Abstract
olecular and clinical studies have linked vitamin D (vitD) de ciency to several aspects of muscle performance. For this retrospective cross-sectional study data from 297 male (M) and 284 female (F) healthy recreational athletes were used to evaluate the prevalence of vitD de ciency in athletes living in Austria and to determine whether serum 25-hydroxyvitamin D (25(OH)D) correlates with maximal (Pmax) and submaximal physical performance (P submax) measured on a treadmill ergometer. The data were controlled for age, season, weekly training hours (WTH), body mass index (BMI) and smoking status. 96 M and 75 F had 25(OH)D levels 20 ng/mL. 25(OH)D levels showed seasonal variations, but no seasonal differences in Pmaxand P submaxwere detected. M with 25(OH)D levels 20 ng/mL had signi cantly lower P submax(p= 0.045) than those with normal levels. In F no signi cant differences in Pmaxor P submaxwere detected. Stepwise multiple regression analysis including all covariates revealed signi cant correlations between 25(OH)D levels and Pmax( = 0.138, p= 0.003) and P submax( = 0.152,p= 0.002) in M. Interestingly, for F signi cant correlations between 25(OH)D and both Pmaxand P submaxdisappeared after adding WTH to the model. In conclusion, our data suggest that 25(OH)D status is associated with physical performance especially in M, while in F, WTH and BMI seem to affect the correlation. Keywords: vitamin D; maximal performance; submaximal performance; 25-hydroxyvitamin D (25(OH)D); physical activity; treadmill ergometer; athlete 1. Introduction The discovery of vitamin D and its role in musculoskeletal pathways at the end of the 20th
conclusion, our data suggest that 25(OH)D status is associated with physical performance especially in M, while in F, WTH and BMI seem to affect the correlation. Keywords: vitamin D; maximal performance; submaximal performance; 25-hydroxyvitamin D (25(OH)D); physical activity; treadmill ergometer; athlete 1. Introduction The discovery of vitamin D and its role in musculoskeletal pathways at the end of the 20th century has led to a signi cant decline in cases of rickets in children, with the awareness about vitamin D de ciency and its consequences getting a second boost at the moment [1]. Vitamin D is known for its critical role in musculoskeletal health through maintaining a homeostasis of calcium and phosphate by enhancement of their absorption in the small intestine [2] thus preventing bone fractures and falls, especially in older populations [3,4]. Moreover it has been proposed to play an important part in reducing the risk of multiple types of cancer [5,6] as well as several autoimmune diseases [7,8] and osteoarthritis [9]. A recent meta-analysis of observational studies by Ekmekcioglu et al. [6] showed that higher 25(OH)D levels are associated with lower risks for type 2 diabetes mellitus and colorectal cancer. Additionally vitamin D de ciency was shown to be associated with higher blood pressure [10], obesity [11] and higher blood lipid levels [12]. Int. J. Environ. Res. Public Health2018,15, 2724; doi:10.3390/ijerph15122724
Int. J. Environ. Res. Public Health2018,15, 2724 2 of 14 Out of the many forms in which vitamin D presents itself, vitamin D2(ergocalciferol) and vitamin D3(cholecalciferol) are the physiologically most important ones. While vitamin D3can be found in fatty sh, egg yolks, liver and dairy products, mushrooms are a source for vitamin D2[13]. The main source for vitamin D however, is the synthesis by the skin in reaction with UVB light. The metabolic steps in the liver and kidney leading to formation of the main circulating metabolite 25-hydroxyvitamin D (25(OH)D) and the active metabolite 1,25-dihydroxyvitamin D (1,25(OH)2D) are regulated through multiple feedback mechanisms and are still under investigation [14]. According to recent ndings the enzymatic systems needed to perform the activating hydroxylation step as well as the nuclear hormone receptor, vitamin D receptor (VDR), were detected in several cells all over the body, including muscle tissue [15]. Subsequently it is assumed that 1,25(OH)2D also in uences skeletal muscle via molecular pathways [16].This hypothesis was tested in numerous observational and interventional studies. Most of the studies have been conducted in elderly populations, linking vitamin D de ciency to several aspects of muscle strength and performance such as handgrip, lower limb strength, balance, timed up and go test and gait speed [1719]. However, also divergent results have been reported, showing no improvements in strength parameters following vitamin D supplementation in elderly populations [20,21]. Especially in the last two decades there have been increased efforts to determine the role of vitamin D in physical performance in younger populations and athletes. Vitamin D levels have been associated with various aspects of muscle strength and recovery in these populations as well, including for example hand-grip strength, gastro-soleus strength and walking distance in a group of young vitamin D de cient Asian Indians [22], hand-grip strength in a group of severely vitamin D de cient Somali women [23], isometric strength and relative grip strength in young hockey players [24], vertical jump and reduced injury risk after winter vitamin D supplementation in elite ballet dancers [25], 10 m sprint times and vertical jump height following supplementation in
young vitamin D de cient Asian Indians [22], hand-grip strength in a group of severely vitamin D de cient Somali women [23], isometric strength and relative grip strength in young hockey players [24], vertical jump and reduced injury risk after winter vitamin D supplementation in elite ballet dancers [25], 10 m sprint times and vertical jump height following supplementation in a sample of vitamin D de cient young athletes [26] as well as improved phosphocreatine recovery half-time of the soleus muscle in severely de cient individuals [27] and enhanced recovery following intense exercise in active male adults [28]. Again, several studies have yielded contradictory results showing no association between vitamin D levels and muscle strength parameters [2932]. In accordance with the rising interest on vitamin D, study groups all over the world have conducted research to determine the prevalence of vitamin D de ciency, with alarming results. According to Hilger et al. [33] 88.1% of examined samples worldwide showed mean 25(OH)D levels beneath 75 nmol/L (30 ng/mL; conversion factor: 1 ng/mL = 2.496 nmol/L), 37.3% with mean levels <50 nmol/L (20 ng/mL) and 6.7% with mean levels <25 nmol/L (10 ng/mL). Similar prevalence rates have been found for athletes [34]. However, a huge variability in all the collected data partly limits the signi cance of possible estimates on vitamin D levels. Considering that there has not been a universal consensus on optimal vitamin levels yet, the comparison of studies investigating the results of insuf cient vitamin D status on different health outcomes is particularly challenging. While the US Institute of Medicine Committee (IOM) proposes that 20 ng/mL (50 nmol/L) as a cut-off level meets the needs of almost all of the general population [35], the US Endocrinology Society regards any 25(OH)D levels under 30 ng/mL (75 nmol/L), and above 20 ng/mL (50 nmol/L) as insuf cient [36]. The aim of this trial was to evaluate the prevalence of vitamin D de ciency in a large sample representing healthy recreational athletes in Austria and to determine whether vitamin D status correlates with both, maximal and submaximal physical performance on a treadmill
25(OH)D levels under 30 ng/mL (75 nmol/L), and above 20 ng/mL (50 nmol/L) as insuf cient [36]. The aim of this trial was to evaluate the prevalence of vitamin D de ciency in a large sample representing healthy recreational athletes in Austria and to determine whether vitamin D status correlates with both, maximal and submaximal physical performance on a treadmill ergometer. 2. Materials and Methods The study was approved by the ethics committee of the Medical University of Vienna (EK-Number 1929/2016). The data used for this study were generated in Sportordination, a multidisciplinary medical centre for sports medicine and sports sciences, located in 1080 Vienna, Austria. On their rst visit all athletes had to ll out a questionnaire on their medical history, weekly training hours and
