Are rows or pull-downs better for back width?
It is practically written in stone in bodybuilding lore that rows develop back thickness and pull-ups develop back width. However, the truth may be more complicated than that, as the following study illustrates.
The study: Variations in muscle activation levels during traditional latissimus dorsi weight training exercises: An experimental study. By Lehman, Buchan, Lundy, Myers and Nalborczyk, in Dynamic Medicine, 2004
What is the background?
Various exercises are used to develop the back muscles. Two of the most popular machine-based exercises are the lat pull-down and the seated row. Many experienced weight lifters believe that using these machines in slightly different ways can target different muscle groups.
For example, it is generally believed that using a pronated grip for the lat pull-down preferentially recruits the lats over the biceps, while using a supinated grip preferentially recruits the biceps over the lats.
Similarly, it is also believed that the seated row develops the middle trapezius/rhomboids more than the lat pull-down, while the lat pull-down is thought to recruit the latissimus more than the seated row.
Hence bodybuilders often refer to the vertical pulling exercises like the lat pull-down as exercises for “back width”, referring to the effect of the lats in widening the physique. On the other hand, they refer to the row as building “back thickness”, referring to the effect of the mid-back in creating greater size when examining the physique from the side view.
But it’s not just the bodybuilders that have opinions about back muscles. Researchers have also proposed that rows with retracted scapulae may produce different activity in the middle trapezius and rhomboids than rows where protraction is allowed. Whether this is the case, however, is not known.
What studies have been published before?
Muscle activity of the latissimus dorsi
Signoreli (2002) investigated the influence of grip width and type of bar used during the lat pull-down on the muscular activity of the posterior deltoids, latissimus dorsi, pectoralis major, teres major and long head of the triceps. Signoreli found that the wide-grip, pronated pull-down produced greater muscle activity than pull-downs using a supinated grip. They also found that pull-downs anterior to the head produced greater activity of the latissimus than pull-downs to the rear of the head.
Muscle activity of the middle trapezius and rhomboids
Ekstrom (2003) investigated middle trapezius activity during various strength training exercises. They found that the single-arm row activated the middle trapezius to 79% of MVC but did not comment on whether this was performed with a protracted or retracted scapula.
What did the researchers do?
The researchers recruited 12 healthy males who had at least 6 months of weight training experience. They asked the subjects to perform four different exercises: a lat pull-down with pronated grip, a lat pull-down with supinated grip, a seated row with scapulae protraction and a seated row with scapulae retraction. The subjects were asked to perform the exercises with their estimated 10 – 12RM in each case.
However, the exercises were not performed dynamically. Instead, they were performed as 10-second isometric contractions. In the case of the pull-downs, they were held at eye-level. In the case of the rows, they were held at 90 degrees of elbow flexion and 0 degrees of shoulder flexion.
In order to make good sense of the EMG activity data that researchers gather, they very frequently use a normalization process. The most commonly used normalization process is to express the size of the EMG signal as a percentage of a maximal voluntary isometric contraction (MVC).
Clearly, when using an MVC, the way in which the contraction is performed is very important, as if the contraction measured as the MVC is not actually maximal, then the percentage of MVC that is calculated in the study will appear oddly high. Various factors affect the force that a muscle is able to produce but none are so important as muscle length.
The length-tension relationship states that muscles produce maximal force a two points: (1) at an optimal length that is neither very long nor very short, and (2) at very long lengths because of the passive elastic elements. Muscle length is defined in the human body by joint angle. Consequently, the joint angles at which MVCs are performed is important for making sense of a study and comparing to other studies. Such joint angles are often referred to as “MVC positions” or the positions in which the MVC was performed. In this study, the researchers used MVCs.
The MVC position for the latissimus dorsi, was a lat pull down against an immovable resistance, although the researchers do not specify which hand position they used, which grip width or at what angle the shoulder and elbow were at during the MVC. For the biceps, they explained that they used a position with the elbow in 90 degrees of flexion. For the middle-trapezius and rhomboids, the researchers used scapular retraction against manual resistance.
The main result that the researchers found was that the EMG activity of the latissimus dorsi was actually higher during the seated rows than during the pulldowns, as can be seen in the chart below:
The researchers note that while there was a trend for the retracted scapulae seated row to demonstrate greater latissimus and middle trapezius/rhomboid activity than the protracted seated row, these results were not statistically significant.
