The Safety of Artificial Turf vs. Grass as a Sport Playing Surface

I was interviewed last week for a newspaper article which looked at the debate over a local school’s intention to transform a grass playing surface to artifical turf.

Among the controversies in sports medicine, the turf vs. grass wars are not the loudest nor the meanest, but they have been among the most persistent ever since 1966, when the Houston Astros first introduced a synthetic turf playing surface in the Astrodome, and dubbed it Astroturf.

Reliant Astrodome

The history of the Astrodome makes for interesting reading:  of note, the original intention was for the surface to be natural grass, and the makers of the dome had installed traslucent skylights to allow for grass to grow on the indoor surface.  Alas, not enough light made it to the playing surface, the grass died, and Astroturf was born.

The progress of science and technology have seen Astroturf give way to newer, so-called second-, third-, and even fourth-generation turfs. The  sporting world has even demonstrated that an indoor venue can sustain a natural grass pitch: witness the luminous Forsyth Barr stadium  in Duenedin, NZ, which has hosted matches from the 2011 Rugby World Cup to a recent Aerosmith concert.

Natural grass surface on Indoor Forsyth Barr Stadium, Dunedin, NZ

Aerosmith, recent performers on the grass of Forsyth Barr Stadium

So, the question from the interviewer to me essentially reduces to,  ‘Grass:  if its good enough for Steven Tyler and Joe Perry, is it good enough for all of us?’

Prior to the interview, I came across and read a wonderful piece from this very blog, written in 2011  by my predecessor and now CJSM Executive Editor Chris Hughes:  The Injury Risks of Artifical Turf in Soccer.   A focus of Chris’ piece is on a meta-analysis performed by Williams et al., “A Review of Football Injuries on Third and Fourth Generation Artificial Turfs Compared with Natural Turf.”  This study looked at injury rates of newer generation turf vs. grass playing surfaces across three football codes (soccer, rugby union, American football) and essentially found comparable rates of injury, with a slight increase in ankle injuries on turf.  The study group recommended that further study would be needed: “Clarification of effects of artificial surfaces on muscle and knee injuries are required given inconsistencies in incidence rate ratios depending on the football code, athlete, gender or match versus training.”

I commend Chris’ blog  piece to our readers as a complement to this blog entry posted two years later by an American clinician.

I stress the “American” adjective to note that the proverbial elephant in the living room in most of these discussions in American sports is American, gridiron football.  With my country’s idiosyncratic penchant for exceptionalism, we refer to the sport as ‘football,’ naming what I would guess six billion folks on this planet know as football as ‘soccer’.  In the discussions of turf vs. grass that occur in this country, the typical focus will be on the surface’s effect on American football.   This is not to say, of course, that the fields will not be used for a multitude of sports, including soccer, field hockey, softball, lacrosse; it is merely to note that in most of these multi-use venues, the ‘tail that wags the dog’ is the American football program and its needs or desires.

So, with that in mind, I set out to look at what evidence might exist in the medical literature that might be newer and complementary to what Chris found a couple of years ago, with a focus on epidemiologic studies focusing on injury rates in American football.  I found a couple of  interesting studies.

Dragoo et al. published an epidemiologic study in 2012 in The American Journal of Sports Medicine which concerned   the effect of turf on the incidence of ACL injuries in NCAA football.  The retrospective study looked at National Collegiate Athletic Asssociation (NCAA) Injury Surveillance data to calculate the association of ACL injury rates in American football players with various influencing factors (e.g. practice vs. competition, turf vs. grass, etc.)  The study authors found that there was a slight, but statistically signficant, increase in ACL injury rates on artificial playing surfaces compared with grass.  They calculated an injury rate of 1.73 ACL injuries per 10,000 athlete exposures (AEs) on turf as compared with 1.24 ACL injuries per 10,000 AEs on grass, a rate 1.39 times higher (95% C.I. 1.11 – 1.73).  They concluded, “The increased injury incidence on third-generation artificial turf may suggest that this surface is not an adequate substitute for natural grass….”  One strength of this study was its reliance on a well-established injury surveillance system.  The authors note several limitations, including, importantly, “…a large percentage of missing data, which may have ultimately skewed the results.”

