Starting to Do Research

After I entered the graduate school of the University of Tokyo in April 1966, I started doing research on sports science in Prof Ikai’s laboratory. At that time, new theories of exercise training were being introduced to Japan following the 1964 Tokyo Olympic Games. I was particularly impressed by Dr Hettinger’s work on isometric training. When I started to work in Prof Ikai’s laboratory, I was eager to do a scientific study of training, but Prof Ikai suddenly ordered me to investigate ultrasound. That determined the direction of my research. However, I did not understand the relation between research on strength training, which was what I wanted to do, and studying ultrasound, which seemed to have little to do with training for sports. This worry continued during my graduate school days. It was more than a year later before I came to a dim understanding of the relation between ultrasound, muscle strength, and training. However, the mental power that I obtained from my concerns about the relation between ultrasound and the science of training was very useful to me in my subsequent research.

Struggling With Ultrasound

At that time, around 1966, ultrasonic imaging of the inside of the body (cross-sectional images) had just begun on a trial basis by scientists around the world. The equipment available was not sophisticated and tended to have problems that often interfered with research. Textbooks of anatomy showed only longitudinal views; no texts had anything on cross-sectional anatomy. Therefore, I attended an anatomy class at the University of Tokyo’s Faculty of Medicine, in which I dissected human cadavers and compared actual photographs of dissected limbs with ultrasonic images.

By persistently repeating ultrasonic measurements every day, I could, to some extent, get an image that looked like a cross-sectional image of an arm, and so I did my thesis on the relationship between muscle strength and muscle cross-sectional area. In a publication based on my master’s thesis,¹ we reported that muscle strength per unit area was 6 kg/cm². However, my concerns about the accuracy of the ultrasonic sectional images persisted. After I was hired as an associate professor at the University of Tokyo in 1976, I replicated my doctoral dissertation research. The result, 6 kg/cm² of muscle strength per unit area, was the same value that I had found in 1968. Furthermore, in 1990, Dr Kawakami’s results from measurement with magnetic resonance imaging (MRI) were almost the same (6 kg/cm²).² When I saw that the results reported in Dr Kawakami’s publication matched my results, my lack of confidence completely disappeared.

From Cross-sectional Slicing to Longitudinal Slicing

Because I had been able to replicate the results of my doctoral dissertation, I gained confidence in using ultrasound to measure cross-sectional area and became proud of being considered a specialist in ultrasonic cross-sectional measurement. However, in the late 1980s, the emergence of computed tomography (CT) and MRI methods was a big shock to ultrasonic cross-sectional measurement specialists, who felt that the time of measuring muscle cross-sectional areas with ultrasound was over. Then, in 1985, I got the idea that, in order for the ultrasonic method to survive, longitudinal slicing had to be done. At the same time, because I was developing an interest in the measurement of the physiological cross-sectional area of muscles using MRI, I asked Dr V.R. Edgerton if I could spend a year at UCLA doing research using MRI for that purpose.³

A longitudinal image from ultrasound looked like a fish that had been split open lengthwise, and the meaning of a visible diagonal white echo was unclear. In 1989, a report on pennation angles measured by ultrasonography was presented at a meeting of the International Society of Biomechanics held at UCLA. At that time, members of our research group had various ideas about a physiological explanation of ultrasonic images of longitudinal slices. In those circumstances, Dr Kawakami showed a strong interest in that method and continued by trial and error. He finally established a method of real-time measurement of the pennation angle.⁴

In the 1990s, research teams using ultrasonography started trying to quantify fascicle length. As a result of this work, measurement of the length of muscle bundles and tendons became possible. The quantification of changes in the length of tendons revealed that the behavior of a muscle–tendon complex depended on muscle activity.⁵ Furthermore, it was confirmed that, at a basic physiological level, the activity characteristics of muscles (isometric, concentric, and eccentric contraction) are not necessarily coincident in living bodies, as had been believed.⁶ In order to check the dynamics of muscles and tendons, movements of muscles and tendons were observed with an ultrasonic probe mounted on the gastrocnemius. This confirmed that the length of a muscle bundle during human walking was almost stable (the so-called demonstration of isometric tension) when the gastrocnemius was very active; that is, the magnitude of EMG discharge was high.⁷

Around that time, our ultrasound team published many papers on the relation between the dynamics of muscle–tendon complexes and their functions, using ultrasonic longitudinal images related to physical exercise. The series of research papers on this topic was recognized by the International Society of Biomechanics when it awarded me the Muybridge Medal in 2003.

From 2008 to 2016, I held the position of president of the National Institute of Fitness and Sport (NIFS) in Kanoya. Managing a university as its president was very stressful. However, it gave me a good chance to realize my dreams that I had had for a long time for sport sciences. In 2015, I had an experimental gym built (70 m long, 50 m wide, and 15 m high), equipped with a 50-m track with force platforms on the surface. Also, equipment was installed to enable real-time analysis of baseball, soccer, and tennis. Having 50-m-long force plates on the track enabled the measurement of ground reaction force to each step during sprint running. Many papers were published reporting research that used this equipment.^8,9

A Lifetime of Continuing to Measure

There are various ways to do research. My way is to start with measurement, concentrating on collecting data about an interesting topic. This method has sometimes been ridiculed as being just a data collection shop. However, when measuring the functions of living people, it makes sense to measure persistently.