Int. J. Environ. Res. Public Health2018,15, 2724 3 of 14 personal best results at races. Following the medical check-up height, body weight and body fat were measured. Afterwards athletes performed an incremental performance test on a treadmill ergometer to subjective exhaustion supervised by a sports scientist. To ensure valid test results, all athletes were asked to refrain from doing any sports, drink enough uid and have a full carbohydrate-based supper on the day prior to the performance test and to make sure that their last food intake was two hours before the test. According to the gathered information on training habits and race performance, the tting protocol for the performance test was chosen. For athletes with half marathon minimum times under 2 h or 10 km minimum times under 60 min the following protocol was used: Initial running speed started at 6 km/h with increments of 2 km/h every 3 min. This protocol has been proven suf cient for incremental performance tests of amateur athletes [37,38]. At the beginning of the test blood lactate and glucose concentration were determined from capillary blood drawn from the earlobe at rest using SUPER GL ambulance (Dr. Müller Gerätebau GmbH, Freital, Germany), a glucose and lactate measurement unit working with a compact sensor. During the test, the treadmill was stopped after each exercise level to determine lactate and glucose concentrations. Heart rate was measured continuously using telemetered electrocardiogram (ECG) recordings and blood pressure was checked regularly. Pmaxwas de ned as highest possible speed (km/h) until subjective exhaustion. While maximal performance represents the overall tness, submaximal performance is far more important to athletes as it determines their training status. During an incremental performance test lactate levels form a characteristic curve [39,40] with a slight increase at the beginning due to accumulating lactate as long as heart rate and capillary dilatation have not adapted, and then stabilization just above baseline when oxidative phosphorylation is responsible for the mean part of energy production. There is a rise in lactate levels as the intensity increases and anaerobic glycolysis begins to take part in ATP production. Within this transition
at the beginning due to accumulating lactate as long as heart rate and capillary dilatation have not adapted, and then stabilization just above baseline when oxidative phosphorylation is responsible for the mean part of energy production. There is a rise in lactate levels as the intensity increases and anaerobic glycolysis begins to take part in ATP production. Within this transition between aerobic and anaerobic energy production oxidative energy supply is no longer suf cient to keep up with the rising need for ATP. The capacity of mitochondria is pushed to its limits resulting in exponential rise in lactate levels as pyruvate production exceeds lactate clearance marking the individual anaerobic threshold (IAT). By measuring glucose levels in addition to lactate only we were able to be more precise in continuously determining which energy substrate was used. We calculated IAT using the concept of Dickhut et al. [41], one of several lactate threshold concepts that can be used to determine submaximal performance and has been proven suf cient for performance diagnostics [42,43]. All healthy recreational athletes aged 18 to 65, who had performed their performance diagnostic test on a treadmill ergometer between 1 July 2013 and 1 February 2017 with valid data on serum 25(OH)D levels were included in the study. 25(OH)D levels were tested in separate laboratories individually prior to the performance test. The main method used was the protein binding immunoassay using Elecsys ® Vitamin D total II (Roche Diagnostics AG, Rotkreuz, Schweiz), which has been shown to accurately measure 25(OH)D levels [44]. Patients who had had their blood taken over three months prior to their performance test were excluded. All patients who had performed their performance test on a cycle ergometer were excluded. To ensure valid data, only athletes following the protocol mentioned above were included in this study. A total of 581 healthy athletes (284 females and 297 males) were included (Tables). They were considered healthy as long as there were no medical contraindications for stress tests such as unstable cardiac disease or acute in ammatory disease [45,46]. Out of these only 224 female and 253 male athletes
following the protocol mentioned above were included in this study. A total of 581 healthy athletes (284 females and 297 males) were included (Tables). They were considered healthy as long as there were no medical contraindications for stress tests such as unstable cardiac disease or acute in ammatory disease [45,46]. Out of these only 224 female and 253 male athletes had their body fat measured. Body fat was therefore not included in further statistical analyses. Apart from body fat levels the data of all the athletes were complete.
