximum amount of weight. As discussed above, arching allows the user to lift more weight.

Second, in a bench press the support from the bench gives the scapulae a surface to brace against. In the pushup, there is no external stability, so the serratus anterior has to create the platform for pushing.

Cues have a time and place. The right cue applied to the wrong lift creates confusion and frustration. “Shoulders down and back” has its place in the bench press, but generally should not be encouraged in the overhead press or pushup.


  1. Rippetoe, M., & Kilgore, L. (2011). Starting strength: basic barbell training. 3rd ed. Wichita Falls, TX: Aasgaard Co.

2. Levangie, P. K., & Norkin, C. C. (2005). Joint structure and function: A comprehensive analysis. Philadelphia, PA: F.A. Davis Co.

Which Measurements Matter to Patients?

“When faced with a difficult question, we often answer an easier one instead, usually without noticing the substitution.” -Daniel Kahneman

As physical therapists we have many tools to track patient progress; outcome surveys, range of motion measurements, etc. However, which of these measurements meaningfully reflect patient outcomes?

This might seem obvious. But we often forget to ask these big picture questions in the clinic. We also forget such questions when reading research or choosing con-ed courses.

To pick the right measurements, we need to understand “clinical endpoints“. This term comes from clinical medicine trials. A clinical endpoint is “an event or outcome that can be measured objectively to determine whether the intervention being studied is beneficial”.1

There are three types of clinical endpoints2:

1. Direct clinical endpoints
These directly relate to patient outcomes. For example, consider a soccer player with knee pain. A direct clinical end point would be their ability to play soccer with an acceptable level of pain. We could measure that with the Victorian Institute of Sport Assessment–Patella (VISA-P). An improvement in the VISA-P likely means an improved ability to play soccer with an acceptable level of pain.

Of course, direct clinical endpoints are not the only meaningful measurements. But they most directly relate to patient outcomes.

2. Surrogate endpoints
These predict patient outcomes. We use these if we can’t directly measure an outcome. For example, we don’t know whether an elderly patient will fall – until they fall. But a measure like the modified 30-second Sit to Stand (m30STS) predicts fall risk.3 If we improve the patient’s m30STS score, they should be at a lower risk of falling.

A surrogate endpoint should be validated. Meaning that data supports its ability to predict an outcome.2

However, some surrogate endpoints are not validated. Such endpoints theoretically relate to an outcome. But data doesn’t show that connection. For example, consider the elderly patient at risk of falling. Low calf muscle strength might correlate with falls. However, there are many other factors related to fall risk, so we can’t predict fall risk with calf muscle strength alone.

3. Biomarkers
These correlate with biological activity in a patient. They are least relevant to patient outcomes. For example, a PT might use a tool shown to “increase blood flow” in a patient with low back pain. “Increased blood flow” is a biomarker. This biomarker might correlate with tissue healing. And tissue healing might decrease low back pain. But this is a weak connection. The increased blood flow is an interesting effect, but may not be relevant to improving low back pain.

As we see, clinical endpoints are not all equally relevant. Understanding them is crucial to interpreting research findings, picking con-ed courses, and tracking patient progress. The goal is to not forget the goal: getting patients back to activities they love.


  1. NCI Dictionary of Cancer Terms. (n.d.). Retrieved November 13, 2020, from
  2. Sullivan, E. (n.d.). Clinical Trial Endpoints [Powerpoint slides]. Retrieved November 13, 2020 from
  3. Applebaum, E. V., Breton, D., Feng, Z. W., Ta, A. T., Walsh, K., Chassé, K., & Robbins, S. M. (2017). Modified 30-second Sit to Stand test predicts falls in a cohort of institutionalized older veterans. PloS one12(5), e0176946.

Is pain during rehab exercise “OK”?

*Medical Disclaimer: Please be advised, the information provided in this article is educational in nature and not meant to diagnose or treat any disease, illness, or condition. For individualized recommendations it is best to follow up with a licensed provider, like myself or another physical therapist.

A common concern among clinicians is when a patient experiences pain during exercise. Some clinicians rush to modify the exercise so that the patient has no pain whatsoever. Others ask the patient to describe the pain, trying to differentiate between muscle fatigue and nociceptive pain. And some say it’s just part of the rehab process.

Thankfully in the past couple of decades research has been building on this question. In 2017 Smith et al. did a systematic review trying to answer this question. Their review included studies that:

-Were randomized control trials that compared exercise where pain was allowed/tolerated vs exercise that was pain-free
-Had adults with chronic musculoskeletal pain (defined as >3 months)
-Measured pain, disability, or function

*Note that adults with “non-musculoskeletal pain” were excluded (this list ranged from cancer to headaches to fibromyalgia).

