Study Showing Football Helmet Design Reduces Concussion Rate Raises Many Questions, Says NOCSAE's Oliver

A 2014 study (Rowson S, Duma SM, et al 2014) reporting that football helmet design can reduce concussion risk has prompted criticism from those with a stake in football helmet design.

To get reaction to the study, I asked Mike Oliver, Executive Director of the National Operating Committee Standards and Equipment (NOCSAE), the non-profit group that sets standards for football helmets, to comment. Here is what he told me:

Question: Does NOCSAE have a comment on the study reported in the Journal of Neurosurgery (Rowson S, Duma SM, et al 2014) finding that helmet design can reduce the significantly reduce the risk of concussion?

Oliver: We are still reviewing the study and its conclusions, and it is hard to tell at this point whether the conclusion is that the Riddell Revolution helmet model is better than the Riddell VSR 4 model, or whether that conclusion also applies to other helmet brands and models. Those two helmets are based on different design concepts, with the main difference being that the Revolution model has a larger shell with more padding on the inside. This type of design difference also exists in almost all other brand helmet models introduced since 2000 when the Revolution was first made available, and this new design type has replaced the older style helmets over time, but we see an increase in concussions, which is not consistent with the study findings. So there are lots of questions and information that the study does not address.

Question: NOCSAE has criticized the Virginia Tech STAR helmet ratings on a variety of grounds. In a July 2013 statement, NOCSAE said, "A concussion in football is a very complex event involving different and changing forces, linear and rotational accelerations, helmet fit, player position, impact duration, player concussion history and overall health, and potentially even genetics. The Virginia Tech Helmet Ratings system approaches the very broad and complex issue of concussion protection from a narrow vantage point of linear accelerations only and does not address other biomechanical variables such as rotational accelerations, particularly where rotational accelerations precede the linear acceleration in a hit. The consensus of scientific experts agree that rotational accelerations are involved in many, if not almost all, concussive events, although no correlated injury threshold for rotational accelerations has been found." Given this new study do you stand by your earlier statement?

Oliver. Yes, we do still stand by the statement. Research recently presented at our Winter meeting supports the conclusion that rotational accelerations are a primary causative force in most concussions, and that whether or not a concussion will happen in a particular event is determined by many factors unrelated to the magnitude of the linear force. The STAR system as currently reported relies on NOCSAE drop test results which involve only linear forces, and the ranking/rating only applies to adult large helmets used by collegiate players.

Question: In light of the study's finding that the Riddell Revolution reduced concussion risk compared to the Riddell VSR4 by 54%, as predicted by the Virginia Tech STAR helmet rating system, does NOCSAE continue to feel that that criticism is valid? If so, why?

Oliver: Yes, until the data behind the study is made available for review and analysis, it is very difficult to evaluate the study. Impact and concussion data has been collected by the Riddell/SIMBEX owned HITS system from collegiate players beginning in 2003 through 2013 but the study only evaluated data between 2005 and 2010. Almost all of the original data collected by Virginia Tech in 2003 and 2004 was done in the VSR 4 helmet, but was not included in the study. Additionally the data collected at Va Tech after 2010 was done with the Revolution helmet and did not involve any VSR 4 helmets, but that data was not included in this study. We also know that SIMBEX filters the impact data using an internal and propriety algorithm that culls out a percentage of the impacts before being used in the various published studies, and the current study does not address how many of those filtered and excluded impacts involved Revolution helmets.

Question: Another of NOCSAE's criticisms of the STAR ratings was that, "According to an independent statistical review of the scores and categories upon which the STAR number is based, there is no significant statistical difference between helmets in the 5 STAR, 4 STAR, and 3 STAR categories." While the current study tested a 4-star helmet and a 1-star helmet, do the findings suggest in any way that 5 star helmets may do a better job at reducing concussion risk than 4 star, or 3 star helmets, or does it simply show that a 4 star helmet reduces concussion risk significantly more than a 1-star helmet?

Oliver. No. The study only looked at two models from the same manufacturer, and does not include any other brands or models. The study does not appear to extend to any particular STAR value model, but is limited to two models made by the same manufacturer. Our statistical evaluation does recognize a statistical difference between the 2 and 3 STAR group of helmets and the 3-5 STAR group of helmets, but not between those in each group, but even that statistical difference doesn't take into account any other variables such as helmet fit.

Question: In addition, NOCSAE criticized Virginia Tech for its failure to test helmets under game conditions: "There are a near infinite number of ways to test helmets (varying temperatures, impact location, helmet size, drop height, etc.) and, therefore, generalizations were made so that the helmets could be tested in a practical manner. Helmets were not tested under game conditions. For example, air bladder fitting and protection systems were not inflated to achieve it, even though the NOCSAE standards require that manufacturer fitting instructions be followed." In view of the new study, which tested helmets under game conditions, and involved data collected from 1833 athletes over a six year period, is this criticism still valid?

Oliver: The quote you have is not my statement, it is the statement of Dr. Duma in their FAQ section. While game conditions (such as temperature) may make a difference in helmet performance, the issue we raised with the STAR methodology on this issue is the helmet fitting to the test headform. As Dr. Duma recognizes in the quote, specific helmet fitting instructions were not followed, but instead a general fitting protocol was used with all helmet models. NOCSAE standards recognize that helmet models may have specific fitting requirements, including the use of the air bladders, and that those instructions should be followed when testing the helmet. We also know that the air bladder fitting system does provide measurable impact performance at very low impact velocities, even though that is not their purpose. By not inflating the air bladder system in the STAR testing program, some helmet models will test inaccurately at lower velocity impacts. Again, the study only compares a Riddell Revolution with the Riddell VSR 4. The statement regarding testing helmets under actual game conditions

Question:  How, if at all, are the results of the study going to inform NOCSAE in establishing new football helmet standards, as it says it is prepared to do at its June 2014 meeting in Boston?

Oliver: It is difficult to know the answer. There are at least two studies either published or to be published which conclude that there is no measurable difference in concussion rates between helmet brands, models, and age, so more information would have to be developed in order to answer the question.