Osteoporosis is characterized by a deterioration of bone microarchitecture, low bone mass and an increased risk of fractures affecting mainly menopausal women.1 The prevention of this disease at early ages of life is a determining factor in reducing the risk of fracture in the elderly.2 Several studies have explored the osteogenic effect of physical activity.3–7 Adolescents engaged in sports showed lower incidence of traumatic fractures.2 Moreover, adults engaged in non-professional sports in early life have higher BMD than their inactive peers, especially women.8 Particularly, sports involving high level impact from ground reaction forces are more osteogenic than non-weight bearing activities, like cycling and swimming. Elite water sports9 can lead to an osteoporosis risk situation when started at earlier ages, or at any case not improving the bone consolidation for the senescence. At the elite level, artistic swimmers (AS) underwent a greater amount of water training hours per week for many years: eight to ten pool sessions per week, with speed swimming and artistic swimming-specific skills. The effect this may have on the future development of osteoporosis has not been defined.10,11 To the best of our knowledge, studies regarding BMD in highly trained older women athletes in aquatic sports have not been explored thorough fully before.

Several research related to different sports has tested the positive effect of a plyometric jump training program on premenopausal women's bone,12 even when osteopenia is already established.13 A high-intensity, low repetition, short-lived, multidirectional jump program is also very effective in producing an osteogenic response, which can be introduced into daily life.14 It is important to find out jump programs that can counteract the potential negative consequences of non-osteogenic sport on bone health. The unprecedented worldwide confinement, caused by COVID-19, in which Spain has been one of the most affected-with severe rules governing confinement may have changed physical activity and sedentary habits due to prolonged stays at home.15

In an attempt to resolve these issues, ex artistic swimmers were included in a two –phases study (I) to determine the effect of being involved in elite artistic swimmers during their youth on bone mineral density (BMD), bone mineral content (BMC) and body composition (BC) compared with sedentary controls; and (II) to evaluate the effects of a jump program on bone mineral density and body composition in a group of ex-Artistic Swimmers.

The present investigation includes two phases, first (I) a cross-sectional observational study that compares bone variables in ex artistic swimmers with sedentary controls and (II) a randomized controlled trial, pre-post intervention jump program in a group of ex artistic swimmers during the COVID-19 confinement.

Study design, protocol and informed consent forms were reviewed and approved by the Ethics and Clinical Research Committee of the Consorci Sanitari de Terrassa (CST) and met the requirements of the Declaration of Helsinki for research on human beings. Informed written consent was obtained from all participants included in the study.

The characteristics of the groups of the phase I are summarized in Table 1. There were no significant differences in age or body weight between groups. Artistic swimmers were taller (3.7%, ES= 1.192, p = 0.001); they had lower body fat percentage (4.6%, ES= 0.438, p = 0.001); and lower fat mass (13.6%, ES= 0.531, p = 0.038) than the sedentary control group.

Artistic swimmers and age-matched controls were into peri-menopause and without oestrogen replacement therapy.

Table 2 shows physical activity levels for a period of more than 15 years. The artistic swimmers had significantly higher activity levels during their time of competition compared with the sedentary control group (p < 0.001). Water training was about 23 hours/week. The fitness routines included three-sessions/ week for 1 hour 30 minutes. Not all the swimmers continued to practice sport. The sports practiced since their retirement were Pilates, running, swimming; they were only active about 2 hours per week. The control group reported being active at gymnasium for about 1-2 hours per week. At the time of this study, there were no significant differences in activity levels between artistic swimmers and the sedentary control group. No significant differences in current life styles were found between groups.

Table 3 shows regional BMD and BMC of the both groups. Z-score and T-score of total BMD (ES= 0.932; p = 0.001 and ES= 0.626; p = 0.016, respectively) and trunk's BMD (ES= 0.754; p = 0.002) in the artistic swimmers group were significantly higher compared with the sedentary group. Artistic swimmers had significantly higher Z-score of the femoral neck too (ES= 0.352; p = 0.046). No differences between both groups were found in total BMD, arms’ BMD and legs’ BMD. BMD of the lumbar spine was significantly lower (ES= 0.594; p = 0.017) in the control group compared with the artistic swimmers group. Further, the artistic swimmers had significantly higher total BMC (ES= 0.769; p = 0.002) and arms’ BMC (ES= 1.108; p = 0.001) than the control group. No differences were found in legs’ and trunk’ BMC.

