Lessons from Masters World Records: Lack of Gender Differences In Aging Muscle Decay Rates

Guidolin D1, Gava P2, Ravara B2,3, Kern H4, Amber Pond L5 and Albertin G1,6* 1Department of Neuroscience, University of Padova, Italy 2A&C M-C Foundation for Translational Myology, Italy 3Department of Biomedical Science, University of Padova, Italy 4Ludwig Boltzmann Institute for Rehabilitation Research, Austria 5Department of Anatomy, Southern Illinois University School of Medicine, USA 6Human Movement Sciences, University of Padova, Italy

proportional to the power developed by the athlete in carrying out the athletic gesture. Such a parameter is a dimensionless number ranging from the maximum value of 1 (for the absolute world record, that is, those of the Senior Athletes) to decreasing values corresponding to the decreasing performance of the Masters athletes. The many confounding factors found in clinical studies on age-related performance usually decay, so the different lengths of clinical longitudinal studies [11,[14][15][16] or the use of different modalities to measure strength, power, and resistance to fatigue [17] were avoided. The results of this normalization were rows of up to 16 performance parameters conveniently interpolated with polynomial trend-lines that allow for the direct comparison of the performance of very different athletic gestures. Using normalized data, the main conclusion of the Gava et al. [5] study was that the performance decline noted in the different Track and Field events was actually very similar across all events and that this variability was more related to the methodology used for the normalization of the records of the different athletic specialties. In a second study, these analyses were extended to include comparisons of male and female Masters athletes in 19 Track and Field competitions, providing physical [18] and statistical evidence [19] that there are only very minor gender differences in the age-related performance decay between female and male Masters athletes. Since our initial conclusions, according to the criticisms of external evaluators, were based on a naïve statistical approach, we will here report the results of more robust statistics (i.e., weighted regression analysis) [20].

World records database
First, our approach required a database with all the world records from the main disciplines of athletics: 11 track, 4 throwing, and 4 jumping events. These collated records contained data from all the Masters athlete categories, both female and male (19 categories per gender). Data from the world records of Senior athletes and on all the Masters athletes was collected from the official archives of different world athletic associations: IAAF (International Association of Athletics Federation [http://www.iaaf.org/]) [21] for the absolute world records of Senior athletes and the world Masters athletics (http://www.world-masters-athletics.org/) [22] for the world records of Masters athletes. All data used in this study was public data collected from events officially recognized by the World Federation of Athletics. They are the official records valid as of May 2013, but similar results could be obtained for previous and successive years.

All athletic performances are "power performances"
The work to displace the body of the athlete from the start to the finish line divided by the time spent is directly proportional to the power developed (the less time, the more power). Meanwhile, the distance reached by the piece of equipment in throwing is directly proportional to the kinetic energy transmitted to the equipment by the athletic action, again proportional to the physical power developed by the athlete.

Normalization procedure
After collection of the database, the data from each event in the Masters world records were "normalized" using each relevant absolute world record, taken as the reference value of 1. Specifically, the data from the Masters world records, which increased generally with age (such as time for the running events), was normalized by dividing the absolute record by the Masters record. On the other hand, the data on the events, which decreased with age (such as the height for the high jumping events and distance for the throwing and long jumping events) were normalized in an opposite manner: the Masters record was divided by the absolute record. Consequently, the normalized records of each event were sets of dimensionless values that decreased with age, from 1 (normalized value of the absolute record) to 0 for a null value. With such a procedure, the performances always decreased with age, in line with the usual idea of decline, regardless of the fact that the measured performance values increased (running) or decreased (throwing and jumping) with age. This was to allow for a consistent form of measurement and graphical representations across the different events, which reflected the fact that increased age was associated with a decrease in performance. We will later discuss the limitations imposed on the normalization of the other variables needed to compensate for the other aspects of Masters athletes competing at increasing ages (see two-step normalizations).

