HORMONAL CHANGES AT REST IN OVERTRAINED ENDURANCE ATHLETES

Abstract

Resting levels of testosterone, luteinising hormone, cortisol, prolactin, and the testosterone/cortisol ratio were studied in a group of endurance athletes before and throughout a high intensity–volume phase of their yearly training cycles. During this period four athletes developed overtraining characteristics. These subjects were physically matched with control subjects conducting comparable training but exhibiting no overtraining symptoms. Hormonal comparisons between the groups revealed no significant differences existed prior to beginning the intensive training; nor did the hormonal levels of the control subjects alter significantly (P>0.05) due to the training. The overtrained subjects, however, did have a significant (P<0.01) decline in testosterone (6.8±1.0 to 4.4±1.0 ng/ml; MEAN ± SE). Furthermore, prolactin was found to significantly (P<0.05) elevate in the overtrained subjects (8.1±2.0 to 13.2±1.1. ng/ml), while the testosterone/cortisol ratio showed a significant (P<0.005) decline (0.83±0.26 to 0.36±0.08). No other significant changes were noted. These findings support the theory that the overtraining syndrome in athletes may represent a neuroendocrine dysfunction.

Introduction

The coach and the athlete must balance the physical and emotional aspects of training carefully to insure optimal performance. Thus, the act of training an athlete is many times viewed as being more an art than a science. For this reason athletes in many instances are undertrained or overtrained. In recent years the latter condition, overtraining, seems to be a more common occurrence within the athletic community [3,7].

Little is known about the overtrained state. In fact, the research literature is quite unresolved even as to how to define the term overtraining. Our laboratory uses the term “overtraining” to indicate an excessive extension of the training process that results in a decline of physiological and psychological performance. Other investigators have used the terms or phrases “burnout”, “withdrawal”, “overtraining syndrome”, and “staleness” [7]. Some of the common signs and symptoms of the overtrained athletes which have been reported are listed in Table 1 [1,3].

Table 1.

Signs and symptoms of overtrained athletes

a. Apathy

b. Lethargy

c. Appetite loss

d. Sleep loss

e. Mood changes

f. Weight loss

g. Heavy feeling

h. Gastrointestinal disturbances

i. Lymphadenopathy

j. Drawn appearance

k. Performance decline

l. Retarded recovery after exertion

m. Muscle pain or soreness

n. Elevated resting heart rate

o. Elevated resting blood pressure

p. Abnormal resting hormonal profiles

Adapted from Barron et al. [1] and Hackney et al. [3]

To date very few studies have objectively examined the question of what are the physiological aberrations associatedwith the overtrained athlete. In recent review articles Kuipers and Kaizer [7] and Hackney et al. [3] concluded from the antidotal and limited objective data available that. neuroendocrine dyfsunction seems to accompany overtraining, and may be the causative factor for the psychophysiological changes noted. The intent of this study was to add to the sparse data available and to examine resting neuroendocrine status in overtrained athletes.

In the present study select resting hormonal levels in a group of overtrained athletes were measured and compared to those from a group of similarly training athletes who were not exhibiting overtraining symptoms.

Materials and Methods

A group of endurance athletes (cyclists and runners) were monitored before and during a high intensity-volume phase (8–10 week duration, training several hours a day) of their training cycles. Blood samples were taken before and during this training phase. Samples were collected via venipuncture from a vein in the antecubal area of either arm. Samples were obtained in the morning in a 12 h-postprandial state and after 24 h of physical inactivity. Collected samples were placed on ice, allowed to clot, centrifuged at 3000·g (4°C), serum aliquoted and stored at −80°C. The serum samples were analyzed for testosterone (T), cortisol (C), luteinising hormone (LH), and prolactin (PRL) by standard radioimmunoassays as previously described [5]. Additionally, the T/C ratio was calculated as an index of the anabolic-catabolic status of the subjects [8].

During the course of the study four subjects exhibited signs and symptoms of overtraining. The displayed symptoms corresponded with items a - m listed in Table 1. The symptoms were determined qualitatively from a question/answer “check-list” administered by the investigative team. Unfortunately, no quantitative assessment of the degree, to which the symptoms occured. was made. The overtrained subjects (denoted as the OVT group) were physically matched to four control subjects, denoted as the CON group), undergoing nearly identical training, who developed no symptoms of overtraining. When the OVT group developed their overtraining characterictics, both groups of athletes had blood samples withdrawn. Two sets of samples were collected to reduce the influence of sampling error (i.e. of an isolated sample [3]. The first blood sample was obtained at approximately the 8th week of training and the second one 10 – 12 days later. The types of overtraining symtoms identified did not change in this period, and all seemed to intensify. The hormone values from these sets of blood samples were pooled and statistically analyzed. A non-parametric statistical analysis was conducted to examine the between- (Kruskal-Wallis) and within-group (Wilcoxon) differences. Results were considered significant only at P<0.05.