Artistic swimmers had lower total body fat mass (ES= 0.531; p = 0.038), trunk (ES=0.708; p = 0.007), ginoid (ES= 0.818; p = 0.001) and android regions (ES= 0.823; p = 0.002). Fat free mass was significantly lower in the control group (ES= 0.438; p = 0.075) with significantly differences in arms (ES= 0.844; p = 0.001), legs (ES= 1.068; p = 0.001), ginoid (ES= 2.055; p = 0.001) and android region (ES= 0.804; p = 0.003) (Table 4).

In phase II, sample size reduced to 19 participants due to personal reasons during the COVID-19 pandemic and two participants reported episodes of tibia periostitis due to jump training and didn't continue the study. The characteristics of the groups are summarized in Table 5. Subjects were randomly assigned to one of the both jumping groups during COVID-19 confinement: the CMJ group (n = 9) and the non-intervention group (n = 10). The intervention compliance was 92%.

There was a significant difference between the non-intervention group and the CMJ group for femoral neck’ BMD in the pre intervention in favour of those that do not jump (ES=1.154; p = 0.023), which was maintained in the post intervention (ES=1.430;p = 0.006); and for femoral neck’ BMC in the pre intervention (ES=1.136;p = 0.032) and post intervention (ES=1.621;p = 0.008).After the intervention period, the non-intervention group presented a significant decrease in femoral neck’ BMC (ES=0.288; p = 0.019) and a significant increase in fat free mass (ES=0.138; p = 0.001) and in FM (ES=0.96; p = 0.001). Furthermore, CMJ group showed a significant increase in Z-score lumbar spine BMD (ES=0.422; p = 0.003) and a significant decrease in fat free mass (ES=0.130; p = 0.001).

During the jump assessment before starting the intervention program, significant differences were found between the reactive jump and the final landing in the PVRF (ES= 1.363; p <  0.001) and in the ILR (ES= 0.83; p = 0.002). Participants obtained an average of PVRF of 4.08 ± 0.90 BW with an ILR of 45.86 ± 17.08 BW/s in the reactive jump and an average of PVRF of 3.06 ± 0.71 BW with an ILR of 35.72 ± 11.33 BW/s in the final landing.

The main finding of the present study was that the long-term exposure to high level artistic swimming training at a younger age produced later significant increase in Z and T-score of Total BMD, Lumbar Spine BMD and Z-score of Femoral Neck BMD compared with controls of similar age and menopausal status. It is important to highlight that lumbar spine and femur measurements are two assessment measures according to the WHO classification for diagnosing osteopenia and osteoporosis 19 and specifically related to the most susceptible sites for osteoporotic fracture. The specific improvement in these areas is probably due to the fact that artistic swimming involves the execution of repeated exercises with an important participation of core and hip muscles.23 Previous studies have suggested a site–specific skeletal response to the type of loading at each BMD site.24 Tveit study25 shows that after 20 to 29 years of retirement from sport, former male soccer players had a higher whole and legs’ BMD than controls, similar to Mantovani study,8 in line to the present study for whole BMD. These results were related with the special biomechanical characteristics of soccer as changes of direction, speed, jump and kicks that offer additional mechanical stress in lower extremities, different from our artistic swimmers. In fact, a previous study conducted by us in young Olympic athletes, aquatic athletes (swimmers, artistic swimmers, water polo players) and non aquatic ones (football, volleyball, field hockey players)10 conclude that mean BMD values were highest in non aquatic sports athletes at all measured sites compared to aquatic athletes. However, all high level sports participants, aquatic and non-aquatic, had higher values than the sedentary controls, in lumbar spine and femur measures, in line with our current study in adult women.

Despite our positive results in lumbar spine and femoral neck, we were concerned about the bone health of the ex elite aquatic athletes. They showed similar results than the sedentary group in total BMD, arms’ BMD and legs’ BMD. In swimmers, low BMD accrual and/or increased bone resorption can occur silently over time, leading to osteoporosis without developing a stress fracture during their competitive years.26 The main factors that determine adult bone health are peak bone mass density at skeletal maturity and the rate of bone loss with advancing age; therefore, maximizing premenopausal BMD should be a critical strategy for the prevention of osteoporosis in this population.14