Int. J. Environ. Res. Public Health2018,15, 2724 4 of 14 Table 1.Descriptive statistics of all female athletes. N Minimum Maximum Mean SD Variance Age 284 18 65 38.66 9.82 96.51 Bodyweight (kg) 284 45.1 99.4 63.42 9.12 83.24 Height (cm) 284 152 188 166.91 6.13 37.62 BMI (kg/m 2 ) 284 17.5 34.2 22.76 2.84 8.09 25(OH)D (ng/mL) 284 7.5 66.5 27.17 10.89 118.57 Weekly training (h) 284 0 15 5.20 1.53 2.34 Pmax(km/h) 284 6.7 18.7 12.94 1.96 3.83 P submax(km/h) 284 5.3 16.2 10.02 1.40 1.95 BMI: body mass index; 25(OH)D: 25-hydroxyvitamin D; maximal (Pmax) and submaximal physical performance (P submax). Table 2.Descriptive statistics of all male athletes. N Minimum Maximum Mean SD Variance Age 297 18 64 40.54 9.16 83.95 Bodyweight (kg) 297 47.6 149.4 80.12 11.43 130.71 Height (cm) 297 155.0 198.0 180.40 6.71 45.04 BMI (kg/m 2 ) 297 18 41.3 24.53 2.81 7.94 25(OH)D (ng/mL) 297 5.2 64.5 24.80 10.15 103.09 Weekly training (h) 297 2 20 6.21 2.21 4.89 Pmax(km/h) 297 8.7 22.0 15.81 1.98 3.92 P submax(km/h) 297 7.1 17.0 11.85 1.59 2.523 BMI: body mass index; 25(OH)D: 25-hydroxyvitamin D; maximal (Pmax) and submaximal physical performance (P submax). Several statistical tests were performed to test whether vitamin D levels are associated with Pmaxor P submax. Males and females were analysed separately because the expected high variance in performance parameters would distort the results for possible correlations with 25(OH)D. For further statistical analysis calculated BMI levels were divided into two groups of normal (BMI < 25 kg/m 2 ) and overweight/obese (BMI 25 kg/m 2 ), and a cut off value of 25(OH)D levels of 20 ng/mL [35], was used to separate athletes according to their vitamin D levels into groups with suf cient and insuf cient vitamin D status. A two-tailed statistical signi cance was accepted at thep< 0.05 level. Seasonal differences in 25(OH)D levels and Pmaxor P submaxwere calculated using an ANOVA. Unpairedt-tests were used to assess the association of Pmaxand P submaxbetween groups of suf cient (serum 25(OH)D > 20 ng/mL) and insuf cient (serum 25(OH)D 20 ng/mL) vitamin D
cient and insuf cient vitamin D status. A two-tailed statistical signi cance was accepted at thep< 0.05 level. Seasonal differences in 25(OH)D levels and Pmaxor P submaxwere calculated using an ANOVA. Unpairedt-tests were used to assess the association of Pmaxand P submaxbetween groups of suf cient (serum 25(OH)D > 20 ng/mL) and insuf cient (serum 25(OH)D 20 ng/mL) vitamin D status and to determine differences in Pmax, P submaxand serum 25(OH)D levels between athletes with normal BMI and overweight athletes as well as differences in Pmaxand P submaxbetween smokers and non-smokers. Additionally, an ANCOVA was performed to test for in uencing variables (age, BMI, WTH and smoking status). Simple associations between primary, secondary and exploratory parameters were tested by Pearson correlations. Finally, stepwise multiple regression analyses were performed to determine whether the obtained exploratory parameters affected the correlation between vitamin D levels and performance. Four models were created for Pmaxand P submaxin male and female athletes, respectively. After performing regression analysis for 25(OH)D as an independent and Pmaxand P submaxas dependent variables, consecutively, the following parameter were added to the model: age, weekly training hours, BMI and nicotine abuse. There were no outliers in the data as assessed by inspection of boxplots. Most of the parameters were normally distributed as veri ed by the Kolmogorov-Smirnov test and inspection of histograms. For those parameters that did not show normal distribution additionally non-parametric tests were performed for comparison, with similar results. All analyses were performed with IBM©SPSS software (Version 23, IBM Inc., Armonk, NY, USA).