For example, I have been measuring my muscle (grip) strength every morning for 20 years. Every year, when I was 60–66 years old, my muscle strength tended to decrease by 5 kg in January and February (40 kg) and increase by 5 kg in July and August (45 kg). This fluctuation of muscle strength may have occurred because of the change in seasons, some individual characteristics, or other factors. It is not possible to find reasons for the changes, but clarification may be found in continued measurement. Daily fluctuations in an individual’s physical functions must be considered when evaluating the effects of exercise training.

I have often regretted measuring what I did not need and not measuring what I did need. Because I do so much measurement, the efficiency of my research work is not high, but this method fits my style. In the usual scientific method, researchers predict possible results before starting an experiment, develop hypotheses, and then start their research, but that is not my way.

Writing Papers: Both a Difficulty and a Delight

Experiments and data collection are followed by presentations at professional meetings and writing papers for publication. For me, generally, preparing presentations for academic conferences is comparatively easy, but writing papers requires considerable time. I enjoy doing experiments and organizing data. Those are exciting activities. But I get nervous when I start to write a paper. I will soon be 80 years old, but I still feel strong stress at such times. I even feel this way while writing this article. For this reason, when a paper has been accepted for publication, I become very happy. I immediately get the feeling that I could write papers one after another, but actually I have never written papers that easily. I do not understand why. I dream of having days in which I can write papers as casually as playing golf, but that dream will probably never come true.

I retired from the presidency of NIFS in Kanoya in 2016. Three years later, that is, in 2019, I started to write a paper on the mechanical energy of sprint running using data that I had obtained in collaboration with Professor Matsuo while I was president of NIFS. Now, I and my collaborators, Professor Matsuo (67 years old, retired from NIFS), Professor Kanehisa (66 years old, Ritsumeikan University), and Professor Kawakami (55 years old, Waseda University), live far from each other, but we are continuing to work together on the paper by meeting online every week. In these meetings, we discuss the analysis of the data, check related publications, construct the main points of the article, and so on, in order to be able to finish the paper. Thanks to recent developments in computer software, we can hold intensive discussions of research regardless of our locations and the locations of our universities and laboratories. Forming a research team regardless of country, university, or academic affiliation is a good model for the future style of laboratories.

Team Building for the Enjoyment of Doing Research

People have their own research interests, interests that are neither forced on them by others nor forced on others by them. Free thinking creates new research results. When many people participate in a study, various ideas will be mixed together. What is important is to respect individuals’ ideas and to make progress. I believe that the laboratory is a place for training, a kind of Japanese dojo (training hall). Therefore, I do not want members of my team to avoid opportunities to train each other. Our research group has a meeting every Wednesday evening in which we get together in order to discuss topics, such as setting the theme of our research, considering various ideas, and going over experimental data and experimental plans. It is a weekly routine (the Wednesday meeting), but I feel that a tense atmosphere is created in the process of setting research themes, deciding on topics, evaluating experiments, and analyzing and summarizing experimental data. In my field of study, it is almost impossible to do all that alone. Naturally, many colleagues are needed. When doing research on the human body, the research participants are also important colleagues. Without their active involvement, it is impossible to collect accurate data.

In other words, an important condition for completing research successfully is how the research team, including the participants, is organized. To enjoy sports, each player should perform optimally in their position with respect to the other players. For example, a pitcher has to have a catcher. Research will proceed smoothly when all members of the research team have their own roles.

A key reason why I have enjoyed doing research in my beloved sport sciences is that it has enabled me to meet various people. The Japanese saying that people are stone walls and people are castles is true. Even if neither money nor experimental equipment is available, as long as people can work together, everything will go well.

In Hagakure, a 17th century book about the Japanese samurai spirit, it is written that there are 4 patterns of human ability: quick-quick, slow-quick, quick-slow, and slow-slow. The quick-quick type understands quickly and acts quickly. The slow-quick type is slow to understand, but quick to act after fully understanding. The quick-slow type understands quickly, but is slow to act, and the slow-slow type is slow to understand and slow to act. I belong somewhere in between the slow-slow and slow-quick types. I am probably a (slow-slow) + (quick-slow) type. In a research team, a mixture of these types will bring about good results. However, whether or not team members are good-natured is another important factor.

There are 2 kinds of people: good ones and capable ones. There are not many people who are both good and capable, but we can form a good and capable team with a mixture of good people and capable people. I was lucky that I was able to meet good people and capable people and form a good and capable research team. I would like to continue to value these qualities in the future.

I would particularly like to acknowledge Dr Hiroaki Kanehisa, former vice president of NIFS, who is both good and capable in addition to being a quick-quick type. He consistently supported the building and retention of our research team.