Some physicians condemn squat, citing how destructive they are to the tibiofemoral joints (knees), despite scientific studies and millions of personal experiences to the contrary. One sports medicine doctor dared to explain to me why squats were considered to be bad for the knees between his sets of squats! Since sports medicine doctors only see people with injuries, one can guess why they may have developed this belief. The individuals they treat certainly do not constitute a random sample, let alone a representative population, which, as any scientist knows, is essential to even attempt to formulate inferences.
The NSCA position statement notes
"Some reports of high injury rate may be based on biased samples. Others have attributed injuries to weight training, including squat, which could have been caused by other factors. Injuries attributed to squat may result not from the exercise itself, but from improper technique, pre-existing structural abnormalities, other physical activities, fatigue or excessive training."
An early study suggested, deep knee bends with weights (squats) were hazardous to ligamentous structures of the knee. Later studies conclude squats improve knee stability if lifting technique does not place rotary stresses on the knee (Fleck and Falkel, 1986). The NSCA state:
"Squats, when performed correctly and with appropriate supervision, are not only safe, but may be a significant deterrent to knee injuries."
The squat develops leg strength by imposing torques on the respective joints. Joint torques are the product of the magnitude of the force and the perpendicular distance from the line of force to the center of the joints.
Torque force is necessary for muscles and joint structures to adapt to respected overload. For example, if the knees do not travel forward during the barbell squat, the quadriceps muscles (ie: knee extensors) are not significantly loaded. On the other hand, an injury may result if knee (or low back) experience greater torque forces than to what they are accustomed.
Contrary to propaganda, prominent weight training authorities demonstrate squat with knees flexing forward at the same distance as hips flex backward. Fredrick Hatfield, Ph.D., the first man to squat over 1000 lbs, recommends knees to extend over feet with back more upright for quadriceps development. Strength Training for Young Athlete" by Steven J. Fleck, Ph.D. and William J. Kraemer, Ph.D., illustrate parallel squats with knees extending beyond feet (knees moving forward with similar magnitude as hips moving backward).
Fry (2003) examined hip and knee torque forces of variations of parallel barbels squats and concluded appropriate joint loading during this exercise may require knees to move slightly past the toes. '*Weight in line with mid-foot throughout squat diagram'
Palmitier (1991) showed that knee shear forces are less in the squat as compared to the seated knee extension. Closed chain exercises like the squat can be more protective than open-chain exercises like the knee extension because it reduces shear forces across the knee.
Typically the torque forces in the barbell squat are slightly greater for the stronger hip joints as compared to the knee joint, although both the knee and hips travel in the opposite direction away from the line of force. During the execution of a barbell squat, knees and hips travel in opposite directions away from the foot, or away from the center of gravity. Torque forces increase through the knee, hip, spine, and ankle as exerciser descends. The greatest torque forces are experienced concomitantly through the hip, spine, knee. and ankle when initiating the rise out of the bottom position.
Torque on the Spine is greater when the torso is at a more bent over angle. See how torso is positioned in a more bent over position during the low bar squat and in individuals with a greater femur to torso ratio.
Rotary forces through the knee can be caused by functional valgus, also known as 'knees caving in'. Valgus during heavy squats is most commonly caused by the Adductor Magnus overpowering the Gluteus Maximus, Medius, and Minimus in the lower concentric portion of the squat. The Adductor Magnus tends to adduct transversely as it extends the hips in but only in the lower concentric portion of the squat. The Gluteus Maximus tends to abduct the hip transversely. The gluteus muscles, particularly the Gluteus Medius, and Minimus normally counter and stabilize these forces as transverse abductors of the hip. Also, see Hip Abductor Weakness.
Valgus during squatting is generally not recommended to maintain joint integrity. However, it is interesting that some individuals seem to tolerate minor Valgus. Some high-level strength athletes such as Dan Green (Powerlifter), Ray Williams (WR Holder Powerlifter), Long Qiingquan, and other successful Chinese Olympic weightlifters regularly exhibit moderate degrees of Valgus during their maximal and near-maximal lifts.
Contreras (2015) theorizes that hip valgus may increase the moment arm of the Adductor Magnus through hip extension, thereby increasing hip extension torque. Contreras coined the term 'valgus twitch' to describe the tendency of the high-level lifter to utilize a quick 'knees in' movement when rising out of the hole during squats. (Contreras 2015)
The practice of adopting foot rotation to selectively strengthen individual muscles of quadriceps is not supported by literature (Boyden 2000; Signorile 1995). Knee rotation during squat can increase risk of injury (Fleck & Falkel 1986). Signorile (1995) states:
"Extreme outward toe point greatly reduces stability, it does not allow the proper drift of hips as lifter descends... Extreme inward toe points are equally dangerous, coupling same problems of stability, base size and lower body drift with added danger of bringing knees together...this movement would place high stress on all connective tissues."
The Quadriceps extend the knee. Its involvement is increased when the knee travels more forward relative to the ankle. Coactivation of the quadriceps and hamstrings occurs to increase knee stability by functionally reduce shear forces and strain across the knees. The three heads (of 4 heads) of the Hamstrings acting as Dynamic Stabilizers by virtually 'shortens' at the hip and simultaneously 'lengthen' at the knee (via quadriceps). When the knee is bent >90, the tension of the hamstrings help stabilize the knee by countering the anteriorly directed forces of the Quadriceps (dislocating force) on the knee (see Components of Force). The countering dislocating component with antagonist stabilizing component is characteristics of closed chained exercises (see Knee Stability). Also see simplified diagram of countering forces next to Hamstring Weakness and more detailed Quadriceps and Hamstrings countering forces in somewhat an opposing movement.
