KEY WORDS: Fat, skinfold, Filipino, karate, pencak silat.

Participants were members of the Philippine men's and women's national teams in karate (12 males, 5 females) and pencak silat (17 males and 5 females). Standing height was measured with a wall-mounted wooden stadiometer to the nearest 0.05 m. Body mass was assessed with a calibrated electronic digital scale to the nearest 0.01 kg.

A Slim Guide skinfold caliper was used to measure skinfold thicknesses at the triceps, subscapula, supraspinale, umbilical, anterior thigh and medial calf. Compared to the Harpenden and Lange calipers, the Slim Guide yielded technical errors of measurement of around 5% (Ross et al., 2000). The reliability correlation coefficient for a range of skinfold thicknesses using the Slim Guide against the Harpenden caliper and Echoscan 1502 Ultrasound system resulted in an R = 0.98 at the minimum (Anderson and Ross, 1986). All measurements, according to the specifications provided by Ross and Marfell-Jones, 1991, were taken three times, unless the first two were the same, and the median used for statistical analysis.

Proportional sum of skinfolds and proportional skinfold patterning of the athletes were based on the Phantom stratagem (Ross and Marfell-Jones, 1991). Proportional sum of six skinfolds was calculated as: sum of six skinfolds x (170.18/height). Phantom skinfold patterning was derived using the following formula that is based on a hypothetical human population (Ross and Ward, 1984):

z = 1/s [v (170.18/h)^d - p]

where z represents a proportionality value in z scores. s is the standard deviation of the Phantom value for the variable of interest. v is the observed size of that variable. 170.18 is the constant for Phantom height in cm. h refers to the observed height of the participant. d is a geometrical exponent and equals 1 for all lengths, widths, girths and skinfold thicknesses; 2 for all areas and 3 for all masses and volumes. p refers to the Phantom value for variable v.

The data were analyzed for skewness and kurtosis, while the Kolmogorov-Smirnov test was used to assess normality. A two-way (Gender*Sport) ANOVA was used to determine the differences between men and women by sport in RPI, absolute and proportional sum of skinfolds as well as proportional skinfold patterning. The L statistic (e.g., Thomas and Nelson, 2001) was employed in cases where the distributions were not normal. Due to the multiple comparisons, a Bonferroni-adjusted alpha was used for all analyses to prevent Type 1 error (e.g., Ntoumanis, 2001).

Table 1 shows the means and standard deviations of the body composition markers by combative sport and gender. Although there was a combative sport difference in proportional sum of six skinfolds (p = 0.048, eta² = 0.107), this was not significant anymore after the Bonferroni correction. However, there was a gender effect for absolute sum of skinfolds with the females recording a higher value (74.90 ± 24.36 mm and 51.35 ± 20.97 mm for the women and men, respectively, p = 0.004, eta² = 209). They also had a higher proportional sum of skinfolds (80.19 ± 25.31 mm vs. 51.77 ± 21.13 mm, p = 0.001, eta² = 0.275).

Figure 1 displays the proportional skinfold patterning by gender and sport. Collapsed over gender, there was a difference between karate and pencak silat in proportional anterior thigh skinfold (p = 0.002, eta² = 0.246) with the former scoring lower (-2.19 ± 0.66 vs. -1.76 ± 0.75). The males had lower proportional triceps (-1.78 ± 0.62 vs. -0.35 ± 0.75, p < 0.001, eta² = 0.506), supraspinale (-1.87 ± 0.76 vs. -0.93 ± 0.96, p = 0.002, eta² = 0.234) and anterior thigh skinfolds (-2.18 ± 0.46 vs. -1.26 ± 0.95, p < 0.001, eta² = 0.385).

Body fat
Neither the men nor the women carried more weight for height as expressed by the RPI. However, the males had a lower proportional sum of skinfolds, which suggests that, although there was no difference in weight for height between gender, the females may carry more fat. Although the sample size of the women was smaller than that of the men, the findings confirm those of previous studies (e.g., Callister et al., 1991; Gualdi Russo et al., 1992).Caution is warranted, however, because the effect size for the gender effect for fatness is small.

Compared to a combined sample of Italian judo, karate and wushu college athletes, the RPI scores of the Filipinos were similar. The Italian males recorded an RPI of 41.86 cm/kg^0.333 and the females, 42.32 cm/kg^0.333 (Gualdi Russo et al., 1992). A combined sample of male Saudi national elite judoka and karateka had an RPI of 41.31 cm/kg^0.333 (Chukwuemeka et al., 1992), which is again similar to the values of the Filipinos.

