Mood States (POMS) has shown that overtraining commonly results in significant mood disturbances in swimmers and other endurance athletes (15).
84
Development of a Scale for Use in Monitoring Training-Induced Distress in Athletes J. S. Raglin I, W. P. lv/organ:1 I 2
Indiana University. Bloomington University oCWisconsin-Madison
Introduction Abstract
J. S. Raglin and W. P. Morgan, Development of a Scale for Use in Monitoring Training-Induced Distress in Athletes. Int J Sports Med, Vol 15, No 2. pp 84-88, 1994. Accepted after revision: September 30, 1993 In an attempt to identify athletes at risk of training-induced distress, a scale was developed using items from the Profile of Mood States (POMS). POMS data were collected monthly from 170 varsity college swimmers (70 female, 100 male) during four competitive training seasons. The team coaches identified swimmers who showed signs of compromised performance in response to intensive training (i.e., distress), and several series of discriminant function analyses were conducted in attempts to identify these in dividuals using POMS scale items. Discriminant function equations derived from subsets of items from the entire POMS items pool resulted in a mean prediction rate for identifying distressed swimmers of 93.9% for men and 100% for women. Seven POMS items which entered the equations most frequently served as predictor variables for analyses of a sample of 33 men swimmers. Predictions of distress were conducted at each monthly assessment. The prediction rate of distress achieved with the 7-item set of POMS items averaged 69.1 %, and this level ofaccuracy was 37.0 % above the chance rate of prediction (p < 0.05). In a cross-validation on 29 collegiate track and field athletes, the prediction rate of the 7-item scale was 66.7% (p
Depression Depression
PO,\.IS items subset It was found that seven POMS items were most effective in identifying distressed swimmers, and these items are presented in Table 2. Subsequently, these items were used as predictor variables to determine their accuracy in correctly classifying distressed athletes from a sample of 33 male swim mers. Nine swimmers in this sample were rated by the coach as distressed. Table 3 presents the trial by trial prediction rates achieved by discriminant function analyses based on the seven item POMS subset. The mean prediction rate for distress yielded by the 7-item scale ....-as 69.1 %. Monthly predictions ranged from 55.6 to 77.8 %. and each was significantly (p < 0.05) better than the chance level of prediction. In some cases not all of the items entered the discriminant function equation. but the pre diction rates were not improved when these items were forced into the equation.
Cross validation: track and field athletes The ability of the 7 -item distress scale in iden tifying distressed athletes from a different sport was then assessed. Twenty-nine members of a men's collegiate track and field team served as subjects (mean age = 19.4), and these in dividuals completed the POMS once during the heavy training season. Using the same criteria as the swimming coaches, the track coach was asked to identify those athletes who ex.hibited performance problems specifically related to training which persisted for more than several days. Sick or injured athletes
Training Distress
Int. J. Sports Med. IS (1994)
Table 3 Accuracy, Gain over Chance. and A-Values for the Prediction of Distress Based on 7-Jtern POMS Scale Scores.
Men (N
Accuracy (%> % Gain MultipJe-R N
=33)
1
2
3
77.S" (49.7) .33 32
77.S" (50.5) .56'" 33
(4O.6) .59 23
66.r
Trial 4 70.0° (32.5) .71 " 24
Overall 5
6
7
Mean
55.6" (19.6) .25
66.7° (29.2) .5So 24
69.1 (37.0)
25
.SO
"p < 0.05. -p < 0.01
Notes: Due to logistical constraints no mood data was collected at trial 5. Because of miSsing data. there were only 6 distressed athletes at trial
were excluded from consideration. Six athletes (20.7%) out of the team were identified as distressed Scores from the 7-item scale were used in an attempt to gauge its effectiveness in identifying distressed track athletes. To determine if a simpler method for predicting dis tressed athletes would be effective, classifications were based on mean distress scores rather than by discriminant analyses. Athletes who possessed scores on the 7-item scale of one stan dard deviation or more above the team mean were classified as distressed. Previous POMS research (10,16) has found that this criterion (Le., mean + one standard deviation) has been success ful in identifying athletes who show signs of distress during intensive training. In an attempt to determine if the 7-item scale was more effective than POMS scales, this procedure was re peated using POMS total scores and POMS depression scores. All predictions were made without knowledge of which athletes were rated as distressed by the coach. Eight athletes possessed scores one standard deviation or greater than the mean for POMS total scores. Four of these athletes were classified as distressed using the POMS total scale score, resulting in an accuracy of 50 % (4 of 8 predictions). For POMS depression, four of the seven athletes (57.1 %) with elevated scores were distressed. and the level of accuracy was 66.7% (4/6) for the 7-item scale. Each of these rates was significantly (p
