4 Results

No significant difference was noted between the profile of mood (POMS) questionnaire of the pre (the month prior to the study) and the test month for all women in the profiles during the different menstrual phases. The time course of the profiles including the sub scales and the overall score showed no differences during the months (figure 4-1).

	Figure 4-1: Similar profiles of mood state (POMS) separated into depression, fatigue, vigor and anger (subscales of the POMS) in the pre and the test month of women; no significance was noted

There was no significant gender difference in the overall score of all subjects (trained and untrained men and women) of the study month (figure 4-2) whereas differences between trained and untrained subjects and differences between male and female athletes and sedentary subjects were noted.

	Figure 4-2: Overall score of POMS in pretest and test month of women compared with men; overall scores were not significantly different in men and women

Trained compared with untrained men had significantly higher sub scores in depression (p<0.01) and significantly higher sub scores in fatigue (p<0.05). Still the overall score was similar in trained and untrained men. Trained and untrained women were similar in sub scores and the overall score.

In the trained group, women had significantly higher sub scores in depression, fatigue and anger compared with the male athletes (p<0.001); still the overall score was similar. In the sedentary group, women had significantly higher sub scores in depression and fatigue than the untrained men but with comparable overall scores (p<0.01).

The results of the blood borne parameters which include the hormonal fluctuations in women, the comparison of blood glucose and insulin in trained and untrained subjects and the electrolytes are presented in the following chapters.

All women had a normal ovulatory cycle. No difference between trained and untrained women was noted. Female athletes did not show affected menstrual cycle phases and/or modulated hormonal concentrations compared with the sedentary females. Despite typical hormonal fluctuations, some women had levels of LH and FSH which were constantly below the 95% range. The reference ranges of LH, FSH, E2 and P was generated by DPC Biermann. The hormonal concentrations were significantly different during the five days of the menstrual cycle (p<0.001).

The menstrual cycle monitoring is based on hormonal levels measured at M (menstruation), MidF (middle of follicular phase), O (ovulation), MidL (middle of luteal phase) and PreM (pre menstruation phase) in figure 4-3 and 4-4.

Total testosterone (tT) and SHBG were analysed whereas the free androgen index (fAi) and the percentage of free testosterone (fT) were calculated based on tT value. No significant differences in tT, SHBG, fAi and fT was found during the study month. The hormonal levels of trained and untrained men were individually different but similar during the study month. Therefore, men did not have any hormonal fluctuations in the course of one month.

	Figure 4-3: LH and FSH concentration all women; LH and FSH peaks are noted around the ovulation day (O)

	Figure 4-4: P and E2 concentration of all women during five phases of the menstrual cycle; E2 peak around ovulation (O) combined with a constant increase of P

Insulin (INS) levels were similar between the sedentary subjects and the athletes without gender differences. Nevertheless trained individuals had lower INS levels than untrained subjects whereas the INS levels were stable in course of the study month in men and women (figure 4-5).

In contrast to INS, the blood glucose (BG) was similar in trained and untrained subjects. No fluctuation of BG was found in course of the study month in men and women (figure 4-6).

The five menstrual cycle phases of women are identical with the five study days of men characterized as follows; M=1, MidF=2, O=3, MidL=4, PreM=5.

	Figure 4-5: INS concentrations of trained and untrained men and women during the study; lower INS levels in trained and higher ones in untrained subjects were noted

	Figure 4-6: BG of athletes and sedentary men and women during the study; similar levels were noted in athletes and sedentary subjects

Electrolytes Na+, K+, Ca²+, Mg²+ and Cl-, the white (WBC) and red blood cells (RBC), haemoglobin (Hb) and hematocrit (Hc) were in the normal range and did not fluctuate in course of the study month. No difference was noted between athletes and sedentary subjects and between men and women with exception of the hematocrit which was lower in women than in men as presented in figure 4-7. Figure 4-8 till 4-12 present the course of Na+, K+, Ca²+, Mg²+ and Cl- during the study and between men and women.

As mentioned above, the five menstrual cycle phases of women are identical with the five study days of men characterized as follows; M=1, MidF=2, O=3, MidL=4, PreM=5.

