First, the osteoporosis detected in many female endurance athletes is caused by amenorrhoea brought on by excessive endurance training and/or weight-loss regimes bringing them into the less-than-13% fat category. Such amenorrhoea is found in female ballet dancers, middle- and long-distance runners, lightweight female rowers (but not the open weight female rowers) and other lightweight female endurance athletes.

On the other hand, exercise and regular training clearly promotes bone growth, particularly in the exercising and/or weight-supporting limbs. Thus, as would be expected, there will be less prevalence of porous bones in non-menstruating female athletes in those limbs most used in a particular sport. Thus lightweight rowers have less bone loss in the lumbar spine compared to runners because they use their lumbar spines much more. Conversely runners' legs are weight-supporting throughout all their training. Exercise and bone growth, and amenorrhoea and osteoporosis effects should therefore be clarified separately.

With regard to rowing, may I point out that it is essentially an endurance sport raced over 2000m which takes between 5.5 and 9 minutes, depending on the type of boat (eight to single) and the wind direction. Thus it clearly falls into the category of a middle- to long-distance event. The contribution of the energy systems to this performance is 75% aerobic and 25% anaerobic. In addition, our winter long-distance events (run like cross-country running on a timed basis) last a minimum of 20 minutes and provide the endurance foundation training during the preparation period.

Yes, heavyweight rowers also do a great deal of strength and strength endurance training but this is not the priority, and rowing, per se, does not increase muscular bulk. Our lightweight rowers avoid strength training in the repetition range because it enhances muscular bulk.

To return to the article, it would be my guess that in comparing rowers with triathletes the research results showed up basic differences in physique as opposed to any effects of training. The rowers would have big frames and heavy and dense bones whereas triathletes would have lighter frames and smaller, lighter bones. The somatypes of the two groups would be entirely different.

Bone density has a strong relationship to weight bearing and triathletes.who swim and cycle as well as run spend a significant proportion of their time doing weight-supported training. Equally, rowing is a weight-supported sport but this should not be understood to mean that rowers don't use their legs. Indeed, since the invention of the sliding seat during the last century the legs have been the prime movers in rowing performance. Cycling is also a weight-supported sport but everyone knows that in cycling the legs are the main performance muscles. Rowing a 2000m race is is like doing 360 deep squats at a rate of 36-40 per minute for around six minutes. When rowers tie-up let everyone be clear that it is the legs that move the boat and the legs which 'blow' first. Since spectators at rowing regattas see more of the upper bodies than the legs it could easily be assumed that the legs are used less than they actually are. It is the explosive power of the legs which moves the boat initially and the back and arms add to, and follow through, from the legs.