Results

Figure 1 shows the T changes of the CON and OVT groups. Prior to the display of overtraining symptoms no between groups differences existed. Afterwards, however, T was significantly decreased only in the OTV group. Also, the T levels of the OVT group were significantly different from the CON group. With the reduction in T in the OVT group there was a concomitant increase in LH (see Fig. 2). This change, however, did not reach statistical significance. A significant reduction was noted in the LH of the CON group.

Fig. 1.

Mean (±SE) testosterone levels in overtrained (OVT) and control (CON) subjects in the Pretraining period and during the Training (8 – 10 weeks).

* Significant difference (P<0.05) between the two periods (Pre-training and Training)

Fig. 2.

Mean (±SE) LH levels in overtrained (OVT) and control (CON) subjects in the Pretraining and Training (8 – 10 weeks) periods

* Significant difference (P<0.05) between the two periods (Pre-training and Training)

The C changes are depicted in Fig. 3. No significant changes were noted for either group. Although, both groups showed a tendency towards C elevations. In both groups the T/C ratio decreased, but this decline from baseline was statistically significant in the overtrained group only. No significant between group differences, however, existed in the T/C changes (see Fig. 5). PRL at baseline did not differ between the groups. The levels increased significantly in the OVT group during the training cycle (see Fig. 4), but not so in the CON group.

Fig. 3.

Mean (±SE) cortisol levels in over trained (OVT) and control (CON) subjects in the Pre-trai-nmg and Training (8 – 10 weeks) periods.

* Significant difference (P<0.05) between the two periods

Fig 5.

Mean (±SE) values of testosterone/cortisol ratio in overtrained (OVT) and control (CON) subjects in the Pre-training and Training periods

* Significant difference between the two periods

Fig. 4.

Mean (±SE) prolactin levels in overtrained (OVT) and control (CON) subjects in the Pre-training and Training (8 – 10) weeks periods.

* Significant difference (P<0.05) between the two periods

Discussion

Our results demonstrate that endurance athletes who exhibit overtraining characteristics have lowered T and elevated PRL levels. These findings are somewhat supportative of the hypothesis that the overtraining syndrome is related to a neuroendocrine dysfunction status. However, our data are not in total agreement with previously published research.

The finding of reduced T has been reported by Stray-Gundersen et al. [9]. and Urhausen et al. [10]. Neither of these investigators reported LH changes to occur in their subjects. Our data on LH increases in the OVT group are only tenable (P>0.12). Paradoxically, the CON group showed significant reductions at the same time. The changes observed would suggest that the responsiveness of the hypothalamic-pituitary-gonodal axis was different in the two groups. However, these findings are extremely difficult to interpret due to the highly pulsatile nature of LH release [2]. In fact, the work from our laboratory suggests that LH assessment from isolated blood samples is a poor determinant of hypothalamic-pituitary-gonadal status [3].

Urhausen et al. [10] and Hakkinen et al. [6] have shown significant elevations in C and concomitant declines in the T/C ratio in overtrained athletes. Therefore our data are slightly at odds with these findings. Why the findings differ between the studies is uncertain. Possibly, the fact that the mode of training was quite different between the studies, influenced the outcome (rowers and weight trainers for Urhausen and Hakkinen, respectively). Although, our T/C-ratio data did show changes in the same direction (i.e. reductions) as in the abovementioned investigations, as did our C data (i.e. elevations). These findings would infer an enhanced catabolic vs. anabolic status in our subjects.

The increase in PRL observed has been reported for intensively training athletes [4]; however, previous studies on overtraining have apparently not examined this hormone. Thus our findings of increased PRL with overtraining are new. The rise in this hormone is reflective of an endocrine stress reaction [2] and has been correlated with declines in T [4]. However, with the small number of subjects in the present study, an attempt to examine a similar relationship seemed inappropriate.

The restricted sample size within this study limits the degree of interpretation of the findings. Yet, the data would suggest that an increased stress reactivity as well as a shift towards a more catabolic status may be occurring in the overtrained athlete. These conclusions are in accord with previous studies examining overtraining, which have also had comparably small sample sizes [1,9].