This systematic review has been carried out to check the effectiveness of Pilates method on the components of physical fitness (cardiorespiratory endurance, muscle strength, body composition, flexibility and neuromotor fitness) in older people. The objective of this revision makes it different from other published reviews on the use of the Pilates method with older adults. Related to the above-mentioned components on physical fitness, positive evidence was observed regarding neuromotor fitness, especially static and dynamic balance. Related to the other components of physical fitness defined by the ACSM  (cardiorespiratory endurance, muscle strength, body composition and flexibility), contradictory results were observed, a fact that requires the need for further studies on those variables. Following the framework in “Results” section, the physical fitness components and the quality of the studies were analyzed in detail.
Related to such component, the four studies that analyze it do not allow us to draw clear conclusions on the possible benefit of Pilates method on this component of physical fitness. Thus, both Plachy et al.  and Vécseyné et al.  observed significant improvements in the 6MWT, but for both EGs, the one who was intervened with Pilates method and the one who was intervened with aqua fitness, with respect to CG. In both studies, the result differences between EGs are not mentioned, and there is also no mention on the control of possible variables that could be influencing the positive results, like the degree of physical fitness activity carried out outside the established intervention. This would be an important aspect to evaluate, given that both studies were the long lasting of the analyzed studies, 26 weeks of intervention. The long duration of the studies could encourage the participants in participating in other physical activities than the established intervention, something that let us to think that the observed benefits could be related to additional physical activity and not to the intervention program, as we cannot affirm that the benefits observed by Pilates method are better than the benefits by aqua fitness.
Additionally, in the study by Gildenhuys et al. , no significant differences were observed related to cardiorespiratory endurance on the 6MWT.
Those data and the fact that the studies that valued cardiorespiratory endurance were rated as low-quality studies according to PEDro scale leads us to state that there is a limited evidence on the effectiveness of the MP on cardiorespiratory endurance.
Regarding muscle strength, we could start differentiating the studies that value muscle strength of upper limbs [8, 12, 13, 45] and those that value muscle strength of lower limbs [6, 8, 13, 23, 33, 38, 45].
In the case of those studies that valued muscle strength of upper limbs, Boguszewski et al.  and Fernández et al.  did not observe a significant improvement in strength of upper limbs for EG when compared to CG. Positive results were observed for EG in the other two studies [13, 45], but in the study by Fourie et al. , a significant improvement was also observed for CG, and in the study by Vécseyné et al. , no significant differences were observed when comparing EG (Pilates method) and CG (aqua fitness group). Perhaps the explanation of these results could be the fact that the total time intervention and workload was higher in those studies in which positive results were observed. Given the results obtained and the fact that only the study by Fourie et al.  is a high-quality study according to PEDro scale, we could state that there is limited evidence on the capacity of Pilates method in developing muscle strength of upper limbs in older adults.
With regard to those studies that valued muscle strength of lower limbs, data obtained are not better, and, therefore, the evidence on the effectiveness of the method in improving muscle strength of lower limbs is also limited. In this case, some studies report significant improvements for Pilates groups [13, 23, 38, 45], but some of them report no significant improvements [6, 8, 33], or positive results were also observed for CG and/or for other EG, and in this case, no significant differences were observed between groups [13, 38, 45]. As with the results on upper limbs strength, the better results are observed in the studies with long-lasting interventions, something that suggests a minimum workload related to the improvement in muscle strength after a Pilates intervention.
With regard to this component, one study reports significant improvements in lean and fat mass and in body fat , but other studies report no significant differences in BMI  or muscle mass . The study by Gildenhuys et al.  reports no significant post-intervention results. Therefore, there are contradictory results, and also, all the studies in which this aspect is valued are low-quality studies according to PEDro scale; this is why we could state that there is no evidence of the positive effects of Pilates method in body composition on older adults.
There seems to be a moderate evidence of the benefits of Pilates method on flexibility of lower limbs. There is a high-quality study  and two low-quality studies [38, 45] which present significant improvements on flexibility after a Pilates method intervention. However, such improvement does not seem to occur in shoulder joint, except for the study by Plachy et al. . This could be caused by any factor associated to the programs developed in the different studies, but this is an unknown aspect. It is recommendable that future research describes the composition of their intervention programs based on Pilates method, enabling to compare programs and to analyze results in a proper way.