The researchers reported that, counter to the results reported by Signoreli (2002), they did not observe any significant differences between supinated and pronated pull-downs for latissimus activity. They noted that there was only a slight trend for the pronated grip pull-down to display higher latissimus dorsi activity than the supinated grip pull-down. Finally, they noted that there was no difference in biceps brachii activity between any of the exercises.
Ratios of muscle activity
However, when comparing the ratios of muscle activity, the researchers noted that there was a significant difference in the ratio of latissimus-to-biceps activity between the pronated and supinated pull-downs. They also noted that there was a significant difference in the ratio of latissimus-to-biceps activity between the seated row with scapular retraction and the supinated pull-downs, as you can see in the chart below:
This chart shows that while there may not have been a significant difference in muscle activity between the supinated grip pull-down and the other exercises, it clearly recruits the biceps muscle proportionally more than the others. Interestingly, the other ratios involving the other exercises are all very similar.
What did the researchers conclude?
Differences between pull-downs
The researchers concluded that the pronated grip lat pulldown demonstrated a small and non-significant increase in the activity of the latissimus dorsi compared with the supinated grip pulldown while the supinated grip pull-down displayed a small bt non-significant increase in the activity of the biceps muscle.
Comparison between pull-downs and rows
The researchers also concluded that the seated row actually recruits the latissimus more than pull-downs. And it also has the benefit of recruiting the middle-trapezius and rhomboids more than pull-downs as well. So while rows certainly do appear to develop back thickness better than pull-downs, they may well also develop back width better too…
Differences between protracted and retracted rows
The researchers concluded that the seated row with retraction demonstrated a small and non-significant increase in the activity of the latissimus dorsi and of the middle trapeizus/rhomboids compared with the seated row with protraction. Retraction of the shoulder blades during rows may well therefore optimize the benefits of this movement.
What were the limitations?
The researchers noted that their study was limited in that a very low weight was used (less than 40% of MVC) and this may have had significant effects on the relative activity of the various muscles. For example, load changes the relative involvement of the hip extensors and knee extensors in squats and it is anticipated that muscle activity also changes as a result. The same thing could occur with back exercises.
An additional limitation of the study (and a likely contributor to the low percentage of MVC) was the fact that the exercises were performed as isometric contractions and not dynamic movements. Isometric contractions that are set at specific loads (i.e. not maximal contractions) are easier than dynamic contractions for the muscles to perform, which may result in lower activity levels.
Also, isometric contractions can produce lower-than-maximal muscular activity if the joint angle is not optimal. Therefore, if the position selected by the researchers is not the point in the movement at which maximal activity is reached, a dynamic exercise with the same movement could easily produce more EMG activity and this could differ between the pull-down and the row.
Another serious limitation was that the subjects themselves selected a 10 – 12RM for each exercise. This means that if the subjects picked harder weights for the pull-downs and easier weights for the rows (for example), this would lead to overstated relative activity for the pull-downs compared to the rows.
Finally, it is important to note that the researchers did not investigate the other muscles involved in these exercises. For example, there may well be upper trapezius and/or posterior deltoid differences between rows and pull-downs that are important to understand for weight lifters.
What were the key points?
While this study had some large limitations, we can tentatively suggest that:
- Pronated and supinated pull-downs (and possibly also therefore chins and pull-ups) only differ slightly in respect of the relative activity of the latissimus and biceps muscles. These differences are therefore unlikely to make a significant difference to the beginner or intermediate trainee.
- Seated rows produce greater latissimus dorsi and middle trapezius/rhomboid activity than pull-downs of any kind.
- Seated rows tend to produce greater latissimus dorsi and middle trapezius/rhomboid activity when performed with scapular retraction.
What are the practical implications?
For bodybuilders and physique athletes
For someone concerned primarily with physique development, it probably makes sense to build a back routine around rows rather than lat pull-downs.
There is no logic in splitting a back routine into “back width” and “back thickness” days, if this is done based on vertical pull-downs being for back width and rows being for back thickness, as rows produce more activity in both the latissimus and in the mid-back musculature.
For everyone performing strength training
When performing seated rows, make sure to perform scapular retraction at the end of each repetition to improve muscle activity. It’s very likely good for your shoulder health to improve range-of-motion anyway.
Whether you use chins or pull-ups is likely not going to make or break your training routine, so you should probably go with what you prefer or mix it up to prevent things getting stale. If pushed, the chin is probably slightly better because more weight can be used.
For more information just like this, please subscribe to our monthly research review service. For only $10 per month, you get between 45 – 50 study reviews every month, delivered directly to your inbox…