	Figure 4-7: The stable course of the Hc of men and women in % during one month i.e. one menstrual cycle; women have higher Hc (lower %) than men

	Figure 4-8: Similar Ca+ levels in men and women during one month; no menstrual cycle related fluctuations were noted

	Figure 4-9: Cl-levels of men and women in course of one menstrual cycle; no differences were noted

	Figure 4-10: Similar K+ concentrations of male and female subjects during five study days

	Figure 4-11: No significant difference was noted in Na+ levels between men and women and during the study

	Figure 4-12: Mg²+concentration in men and women without any significant differences between genders and in course of the study month

The breathing frequency was different within the groups and during the intervention days. Breathing frequency (BF) of men was significantly lower than the BF of women (p<0.01).

Male athletes had a significantly lower BF than the sedentary ones (p<0.001), but the BF were stable throughout the five study days (table 4-1 and figure 4-13).

Table 4-1: Breathing frequency (BF) of men and women

	Figure 4-13: Breathing frequency was significantly lower in trained than in untrained men (p<0.001) without any significance in course of the study

Women showed the opposite; untrained women had significantly lower BF than trained women (p<0.01); results are presented in table 4-1. Nevertheless, the BF of trained women reacted different during the menstrual cycle whereas the BF of the sedentary females was stable in course of the study. Table 4-2 illustrates the BF throughout the five menstrual cycle phases in trained and untrained women.

Female athletes had significantly higher BF in MidL (p<0.05) and in PreM (p<0.05) compared with the M which is demonstrated in the following table and illustrated in figure 4-14.

Table 4-2: Breathing frequency in course of the menstrual cycle in trained and untrained women

	Figure 4-14: The median of the breathing frequency of untrained women was significantly lower compared with trained women (p<0.01); still the athletes showed significantly enhanced BF in MidL (p<0.05) a in PreM (p<0.05) compared to M whereas the untrained women did not show any significance in course of the menstrual cycle

The results of the heart rate variability (HRV) recordings are presented in the time and the frequency domain at rest, during the menstrual cycle and during the orthostatic test. Finally, the orthostatic test is presented in the five menstrual cycle phases.

The HRV at rest was different in athletes compared with the sedentary subjects. Differences between male and female athletes were noted whereas the sedentary men and women were similar. The results are presented first in the time and second in the frequency domain.

The meanNN was significantly higher in men than in women (for athletes p<0.01, for sedentary subjects p<0.001). Male and female athletes had significantly higher meanNN than the sedentary men and women (for males p<0.05, for females p<0.001).

The SDNN, RMSSD and pNN50 were significantly higher in trained men and women compared with the untrained subjects (SDNN, RMSSD for men p<0.01, pNN50 for men p<0.05 and SDNN, RMSSD, pNN50 for women p<0.01).

The results (median, 1st/3rd quartile) are also presented in details in table 4-3 for men and in table 4-4 for women.

[page 68↓]

Table 4-3: Time domain parameters of trained and untrained men with its significance level (p-value)

Table 4-4: Time domain parameters of trained and untrained women with its significance level (p-value)

The TP and LF power were significantly higher in trained compared with the sedentary subjects (for men p<0.01-0.001, for women p<0.05).

The VLF was significantly higher in male athletes compared with untrained men (p<0.01) whereas trained and untrained women were similar.

The absolute HF power was similar (n.s.) in the athletic and the sedentary groups.

Table 4-5 presents the detailed results of the male and table 4-6 of the female group.

[page 69↓]

Table 4-5: Frequency domain parameters of trained and untrained men with its significance level (p-value)

Table 4-6: Frequency domain parameters of trained and untrained women with its significance level (p-value)

The LFnu, HFnu as well as the LF/HF ratio were significantly higher in trained compared with untrained men (p<0.01). Still, no difference was noted in LFnu, HFnu as well as LF/HF ratio between trained and untrained females.

Male athletes had significantly higher LFnu and LF/HF ratio than female athletes (p<0.05) whereas trained women had significantly increased HFnu compared to trained men (p<0.05).

Table 4-7 and 4-8 present these results in detail.

[page 70↓]

Table 4-7: LFnu, HFnu and LF/HF ratio in trained and untrained men and its significance level (p-value)

Table 4-8: LFnu, HFnu and LF/HF ratio in trained and untrained women

The results of the heart rate variability (HRV) in relation to the menstrual cycle were compared between trained and untrained women. Data are presented in five different menstruation phases, which were individually specified for each woman. The focus of interest was the HRV in course of one menstrual cycle in two different female groups (athletes and sedentary women) whereas only the course of the HRV in five phases is considered.