Table 2 shows comparative RPI data of elite combative sport athletes from various countries. It is self-evident that RPI depends on the weight division of the athlete (e.g., Claessenset al., 1986a; Pieter et al., 1998a). However, it was not always possible to divide the samples studied into different weight categories. Nevertheless, it is instructive to compare the Filipino athletes with counterparts from other countries. The American, Japanese, Korean and Polish male judoka seem to carry more weight for height than the Filipino athletes. In the females, the Korean and Japanese judoka seem to carry more weight for height, with the Filipino heavyweight judoka having much more weight for height than any of the other groups, male or female. However, RPI is a reflection of total body weight, containing both lean and fat tissue. In other words, it may very well be that the lower RPI scores are indicative of more lean than fat mass.

Studies using sum of six skinfolds to represent body fat in combative sports athletes are scarce. Italian college combative athletes had a sum of skinfolds of 61.1 mm (men) and 74.4 mm (females) (Gualdi Russo et al., 1992), both of which were higher than those of the Filipino male and female karateka. Only the female pencak silat athletes had more total relative body fat than their Italian counterparts. Filipino elite female judoka (< 60 kg) had a sum of six skinfolds of 76.0 mm, while their heavier colleagues recorded a value of 136.3 mm (Pieter et al., 1998a). In contrast, American elite male taekwondo-in had a sum of skinfolds of 36.14 mm and their female colleagues, 54.81 mm (Pieter, 1991), both of which were lower than those of the Filipino athletes in the present study. Russian elite male wushu taolu athletes had a sum of six skinfolds of 30.81 mm, while their recreational sanshou colleagues recorded a value of 42.19 mm (Pieter and Gagonin, 1994).

Clearly, sport-specific requirements should be considered when evaluating the athletes' relative total body fat. In karate, for instance, athletes will have to be able to propel the body through space as fast as possible, as is the case in taekwondo. Excess mass, especially in the form of fat, may be detrimental to performance because of its negative effect on the weight-to-strength ratio (Sinning, 1985). In pencak silat, however, slow and fast movements are alternated, which calls for different requirements. For instance, male taiji quan athletes had 13.9% of relative total body fat (Zhuo et al., 1984). On the other hand, American elite taekwondo-in were reported to have 7.5% (men) and 12.0% (women) of relative total body fat (Taaffe and Pieter, 1990). Although body fat will depend on weight division, combative sports where fast movements are required may most likely call for a low fat mass to enhance the weight-to-strength ratio. Nevertheless, even if fast and slow techniques are part of one's sport, such as in pencak silat and taiji quan, a more desirable amount of fat would still be preferable: too much fat will most likely deter the athlete from achieving peak performance in her or his chosen sport (Sinning, 1985).

Proportional skinfold patterning
The differences in skinfold patterning between males and females collapsed over sport were expected: men had lower skinfolds at the tricipital, supraspinale and anterior thigh sites. Sexual dimorphism in skinfold patterning was also found in other sports (e.g., Carter, 1982; Ross and Ward, 1984). However, the effect sizes for more extremity fat in the females are moderate. Future studies should use larger sample sizes for each gender by sport.

It was shown that Filipino female elite judoka had proportionally more truncal fat than American elite female taekwondo-in, who showed more fat on the extremities (Pieter et al., 1998a). It has been suggested that mechanical efficiency may be at the basis of fat patterning that may also be sport-specific (Malina et al., 1982, Mueller et al., 1982). Similar to their Filipino colleagues in judo, the pencak silat women had more truncal fat than their male counterparts or their colleagues in karate, although the differences were not statistically significant. This was also found in adolescent Mexican-American girls involved in running, volleyball and basketball (Mueller et al., 1982).

Central fat may be more advantageous in judo than in taekwondo and karate, where one has to be able to propel the body through space as fast as possible. Having proportionally low fat at various body parts will surely aid the karateka and taekwondo-in accelerating and decelerating rapidly as is required by the nature of both sports. On the other hand, judoka may benefit from a more stable body position, so as not to be thrown. Naturally, one has to weigh the advantages of a more solid positioning of the body against a more advantageous weight-to-strength ratio (Sinning, 1985).

The results of the present study seem to suggest that there was no difference between combat sports in fatness (sum of skinfolds), which is contrary to what was found by others (e.g., Malina et al., 1982). Skinfold patterning was more in line with what was reported in the literature, with the males recording lower extremity fat.

• The purpose of the present investigation was to assess relative total body fat and skinfold patterning in Filipino national karate and pencak silat athletes.
• The results seem to suggest that there was no difference between combat sports in fatness.
• Skinfold patterning was more in line with what was reported in the literature with the males recording lower extremity fat.