A wider stance increases Hip Adductor involvement but does not appear to alter Quadriceps or Biceps Femoris involvement (McCaw and Melrose 1999). The Gluteus Maximus involvement increases with a wider stance, but only at heavier load (eg: 75%). (McCaw and Melrose 1999). A closer stance targeting involves greater activation of the Gastrocnemius (Escamilla 2001).
The Spine is held ridged by the Erector Spinae acting as a Stabilizer with the Rectus Abdominis and Obliques acting as Antagonist Stabilizers countering the pull of the Erector Spinae. Erector Spinaes involvement is increased when spine is positioned at a greater forward angled.
Through the lower leg, Soleus Planter Flexes the ankle allowing the shin to become upright from the forward angled position at the bottom of the squat. Like the Quadriceps, its involvement is increased when the knee travels more forward relative to the ankle. The Gastrocnemius acts as a Dynamic Stabilizer, virtually 'shortened' through the ankle and 'lengthened' at the knee.
The mechanics of the squat change as the load increases. Hay (1982) examining the effects of varying squat load observed changes in the kinematics that effectively amounted to a technique change. At lower loads a larger moment production was required by the knee extensors (quadriceps). As the load increased (40, 60, and 80% of 1-RM) increase in forward inclination of the trunk was noted, shifting the damand to the hip extensors.
During the lower portions of a squat, the lumbar spine may flex if hip flexibility is limited. This posterior rotation of the pelvis is also known as butt wink.
It is thought that the risk of low back injury is increased during spinal flexion under a load, particularly if muscles of the lower back are not strong enough to support the flexed spine or if joint structures have not progressively adapted to such movement. Some suggest that the depth of the squat should stop just short of posterior pelvic rotation.
However, there is no evidence in the research that pelvic posterior rotation increases the risk of lumbar spine injury. In fact, many individuals exhibit butt wink with no lumbar pain or irritation, presumably adhering to Adaptation Criteria. During a loaded squat, there appears to be a relatively safe range of motion in which the lumbar spine can adapt (eg: slight lumbar hyperextension to neutral or very slight lumbar flexion).
Structural factors contributing to butt wink include an individual’s hip socket depth, anteversion angle of the acetabulum or femoral neck, and diameter of the femoral neck. As one descends into the squat, the femoral neck (ball) makes contact with the superior rim of the acetabulum (socket). (Mee 2017)
To continue the descent, the spine must flex as the pelvis posteriorly rotates to compensate for the inability of the hips to continue flexing, in order to keep the weight (barbell and upper body mass) over the base of support (feet). Likewise, inadequate ankle dorsiflexion may contribute to butt wink by preventing knees from traveling forward and effectively shifting hips back. Torso must be angled more forward to maintain the center of gravity over feet, possibly requiring spinal flexion once maximal hip range of motion is achieved. A long femur or short torso can contribute to but wink in much the same way.
A wider stance can both slightly increase hip range of motion and decrease the required magnitude of ankle dorsiflexion, possibly preventing butt wink, decreasing its magnitude, or increasing the depth in which it occurs. With a wide stance squat, the Adductor Magnus inflexibility may have a greater potential of inhibiting hip range of motion contributing to the onset of a hip wink in a deep squat, particularly if ankle flexibility and structural hip issues are not contributing.
If hip mobility or ankle range of motion is insufficient to perform parallel or full squats, particular hip and ankle flexibility exercises can be performed in an effort to eliminate or decrease spinal articulation throughout the lower portion of the squat. See Full Squat Flexibility and Deep Squat Test.
Boyden G, Kingman J, Dyson R, (2000). A comparison of quadriceps electromyographic activity with the position of the foot during the parallel squat. J Strength Cond Res. 14(4): 379-382.
Conteras B (2015). Why Do People’s Knees Cave Inward When They Squat? Bretconteras.com [accessed 15 Oct 2017].
Escamilla RF, Fleisig GS, Lowry TM, Barrentine SW, Andrews JR (2001). A three-dimensional biomechanical analysis of the squat during varying stance widths. Med Sci Sports Exerc. 33(6): 984-98.
Fleck SJ. and Falkel JE (1986). Value of Resistance Training for the Reduction of Sports Injuries. Sports Medicine, 3, 61-68
Fry AC, Smith JC, Schilling BK (2003). Effect of knee position on hip and knee torques during the barbell squat. J Strength Cond Res. (4): 629-33.
Hatfield FC (1989). Power: A Scientific Approach, Contemporary Books, 158.
Hay JG, Andrews JG, Vaughan CL (1982). The biomechanics of strength-training exercises. Proceedings of the 10th International Conference of Sport, Physical Education, Recreation and Dance; 7: 99-109
Kraemer WJ, Fleck SJ (1993). Strength Training for Young Athletes, Human Kinetics.
McCaw ST, Melrose DR (1999). Stance width and bar load effects on leg muscle activity during the parallel squat. Medicine and Science in Sports and Exercise; 31(3): 428-436.
Mee D (2017) What really causes butt wink? Strengthandconditioningresearch.com, accessed May 30, 2019.
Palmitier RA, An KN, Scott SG, Chao EY (1991). Kinetic chain exercise in knee rehabilitation. Sports Medicine; 11: 402-413.
Signorile JF, Kwiatkowksi K, Caruso JF, Robertson B, (1995). Effect of foot position on the electromyographical activity of the superficial quadriceps muscles during the parallel squat and knee extension. J Strength Cond Res. 9: 182-187.
First squat line drawings (side view) from Trainer ClipArt.