No difference was found in the HRV of the time domain between trained and untrained women in course of one menstrual cycle. Both female groups reacted similarly throughout the five menstruation phases and no significance was noted in the time domain parameters as presented in the following figures (figure 4-15 to 4-19).

	Figure 4-15: The course of the meanNN during the menstrual cycle was similar in both groups; no significance was noted

	Figure 4-16: The course of the SDNN was not significantly different between athletic and sedentary females during the five menstruation phases

	Figure 4-17: The course of the RMSSD was similar in both groups; no significance was noted

	Figure 4-18: The course of pNN50 was not significantly different in trained and untrained women during the menstrual cycle

There was no significant difference in the HRV of the frequency domain between trained and untrained women in course of the menstrual cycle. The course of the single parameters was the same in both female groups. Thus, athletic and sedentary women reacted similar throughout the five menstruation phases as presented in the figures 4-19 to 4-25.

	Figure 4-19: The course of the TP was not significantly different in trained and untrained women during the menstrual cycle

	Figure 4-20: The course of the VLF was similar in both groups during the five menstruation phases

	Figure 4-21: The course of the LF power was the same in both groups; no significance was noted during one menstrual cycle

	Figure 4-22: The course of the HF power was not significantly different in athletic and sedentary women throughout the five menstruation phases

	Figure 4-23: The course of the LFnu was similar in both female groups; no significance was noted during one menstrual cycle

	Figure 4-24: The course of the HFnu was not significantly different in trained and untrained women during the menstrual cycle

	Figure 4-25: The course of the LF/HF ratio was similar in athletic and sedentary females; no significance was noted during the five menstruation phases

Differences in the provocation of the orthostatic test induced by an active body position change were investigated in the four subgroups of all study days. Therefore trained and untrained men and women were compared in the HRV parameters of the time and the frequency domain in five orthostatic tests during the study.

The HRV parameters of the time domain reacted similar in all groups. No significance was noted between genders and between trained and untrained subjects. In supine position every single HRV parameter was increased compared with standing whereas while standing the opposite was noted. Still the five orthostatic tests provoked the same reaction in all groups as presented in the following figures (figure 4-26 to 4-30).

	Figure 4-26: MeanNN during the orthostatic tests in trained and untrained men and women; no significance was noted between the groups

	Figure 4-27: SDNN during the orthostatic tests was not significantly different in athletic and sedentary men and women

	Figure 4-28: RMSSD during the orthostatic tests was similar in all groups; no significance was noted

	Figure 4-29: pNN50 during the orthostatic tests in trained and untrained men and women without any significant difference between the groups

The HRV parameter of the frequency domain reacted different between trained and untrained men and women during the orthostatic tests. Thereby male and female athletes were compared as well as sedentary men and women. Different results were noted between the groups which are described in details below the corresponding figures (figures 4-30 to 4-36).

	Figure 4-30: The TP during the orthostatic tests decreased while standing and increased in supine position in all groups whereas the TP of trained and untrained women decreased and increased more significantly compared with men (for athletes p<0.05, for sedentary subjects p<0.01)

	Figure 4-31: The VLF was not significantly different between male and female athletes whereas the VLF reacted opposite in the sedentary group; while standing the VLF increased in untrained men and decreased in womenwhich resulted in a significantly different course of VLF (p<0.05)

	Figure 4-32: The LF power was increased while standing and remained enhanced in supine position in athletic and sedentary men whereas the LF power of women decreased while standing and increased in supine position; thus the LF power was significantly different in athletes (p<0.001) and in sedentary subjects (p<0.001).

	Figure 4-33:The HF power was similar in all groups with a decrease while standing and an increase in supine position; no significant difference was noted

	Figure 4-34: The LFnu was not significantly different in athletes and sedentary men and women; while standing the LFnu increased and decreased in supine position

	Figure 4-35: The HFnu reacted in the opposite way and was increased in supine position and decreased while standing in all groups; no significant difference was noted

	Figure 4-36: The LF/HF ratio decreased while standing and increased in supine position whereas the male athletes had a significantly higher increase and decrease compared with the female athletes (p<0.05); the sedentary groups were not significantly different

The orthostatic test was done five times during the study in women. The main interest was the effect of the menstrual cycle on the orthostatic provocation in women, which was investigated in five different menstrual cycle phases. No difference was noted in the time and frequency domain of the HRV in trained and untrained women. The orthostatic provocation reacted similarly five times during the course of one menstrual cycle. No significance was noted between the orthostatic provocation in athletic and sedentary females. Thereby, the HRV results of the time and frequency domain are presented in one female group.

