There is concern voiced within the aikido community regarding knee injuries, but to date it has been largely anecdotal. The Aikido and Knee Injury Project will gather data on the incidence, causes and effects of knee injuries in aikido. Part One focussed on the incidence of knee injuries, specifically the number of adult students within each responding dojo who had an acute knee injury related to their training during the past five years. While some factors about dojo practice were examined, the goal was not to extrapolate a cause, but rather to begin quantifying the problem of knee injuries and to "take a snapshot" of current practice.
    Movement from the center and the taking of balance in aikido is characterized by circular movements. Most commonly this involves lateral turning of the body; that is, pivots in combination with steps or slides. It is often accompanied with vertical and/or spiral turning, with bending of the knees or a complete drop to kneeling. The overall effect is repeating torquing of the knees. In the course of a typical one hour class. a student might perform technique 120-200 times and receive technique (be pinned or thrown) 120-200 times. For the knees, this represents many repeated incidents of torquing, bending, and impact with the mat.

The survey was placed on two separate interactive web pages to ensure anonymity. The first page included identifying information for the responding dojo. The second page was the survey itself. Dojo e-mail addresses were obtained from the web sites of the various styles and organizations within the aikido community. Over 300 requests were sent to individual dojo. In addition, notices with live links were posted on the aikido-L (an internet aikido mailing list) and on two general information aikido web sites (Aikiweb and the Aikido Journal).

From around the world, 101 dojo, representing all styles and affiliations, sent in complete, usable responses. They comprise 3253 individuals over 21 years old, practicing aikido in 15 nations. Follow-up questionnaires gathered additional information, and 78 of the 101 original participating dojo provided responses to the follow-up.

It needs to be stressed that, while I was delighted to have over 100 responding dojo, this is a relatively small sample. There is no factor that can be pointed to and identified as "a cause of knee injuries." There are some factors that, for reasons that are unclear, do seem to correlate with a higher incidence of acute knee injuries.

    Half of the participating dojo reported NO acute knee injuries in training during the past five years. Fifty one dojo reported a total of 99 persons had experienced acute knee injuries in training. This is 3% of the total respondents; computed on a per-dojo basis the average acute knee injury rate is 4.45%. The figures are virtually identical for men and women.
    Those half of the dojo reporting acute knee injuries were divided into three sections for comparison: those with the highest incidence of injuries (determined as those in the 80^th percentile, which amounted to 11 dojo), the corresponding number of dojo showing the lowest incidence of injuries, and the remaining number comprising "some injuries."
    What is intriguing is that the eleven dojo in the "high incidence" group represent less than 5% of the total adult members being reported on yet account for over 30% of the acute knee injuries. This accounts for the comment by many participating dojo that they really do not perceive knee injuries as a common problem, and raises the question of what, if anything, this group of dojo has in common with each other but not with the rest of the study group.

    The most commonly practiced warm-ups are standing and moving the knees in a circle and sitting in seiza and lying back. What stands out is that a significantly larger proportion of "many knee injury" dojo perform two of the warm-ups:
    Bunny hops are only done in 7% of all dojo, but are done in 18% of the "high injury rate" dojo, double any other category and an increase of 157% above the respondents as a whole.
    Standing with bent knees and moving them side-to-side is done in 64% of the "high injury rate" dojo, representing a 78% increase over the whole survey population.
    It is not clear what this correlation signifies. It seems unlikely that the warm-ups in themselves are a factor. Is it possible that they indicate a certain higher intensity of practice that makes injuries more likely?