The artistic swimmers included in this study presented greater fat free mass (FFM) arms and legs and lower total and trunk fat mass (FM) than the control group, in line with the literature.8,24 Probably, these results are related to the fact that Artistic swimming training involves the execution of repeated exercises with a variety of movements of arms and legs to accelerate, decelerate, rotate, and turn to pull the body out of the water and even jump. Thus, artistic swimmers could develop a large muscle mass in extremities that increase bone mass in legs and femoral neck, a situation that occur in our population. On the contrary, other studies found greater FM in former soccer players and similar FFM than controls.25 Legs were more affected by sports participation probably because these body segments are more exposed to biomechanical ground reaction forces generated by the sport. As well as, fat free mass was the most relevant determinant of all-bone related outcomes. It is stated that FFM greatly influence BMD and BMC in adolescent women.27 Muscle strength applied to bone could be an important predictor of BMD,4,28 even in water sports. Sports participation in early life positively affects muscle mass gains, mainly in adolescence, and these gains in muscle mass could affect BMD through life, in line with these women athletes. The present study provided some evidence that participation in high-level aquatic sports during early life is not as adverse for the bone in adulthood as hypothesized in previous studies in swimmers.29–31 The discrepancy between swimmers and artistic swimmers could be attributed to training differences related to the greater mechanical loading from repetitive pushing against the pool wall in speed swimming,32 propulsive strokes and resistance against water30 and maybe also the related to training exercises out of the water.

As we have previously mentioned, one of the main factors that determine adult bone health is the rate of bone loss with age. Women are more susceptible to bone loss during menopause, mainly because of significant hormonal changes and a decline in exercise solicitation. Intervention approaches should aim to foster more positive attitudes to aging and retirement and promote physical activity at all stages in life.33,34 In line with our concern for the bone health of our retired artistic swimmers we designed phase II that matched with Covid-19 confinement. Results of the intervention program applied on the Phase II showed a decrease in femoral neck BMC’ in the non-intervention group. Furthermore, the CMJ group showed an increase in Z-score lumbar spine BMD’, in line with the literature.21 These improvements may be related to the high-impact tasks performed during the intervention program. The magnitudes of the PVRF (4.08 BW's) and the ILR (45,86 BW/s) for the reactive jumps after the first landing of the CMJ, exceed the defined osteogenic thresholds previously shown to improve bone mass at clinically relevant sites for premenopausal women (>3 BW's and 43 BW•s-1). In line with these results, previous studies with bouts of exercise designs (short duration, high-impact) similar to the present study found improvement in BMD on femoral neck.35–37 A recent study conducted by us showed that after 22 weeks of a jumping rope and whole body vibration program, only 20 min/day, twice week, the olympic artistic swimmers team increased lumbar spine and femoral neck BMD’ compared to the same group of athletes without intervention program.38 It has been difficult to follow the program during COVID-19 confinement, in line with a study that showed that the Spanish adult population decreased daily self-reported physical activity and increased sedentary time during COVID-19 confinement. The lockdown and social distancing caused by COVID-19 has influenced common health behaviour. Vigorous physical activities and walking time decreased by 16.8% and 58.2%, respectively, whereas sedentary time increased by 23.8%.15

The present study has several limitations. The first one is the relatively small sample size in phase I and phase II, determined by the small number of former elite artistic swimmers. Years ago artistic swimming did not have many practitioners. The difficulty in the recall bias from collecting information on physical activity levels from 15 years ago may exist. In phase II, the main limitation is the compliance with the intervention exercise program with the COVID-19 confinement.

On the other side, the present study has also several strengths. Our study is distinctive in the way that it presents lifetime bone trait evaluation in high level artistic swimmers, who trained regularly at a high intensity, volume and frequency, the perfectly standardized sport performed during early life, unthinkable in general population where everyone has dietary and exercise habits and different lifestyles. There are few studies in the literature that have taken into account the physical activity levels, bone variables and body composition of former elite artistic swimmers. The current study presents a high-intensity, low repetition, multidirectional jump program that could be introduced into daily life.

Professional sport participation during childhood and adolescence was positively associated with bone health in adulthood, even in aquatic sports, and specifically in artistic swimmers.

Short bouts of home exercise should be considered to reduce the risk of bone lost in retired artistic swimmers. After 8 months of high-impact jump training, BMD can be maintained or improved in premenopausal women.

No financial or material support of any kind was received for the work described in this article. No funding received.

All authors contributed to the study conception and design. MB wrote the first draft of the manuscript and all authors commented on previous versions of the manuscript. MB carried out the bone densitometry studies, analysed the results and drafted the manuscript; FD conceived and coordinated this study, participated in its design and reviewed the draft; VF and EJ helped in the analyses of the results and reviewed the draft; LD participated in the bone densitometry studies; LG controlled the jump program and bone densitometry studies; VF and AT reviewed the document. All authors contributed to the editing and finalization of the manuscript. They have read and approved the final version of the manuscript, and agree with the order of presentation of the authors.