In relation to this component, it is worth noting that it is the widely aspect valued among the analyzed studies. Thus, 11 of the 17 studies assess balance in their sample [4, 6, 8, 17, 19, 22, 23, 36, 41, 43, 45]. Among them, significant benefits on static balance [6, 19, 22, 43] and dynamic balance [4, 6, 17, 23, 36, 41, 43, 45] are observed. The only study in which no significant results are observed is the one by Boguszewski et al. . For both aspects of balance, there seems to be a strong evidence about the benefits of Pilates method on older adults indicated by the two high-quality studies and other low-quality studies according to PEDro scale that confirm those positive results. Such results also could award Pilates method a preventive tool in diminishing risk of falls in older adults and consequently in diminishing fractures prevalence and their functional and social consequences.
The studies analyzed are low-quality studies in general lines (mean 3.8 ± 1.2; range between 1 and 6). On the one hand, the main items achieved by the 17 analyzed studies according to PEDro scale were those related to “Point Measure and Variability”, “Groups Similarity at baseline”, “random allocation” and “between-group comparisons” (see Table 1). On the other hand, in none of the analyzed studies, subjects and therapists were blinded, and only in one study the assessor was blinded , only in two studies the “follow-up 85 %” criterion was met [6, 33] and only three studies met the criteria “concealed allocation” or “intention to treat analysis” [4, 6, 19]. Therefore, data on quality assessment suggest that the strong points of the studies are related to the allocation of the sample and data analysis, while weak points are those related to blinding and the explicit information on dropouts and the lack of the intention to treat analysis. Although in several clinical trials it is difficult to blind subjects and therapists, it would be interesting that pieces of information on dropouts were detailed because of the possible influence of this aspect on the determination of group differences. The failure to meet the criteria of “follow-up 85 %” and “intention to treat analysis” introduces a bias that directly affects internal and external validity of any study.
In short, the low scores in PEDro scale by the analyzed studies mean their weakness in the experimental methodology, especially due to the lack of blinding, dropouts control, intention to treat analysis and allocation concealing. Such weakness is also reinforced by other aspects like the lack of standardized intervention protocols based on Pilates method (number of sessions per week, intervention duration, duration of each session, sample characteristics and if therapists are certified in Pilates method). All this could be also a reason to understand the fortitude of the evidence about the benefits of Pilates method on physical fitness components in older adults as limited.
Further to the above-mentioned on methodological quality of the studies, other limitations could be mentioned. First, samples are scarce in general. Only the study by Irez et al.  counts with 30 subjects in each experimental and CG. Additionally, in none of the studies, the calculus on the representative sample of each population was carried out, a fact that complicates the extrapolation of the observed results. In most of the studies, the sample is formed only by women [8, 13, 14, 17, 22, 23, 35, 36, 41, 43] and in one study by hospitalized patients . Some of the observed differences regarding the results could be due to sample ages. Although, in all of the studies, 65 years or older is the average age; this is also true that the ranges, in those studies in which they are indicated, are quite variable. This revision includes studies with age ranges of 65 to 74 years , age ranges of 55 to 76 , of 60 to 78  or of 62 to 80 .
Other important limitation of the analyzed studies regards the developed Pilates program. There are several differences among the programs and, also, the lack of information on the specific exercises used, something that complicates the replication and, consequently, the comparison of such studies. In one study, the duration of the intervention is unknown and, consequently, it is also unknown the workload burden to the EG . There are studies with a total intervention time from 4  or 5 weeks  to 26 weeks [38, 45]; one includes one session per week  and some include three sessions [6, 12–14, 17, 22, 23, 33, 35, 36, 38, 45]; some include a duration of the session of 30–40 min [22, 33] or 90 min ; some include a total intervention time under 20 h [6, 8, 33, 41, 43] or over 30 h [23, 36, 38, 45]. It would be appropriate for a greater uniformity on protocols in order to make possible comparisons between studies and the replication of the study.
Given the above-mentioned, future researches should include bigger and representative samples, a random and concealed allocation of subjects into experimental and CGs, blinding of subjects and assessors, dropouts control, the detailed protocol intervention (exercises, number of repetitions, duration and frequency of the sessions and the number of the total weeks intervention).