Int. J. Environ. Res. Public Health2018,15, 2724 5 of 14 3. Results Female athletes showed non-signi cantly higher serum 25(OH)D levels than male athletes (Figure). 75 females and 96 males had 25(OH)D levels 20 ng/mL. Therefore, according to the cut off levels de ned by the IOM out of all the included athletes 26% of females and 32% of males would have an insuf cient vitamin D status. However, if our data were analysed according to the US Endocrinology Society guidelines 66% of females and 76% of males would have insuf cient vitamin D levels ( 30 ng/mL). Figure 1. Mean serum 25(OH)D (25-hydroxyvitamin D) levels in females (f) with 27.17 ng/mL were found to be non-signi cantly higher than mean 25(OH)D levels in males (m) with 24.80 ng/mL. As expected, our results showed that 25(OH)D levels varied signi cantly between seasons with lowest values in winter and highest values in summer in female (F = 11.538,p= 0.000) and male athletes (F = 22.345,p= 0.000), respectively (Figure). Furthermore, our results showed that 57% percent of our athletes had vitamin D levels 30 ng/mL even during the summer months of July, August and September, however only 8% had levels 20 ng/mL. Figure 2. Signi cant variations in 25(OH)D (25-hydroxyvitamin D) levels between seasons (p= 0.000). There were no seasonal differences for Pmaxand P submax(data not shown) but male athletes with suf cient (>20ng/mL) 25(OH)D levels showed signi cantly higher P submaxlevels compared to those
Int. J. Environ. Res. Public Health2018,15, 2724 6 of 14 with levels 20 ng/mL (Figure). In females the difference in P submaxwas insigni cant (p= 0.08). Any detected differences between vitamin D groups ( 20 ng/mL vs. > 20 ng/mL) in Pmaxand P submax for both male and female athletes were insigni cant when controlled for modifying variables (age, BMI, WTH, smoking status) (data not shown). Figure 3. Signi cant differences (p= 0.045) in submaximal performance between groups of suf cient vitamin D status (25(OH)D > 20 ng/mL) and insuf cient vitamin D status (25(OH)D 20 ng/mL) in males but not in females. 25(OH)D: 25-hydroxyvitamin D; * submaximal performance signi cantly higher in male athletes with 25(OH)D levels >20 ng/mL. Concordant with other ndings [47] Pmaxand P submaxwere signi cantly lower in overweight males compared to those with normal BMI (p< 0.001) with no signi cant difference in 25(OH)D levels between the two groups (data not presented). In female athletes we found signi cant differences in Pmax, P submaxand 25(OH)D (Table). Table 3. Differences in Pmax, P submaxand 25(OH)D levels between female athletes with normal and high BMI. F p-Value BMI (kg/m 2 ) N Mean SD Pmax(km/h) 5.075 0.000 <25 (normal) 234 13.23 1.78 25 (overweight) 50 11.57 2.19 P submax(km/h) 2.384 0.000 <25 (normal) 234 10.2 1.32 25 (overweight) 50 9.21 1.49 25(OH)D (ng/mL) 0.445 0.012 < 25 (normal) 234 27.91 10.68 25 (overweight) 50 23.67 11.29 t-test for BMI (normal and overweight) groups of normal weight (<25) and overweight ( 25) as independent variables; 25(OH)D: 25-hydroxyvitamin D. No signi cant differences in Pmaxand P submaxwere found between smokers and non-smokers in male and female athletes (data not shown). In female athletes signi cant correlations at thep= 0.05 level were found for 25(OH)D and Pmax (r = 0.143) as well as for 25(OH)D and P submax(r = 0.141), while in male athletes only the correlation between 25(OH)D and P submax(r = 0.139) reached the level of signi cance. Even though signi cant correlations were shown between 25(OH)D and Pmaxand P submaxin female athletes and between 25(OH)D and Pmaxin
Description
This study evaluates the correlation between vitamin D levels and physical performance in recreational athletes.