Turf on local high school softball field, Ohio

The next study I looked at also was published in The American Journal of Sports Medicine: “Incidence, Mechanisms, and Severity of Game-Related College Football Injuries on FieldTurf Versus Natural Grass:  A 3-Year Prospective Study.” The study author, Michael Meyers, followed 24 Division IA NCAA football teams prospectively for three years and analyzed a total of 465 games (230 on FieldTurf and 235 on natural grass) over that period.  This study was exclusively looking at American football game play (as compared to, additionally, practices and scrimmages) and the specific newer-generation turf known as FieldTurf, described by Meyers as “…a polyethylene fiber blend stabilized with a graded silica and cryogenically ground rubber infill.”   The study looked at total injury rates and so-called ‘minor,’ substantial,’ and ‘severe’ injury rates on FieldTurf vs. grass:  ‘minor’ injuries were defined as those resulting in 0 to 6 days of time lost from sport, ‘substantial’ were those resulting in 7 to 21 days time lost, and ‘severe’ were those requiring 22 or more days out of sport.

Meyer used multivariate analyses to demonstrate significantly lower injury incidence rates on FieldTurf than on natural grass.  Rates were calculated as number of incidents per “10 team games.”   For instance, ‘severe’  injury rates on turf were found to occur 2.7 times per 10 games when compared with grass, where such incidents occurred 4.1 times per 10 games.  This finding was indeed statistically significant,  with P = 0.049, a figure coming just under the defined a priori alpha value of 0.05.  Another way of looking at such data, which I typically find more instructive, is the 95% confidence interval (CI):  for severe injury rates on turf, the 95% CI was (2.1 – 3.3) and for grass the 95% CI was ( 3.5 – 4.1);  there was a difference, but it could literally be as small as 2 incidents per 100 games (for an American Division I NCAA football team that would be approximately 9 to 10 seasons).

That is, on the ‘lower end of effectiveness,’  FieldTurf may lower the number of  ‘severe’ injuries an American football team may see by a total of 2 over the course of 10 seasons (instead of 35 such injuries, ‘only’ 33).  That may or may not be ‘clinically significant,’ I would leave it for the reader to decide.

One of the strengths of this study was its prospective nature and the duration for which it was conducted, three years.  As far as limitations, the author notes several, including the possibility that injuries could have gone unreported or that there could be undefined confounders.  One dimension to which I am always particularly sensitive is the funding source for a study.  In this case, the research was funded by the product maker itself, FieldTurf.  My overall take on this study then is that it is intriguing, but the strength of the findings are not overwhelmingly substantial (i.e. its ‘clinical significance’ is unclear to me), and its chief weakness is that the study was funded by the product maker, something I would always rate a potentially significant limitation.

So, where do I stand on this issue?  I do have a bias, and it’s revealed in this photo of me running with my young son:

The grass between our toes….

I acknowledge that there may exist several factors schools or stadium officials may need to factor as they make their decisions regarding turf vs. grass.  There are issues of finance–the capital cost of installing turf, the on-going cost of grass maintenance, etc.–that are most definitely beyond my area of expertise.

However, from a point of view within my scope of expertise–athlete safety–I don’t yet see any solid evidence that turf can lower injury rates in athletes.  The aforementioned article in The Columbus Dispatch quotes me as saying  “If a (school) district asked him whether it should switch to fake turf, MacDonald said he would suggest holding off, at least until the data are more conclusive.”  I stand by that quote.

However, it is my hope that this blog will be a conversation and not a monologue.  I encourage all readers to participate in this conversation, and so please, share your thoughts and opinions, pro and con.  If you are aware of interesting studies in the medical literature that can inform this discussion, pass them on by all means.