In the time domain the HRV parameters reacted similarly throughout the menstrual cycle phases which are illustrated by the following figures (figure 4-37 to 4-40). No significance was noted between the five phases in women.

	Figure 4-37: The meanNN during the orthostatic test was similar five times; no significant difference was noted during the menstrual cycle in women

	Figure 4-38: The SDNN during the orthostatic test was not significantly different in course of the menstrual cycle in women

	Figure 4-39: The RMSSD during the orthostatic test reacted similarly during the five menstrual cycle phases in women; no significant difference was noted

	Figure 4-40: The pNN50 was not significantly different throughout the five menstrual cycle phases in women

The HRV parameters of the frequency domain were not significantly different in the five orthostatic tests. Throughout one menstrual cycle, the results of the orthostatic test were similar in women. The orthostatic provocation of each HRV parameters in course of the menstrual cycle is presented in the following figures (figure 4-41 to 4-47).

	Figure 4-41: The TP during the orthostatic test was similar five times in women; no significant difference was noted

	Figure 4-42: The VLF during the orthostatic test was not significantly different throughout one menstrual cycle in women

	Figure 4-43: The LF power during the orthostatic test was similar in all menstrual cycle phases; the results were not significantly different between the five days

	Figure 4-44: The HF power during the orthostatic test was not significantly different in five menstruation phases in women

	Figure 4-45: The LFnu during the orthostatic test was similar in course of the menstrual cycle; no significant difference was noted between the five study days in women

	Figure 4-46: The HFnu during the orthostatic test was not significantly different throughout one menstrual cycle in women

	Figure 4-47: The LF/HF ratio during the orthostatic test was not significantly different in five study days in course of the menstrual cycle in women

No correlation or regression could be found between any parameter and the HRV parameters.

The POMS, the blood borne parameters, the ergometric data and the respiration did not correlate with any of the parameters of the HRV; also no regression was found. No menstrual cycle phase correlated with the parameters of the HRV and no regression was noted. Additionally no difference was found between men and women.

Finally no correlation and regression could be found during the orthostatic provocation in trained and untrained men and women.

The reliability of the repeated measurements has been controlled by the intragroup correlation coefficient (ICC), the coefficient of variation (CV) and additionally by Bland-Altman plots. The ICC is presented by the mean, the minimum and the maximum in the following table (table 4-9) whereas the CV evaluated in percentage can be noted in table 4-10. The more detailed results of the ICC can be found in the appendix. No universal standards exist in classifying the ICC in the HRV measurements. In accordance to Lee et al. the ICC was considered poor if ICC<0.40, acceptable if ICC ranged from 0.41-0.60, good if ICC ranged between 0.61-0.80 and excellent if ICC>0.81.

The Bland-Altman plots illustrated a dispersion of the meanNN, the SDNN, the RMSSD and the pNN50 within ±2 standard deviations. Nevertheless the dispersions of the frequency domain parameters were not acceptable considering the Bland-Altman plots due to the unacceptable limits of agreements (mean ± 2 standard deviations) which are not related to the original scale of measurement. Exemplary two Bland-Altman plots illustrate an acceptable and an unacceptable dispersion of the HRV parameters (figure 4-48 and 4-49).

[page 92↓]

Table 4-9: Intergroup correlations coefficient (ICC) of trained and untrained men as well as of all men

Table 4-10: Coefficient of variation (CV) of trained and untrained as well as of all men

	Figure 4-48: Bland-Altman plot of pNN50 with an acceptable dispersion (96% of cases) within mean ± 2 standard deviations and acceptable limits of agreements (mean ± 2 standard deviations)

	Figure 4-49: Bland-Altman plot of LF power illustrates an unacceptable dispersion (92% of cases) because the limits of agreement (mean ± 2 standard deviations) are not related to the original scale of measurement; the mean is 616 ms² and the range of limits 7571 ms².