From their search, 7 studies made the cut: 1 trial on low back pain, 3 trials on shoulder pain, and 3 trials on foot/ankle pain. Patients in the studies were instructed that pain during exercise was acceptable (in some studies up to even 5/10 on the VAS), but that the pain should subside afterwards.

They then analyzed the effectiveness in the short (<3 months), medium (3-6 months), and long term (>12 months).

The results?

– Short term, patients who had pain during exercise had statistically significantly LOWER pain scores compared to pain-free exercise
-Reductions in pain levels were similar in the medium and long term between groups
-Improvements in disability and function were similar in the short, medium, and long term between groups

In short, for patients with chronic pain, pain-free exercises and exercises where pain was allowed resulted in similar outcomes (though short term, exercises where pain was allowed led to significantly decreased pain levels). However, we still don’t know about acute injury and post-surgical patients. We likely have to rely on clinical judgement for this. Also, the research doesn’t address other regions of the body (knee, elbow, neck, etc), but I think that we can reasonably generalize the results to other regions of the body-it is likely that chronic pain responds similarly to treatment, regardless of the region of the body. Lastly, a major limitation is that patients with conditions like fibromyalgia and migraine were excluded-this is a major group of patients who suffer from chronic pain! It would be helpful to examine the effects of painful exercise in these populations.

A few questions remain for me:
-Is there a ceiling on the “allowable pain” during exercise? Is there a point of “too much” pain?
-Related, is pain post-exercise “OK”? Would patients have better or worse outcomes if the pain from exercise lingered for a couple days?
-After discharge from therapy if a patient experiences pain with general exercise (a general training program), does this predict future pain?

Here is the original paper for review:

The Real Purpose of Health and Fitness Certifications

People often critique exam based certifications like the CSCS (Certified Strength and Conditioning Specialist) that lack a practical component. Critics say “you just need practical experience” and “it’s just a title”. I think these critics misunderstand the purpose of these certs. Here we’ll look at the purpose of these certs (next time we’ll talk about the limitations).

Exam based certs like the CSCS are a structured, standardized way to learn foundational, didactic knowledge in health and fitness. This serves several crucial roles:

1. Foundational knowledge makes practical experience more valuable

Imagine Jimmy, an aspiring strength and conditioning coach. Jimmy believes all that matters is practical experience. He walks into a gym and says “I want practical experience, can you mentor me?”. The coach says “Great Jimmy, come along and watch. Today my athlete is doing a squat workout; 3 sets of 5 at 80% 1 RM”. Jimmy is quite confused; what’s a squat? , what’s a set?, what’s a rep?, what’s an “RM”? The coach proceeds to define each term and explain basic principles of strength and conditioning…

As we can see without foundational knowledge, the valuable time of a practical experience would be frittered away passing on knowledge that the student could have learned on their own. The practical experience is the time to see foundational knowledge applied, not to learn it.  

2. Foundational knowledge is crucial to understand and apply scientific literature

Similar to practical experience, having foundational knowledge equips you to take advantage of scientific literature in the field. If you are reading a research paper but don’t know basic terms and principles, it would be nearly impossible to understand and apply that research. And with the plethora of health and fitness gurus promoting “cutting edge” workout routines and nutrition supplements, having foundational knowledge and reading research helps you verify claims.

3. Foundational knowledge lets you think for yourself (instead of just copying gurus)

Suppose a coach mostly does “fives” (sets of five) with athletes. Their apprentice who has no knowledge of loading parameters might just assume, “Well, in strength training we do ‘fives”. The apprentice who doesn’t understand the principles that inform that choice has to follow their guru’s advice. That person cannot adapt a training program to different athletes at different times. Ultimately, this person will be unable to modify and create programs.

Foundational knowledge is necessary to optimize practical experiences, use scientific literature, and think for yourself. It is necessary, but not sufficient. Of course, a certification is not the only way to learn and there are certifications that teach low quality information.

Having foundational knowledge in health and fitness doesn’t “make you a coach”, but it does set you up to become one.

Temporomandibular Joint Pain: Why, What, and How

The temporomandibular joint (TMJ) is a small joint that can be the source of major pain in patients. While some TMJ pain is directly from the joint itself, there are MANY other causes of pain in this region and there are many treatments available. This presentation was given to clinicians as a primer on fundamental anatomy of the joint and how to evaluate TMJ pain.