Statistical analyses
For each discipline, a weighted regression analysis was used to estimate the course of the performance as a function of age. The slopes of the best fitting lines (representing the rate of change in performance) obtained from both the males and females were then compared using a Student's t-test [23]. As an index of overall performance at each age point, the mean value of the normalized records was considered. Each of these points, however, did not provide equally precise information on the deterministic part of the total process variation, since the number of disciplines with an available world record decreased with the increasing age of the athletes. To account for this aspect, a weighted regression procedure was used to characterize the trend exhibited by the overall performance score as a function of age [24]. This was done by associating each point with a weight corresponding to the number of normalized data contributing to its value in order to give each data point a proper amount of influence over the parameter estimates. The line slopes estimated for both the males and females were then compared using a student's t-test. For the statistical analysis, the SPSS 13.0 software was used, and p < 0.05 was used to determine statistical significance.

Result and Discussion
The normalized parameters of our study were derived from the world records of Masters athletes (male and female) for 19 athletic disciplines.

Why two-step normalization?
In the case of throwing events, performances needed to be further normalized by taking into account the decreasing weight of the implements with the increasing age of the Masters athletes. For the jumping events, the power developed by the athletes as associated with the displacement of the Center of Mass (CoM) of their own body, which is essentially equal to the jumping length for horizontal jumping events (e.g., Long Jump and Triple Jump), but is not proportional to the height of the cross bar in vertical jumping events (e.g., High Jump and Pole Vault). The height of the vertical jumps was unquestionably mainly linked to the speed of the athlete's CoM at the time of detachment. Such speed, according to Newton's second law of motion, depends directly on the impulse imparted on the athlete's mass at take-off. The impulse, in turn, depends on the athlete's ability to accelerate his CoM, a capacity that is directly linked to the power developed in the athletic gesture. For this reason (in spite of controversial objections raised by some authors; [25][26][27]), we treated the lifting of the athlete's center of gravity in the jump as directly connected to the power developed by the athlete in the execution of the jump [25,28]. Indeed, when performing vertical jumps, the athlete has to raise his own CoM from a starting level of about 110-120cm from the ground at the take-off point to 10-20cm above the crossbar. Thus, vertical jumping performances require a two-step normalization process in order to obtain a normalized dimensionless parameter proportional to the power developed during the performance. Despite some approximations, this method allowed for a conceptually correct "normalization" of vertical jumps.
On the other hand, despite these approximations, all slopes had an R2 higher than 0.90, suggesting a continuous process of performance decay with age that occurring in both male and female athletes throughout life from at least the age of 35 (when the data was first gathered) [5]. Figure 1 shows, for each considered discipline, the slopes of the best fit lines of the normalized records as a function of age for both males and females. The values are represented together with their estimated standard errors. Asterisks indicate disciplines for which a statistically significant difference was found between male and female athletes in terms of the rate of performance change. We were unsurprised that these were jumping and throwing events (see above). Figure 2 shows the weighted linear regression of the overall performance score (mean value across the disciplines of the normalized record data) as a function of age. The obtained line slopes for the males (black dots, solid line) and females (white dots, dashed line) are shown together with their standard errors and degrees of freedom (d.f.). No statistically significant differences were detected when applying the student's t-test (p = 0.171).

Rates of performance decline in master's athletes: Implications and conclusions
The results of our analysis revealed some indisputable elements:

1.
Despite the differences in the absolute performance of female and male skeletal muscles (15% lower in females) [2][3][4], the trend of performance decline with age was similar, if not identical, for the two genders (Figures 1 & 2).

2.
The weighted regression analyses of the male vs female dimensionless parametric data did not reveal any unexpected peculiarities, which is agreement with previous analyses that used different approaches [6,8,10,11,12,24,29,30], with exceptions related to 2-step normalization events (see above).

Conclusion
In conclusion, taken together, the quantitative analyses of world records of Masters athletes described here suggest that there are no significant differences in the rate of "age-related decline in athletic performance" between females and males. This is unexpected in terms of gender-related sport behaviors. The consequent implications may have long-term influences on the biology, physiopathology, and management of early aging, especially in terms of metabolic sarcopenia, cancer cachexia, aging per se, and its complications [58][59][60][61][62][63][64][65][66][67][68][69].