    Most dojo report that they sometimes do seated technique (suwariwaza or hamni handachi), but only for one or two techniques and not often. The only place that this drops to under 60% is for dojo in the "many injuries" category, a quite significant drop to 36%. Forty-five percent of the "many injuries" category of dojo report doing seated technique with high frequency. This is considerably more than is reported by the dojo with lower incidence rates of acute knee injuries.
    I decided to look at those twenty-six dojo reporting higher frequency of seated practice (23 that in most classes do one or two techniques suwariwaza or hamni handachi, plus 3 that often do classes consisting mostly or entirely of suwariwaza or hamni handachi). Six of them are in the "highest injury" group of dojo. Their overall injury rate is considerably higher than for the survey as a whole. When it comes to male-versus-female injury rates, it is the only factor noted in the survey to reflect a major discrepancy. This is not explainable by disparity in relative numbers of female students, who comprise a similar proportion of total members for this group and other groups.
    It is not clear whether this correlation has to do with injury related to seated practice or, as with certain warm-ups, if it is a reflection of an intensity of practice. Whether the jump for women knee injuries is an anomaly is likewise unclear. Hopefully, surveying individuals who have been injured will shed some light on these issues.

    When I first looked at data in a table, it seemed like the number of students per dojo was correlating negatively with the incidence of acute knee injuries. I looked at injury rates for the groups of smallest, the most average sized, and the largest dojo in the survey. Those smallest dojo had up to 9 adult members and a per-dojo injury rate over 5%. A cluster of dojo around the median and mean, with between 30 and 36 adult members, had a per-dojo injury rate of close to 2.5%. The eleven largest dojo had over 55 members each, and per-dojo acute knee injury rate of under 2%. So there was a trend.
    One thought that came to mind was that smaller dojo might represent newer dojo. A follow-up survey was done to ascertain how many years the participating dojo had been in operation, with 78 of the original participating dojo responding. In fact, there was a relationship: Of the ten smallest dojo, 40% had been in operation five years or less, while all of the ten largest dojo had been operating six years or more. Another 40% of the smallest dojo had been operating over ten years, as compared to 60% of the largest dojo.
    Looking at how long dojo had been open, two things stuck out: Of the dojo in operation five years or less, 22% were also dojo with "no injuries," higher than predicted based on their proportion within the survey. This may be because they have a higher proportion of beginners, who are not training at the speed or intensity that is associated with torquing knee injuries.
    Of the dojo in operation for six to ten years, none were in the "small number of injuries" category, but 24% were in the "moderate number of injuries category." I am speculating that this may represent the tendency some aikidoka have noticed for students approaching shodan to become somewhat reckless in their training. For the dojo in operation over ten years, the numbers tend to balance out again.

    Of the original responding dojo, 78 followed up with information on the type of mats used for training. There was a decent balance in this group between dojo reporting no knee injuries (35) and those reporting any acute knee injuries (43). A couple of numbers jump out:
    Wrestling mats seem to correlate with no knee injuries (despite anecdotal complaints that feet stick to them). On the other hand, canvas stretched over foam surfaces seem to correlate with moderate numbers of knee injuries.
    Finally, of the 11 dojo with the highest knee injury rates, eight responded, and 50% of these train on mats of artificial tatami/foam; given that this represents merely 4 dojo out of a small subset of 8, it is unreasonable to draw any conclusions other than to suggest that mat surfaces seems to be an area worth more detailed research.

    Several people suggested that, as long as a follow-up survey was going out, I inquire about the frequency of jumping highfalls. This was defined on the survey as falls involving having the whole body off the mat in forward rotation and landing in a fall rather than a roll.
    There is a disparity noted in the "many injury" category, with the overwhelming majority sometimes doing jumping breakfalls, but with substantially fewer of these dojo reporting that they frequently do them.

    It was most gratifying to have such a positive response to Part 1. I believe that those who participated were open and honest in their reporting, because so many included comments or questions that reflected a real concern about how the art is doing on this issue. We owe all 101 of you a sincere rei.
    I am looking forward to Part 2 of this study, which will gather information from individuals who have sustained acute knee injuries doing aikido. Hopefully it will clarify what some of the contributing factors are, as well as suggesting who might be at highest risk or if certain moves pose more danger to the knees than others.
    As of 2003, Part 2 is on indefinite hold due to time constraints.

Thank you to Jun Akiyama at Aikiweb and to David Cody for technical assistance that made this project possible, to Joel Posluns, and to the members of Aikido-L for their support and assistance with this project.