**4. Discussion**

The present study is the first to analyse and describe the health-related physical fitness together with the adherence to MD of Icelandic adolescents.

As expected after analysing the body composition of the participants, significant differences were obtained between the boys and the girls in weight, height, % of body fat, and waist circumference (see Table 1). These results are similar to those obtained in several studies on the adolescent population [60–63], in which girls had higher levels of adiposity, whereas boys had higher weight, height, and waist circumference values. Both the boys and the girls show average values of BMI, % of body fat, and waist circumference (see Figure 1).

The weights, heights, and BMIs of the present study sample are comparable to the reference values provided by Ortega et al. (2011) [59]. Unlike what was found by Wärnberg et al. (2006) [64], there is no prevalence of obesity in the adolescent participants of this study as the BMI, waist circumference, and % of body fat are considered medium values (see Figure 3).

In relation to the performance in the physical fitness tests, the boys obtained significantly better results in manual dynamometry, the long jump, the 4 × 10 m sprint, and the endurance test (see Table 1). These results are similar to those of previous studies carried out [55,65–67]. However, when analysing the mean values of the different tests in relation to Alpha Fitness categories, the girls show the highest scores (see Figure 2). These results differ from several studies where boys score higher on these tests [66,68]. It is possible that three hours of mandatory PE classes and swimming lessons in secondary education may contribute to obtaining these results [69].

As mentioned before, MD is considered one of the healthiest dietary patterns [70], with benefits on a physical and mental level, among others [71–73]. The results of the present study show a low/poor MD adherence (14.99%) and are superior in an average (60.72%) and high adherence (24.29%). In addition, no significant differences were observed in the adherence to the diet according to the gender of the participants. These results are also superior to those of studies conducted in Mediterranean countries [74–76]. These results are different with respect to a recent study that analysed, in a similar way, the adherence to MD in non-Mediterranean countries, showing worse final results than ours with a poor (39%), medium (47.7%), and high (13.3%) adherence [73].

Several recent articles directly associate the adherence to MD with the weight and the BMI of the participants [77,78]. These data are similar to those obtained here, as the participants with the highest adherence to MD are those who also show regular weights and BMIs. Those with a low adherence to MD show a high BMI, although this is not significant. It should be noted that these values are significant when considering fat % (see Table 2), so it would be substantial to see which variable, BMI or fat %, has more importance in the three subgroups (low, medium, and high) that result from a better adherence to the MD (see Figures 4 and 5).

In contrast to the results obtained by Ozen et al. (2015) [79], which showed important differences in relation to a high and low adherence in the population analysed, with a clear tendency to abandon the MD, the results found in the present study display a tendency to maintain or even increase the MD patterns related to this type of diet, since a medium and high adherence gather 85% of the participants (see Figures 4 and 5).

These results from the KIDMED index are in line with the results obtained in other studies [77]. In addition, the results found are similar to studies carried out in southern European countries, where a large part of the sample is at a medium level of adherence to MD [80,81].

Regarding the relationship of the MD with the waist circumference, Bacopoulou et al. (2017), after studying more than 1600 subjects of a similar age to those of the present research, determined that the increase in adherence to MD was associated with a decrease in the perimeter of the waist, indicating a potential for school interventions to fight against abdominal obesity in adolescents [71]. This matches with the results of the present study and with the findings of Schröder et al. (2010), where more than 60% of the participating subjects presented a medium adherence to MD and medium, low, and very low waist circumference values [38].

As a novel aspect, the present study searched for relationships between health-related physical fitness, adherence to MD, and body composition (see Table 2 and Figures 6 and 7) in Icelandic adolescents. Significant differences were found in the tests of 4 × 10 m and endurance in the boys, and endurance in the girls with respect to those participants that show low adherence, compared with those with a medium or high adherence to MD. The disparity between genders in performance scores can be explained in the different processes of the adolescents' development. Girls experience development earlier than boys, which determines their ability to develop higher levels of strength, speed, and endurance [82].

The results mentioned in the previous paragraph are consistent with the conclusions of recent research. Muros et al. (2017), for example, found a positive relation between a high performance in the resistance test and a high adherence to MD [77]. Evaristo et al. (2018) not only demonstrated the relationship between a high adherence to MD and high levels of health-related physical fitness, but the subjects also showed high levels of health-related quality of life [70].

Despite the strength of the study, it is also important to acknowledge the limitations of the current research, which may restrict the generalisability of our findings and possible alternative interpretations. First, our data are cross-sectional and, therefore, do not enable us to infer the causal direction of our predictions. Nevertheless, they can be used as valuable indications to be considered for future research. Second, some of the data collected (KIDMED) were self-reported, which could lead to an error in the reports and recall bias due to the nature of the study. In addition, it must be borne in mind that the KIDMED questionnaire, although it was used to observe the adherence to MD does not contemplate the content and intake of nutrients consumed by the sample, which may be a confounding factor to be taken into account.
