found Prmted
of Psychosomaric Research.
Vol. 36. No. 7, pp. 657-665.
1992. b
an Great Brimm
PARTICIPATION CARDIOVASCULAR
IN PHYSICAL FITNESS HAVE
ON MENTAL KATHRYN (Received
26 September
HEALTH
THIRLAWAY 1991; accepted
ACTIVITY DIFFERENT
0022-3999192 $5.00+.00 1992 Pergamon Press Ltd
AND EFFECTS
AND MOOD
and DAVID BENTON in revised form
12 February
1992)
Abstract-The cardiovascular fitness, blood pressure, resting pulse, lung capacity and body fat of 246 subjects were measured and each subject completed activity, mental health and mood questionnaires. The results of this study support the hypothesis that participation in physical activity, rather than cardiovascular fitness, is the factor associated with better mental health and mood. Higher levels of physical activity were associated with better mood scores unless the individuals were unfit. This research has implications for how physical activity is prescribed; the emphasis needs to be on performing physical activity rather than improving fitness and should not be beyond the physical capabilities of the individual.
INTRODUCTION THE
[ l-71
ASSOCIATION
. Participation
between physical activity and mental health is well established in exercise is associated with lower anxiety [
al. [ 141 suggested that physical activity rather than increased cardiovascular fitness was important, as moderate activity, but not high activity, was associated with better mood. Steptoe er al. [ 51 also carried out an intervention study which found moderate, but not high intensity activity, was associated with better mood. However, Goldwater and Collis [ 1 ] found that vigorous rather than less demanding cardiovascular exercise was associated with greater decreases in anxiety, and increases in well-being. The Steptoe et al. [ 51 study is one of the few intervention studies that distinguished participation and fitness changes and, despite small sample sizes, found that participation in exercise rather than increased fitness was associated with improvements in mood. The present study investigated the hypothesis that participation in physical activity, rather than cardiovascular fitness, is the factor associated with better mental health and mood. Cardiovascular fitness and activity levels were measured in a nonexperimental population. It was hypothesized that although cardiovascular fitness and physical activity would be positively associated, it would be physical activity and
Psychology
Department,
Swansea
University,
Singleton
Park,
West Glamorgan,
U.K.
K. THIRLAWAY and D. BENTON
658
not fitness psychiatric
that was associated symptoms.
with a tendency
to self report
a lower
incidence
of
METHOD Subjects
Two-hundred and eight of the 246 subjects were employees of West Glamorgan County Council. A staff bulletin was sent out by the personnel department offering a free fitness test, in return for completing three questionnaires. Two-hundred and eight out of approximately 600 white collar workers volunteered to take part. Although the subjects were volunteers they were offered the chance of a free fitness test, in the building they worked in, during office hours, so the bias towards a self selected fit sample was limited. In fact 84 subjects were classified as sedentary. The remaining 38 subjects had booked a fitness test in Swansea Leisure Centre and were requested to fill in the questionnaires after the fitness test. Every subject completed all three questionnaires. There were 155 males and 91 females aged between 18 and 63 yr. The majority of subjects were between 30 and 40 yr. PROCEDURE The subjects first completed a physical activity habits questionnaire and a Profile of Mood States (POMS) [ 191. They were weighed and their resting pulse, blood pressure, body fat and lung capacity measured. Anybody with diastolic blood pressure over 100, or systolic blood pressure over 140, was not permitted to continue with the sub-maxima1 test, one subject was excluded. As long as blood pressure was normal the subjects proceeded with the bicycle assessment. While the results of the test were being calculated the subjects completed a General Health Questionnaire (GHQ) [20]. Activity
levels
Subjects completed a questionnaire concerning their physical activity over the last fortnight. A list of activities was presented and they noted those in which they had engaged, how many times and the duration. If they mentioned running, cycling or walking they were also asked how far they ran, cycled or walked each time. Each type of exercise was assigned a value that reflected the amount of oxygen consumed by the body per min [21]. The total time spent in a fortnight on each activity was multiplied by its value per min and the scores for each activity performed were added to give a total. Having calculated an activity score for each individual they were arbitrarily split into three groups according to activity score: Group 1 None to virtually no activity - activity score O-20. - activity score 20-60. Group 2 Some activity Group 3 Active - activity score 60-high. To score 0 you must report no physical activity at all. To score 20 points in a fortnight you would have, for example, walked for 20 min one day or swum for 40 min once in the fortnight. To score 60 points you may have walked for 70 min daily or swum for 30 min eight times in a fortnight. Naturally, points are earned typically from a range of activities. These activity ratings have been previously found to relate to an estimate of VO, max [6, 221. Sub-maximal
test of VO, max
A sub-maximal estimate of VO, max is based on the high correlation between cardiac output during exercise and oxygen uptake. When exercise is performed on a cycle ergometer there are relatively small individual variations in mechanical efficiency, so oxygen uptake can be predicted from the external power to which the subject is subjected [23]. A Monark ergometric bicycle (Monark-Crescent AB Varberg, Sweden) was used for the sub-maximal VO, max test. A Sports Lab Computer and Heart Rate Monitor (Triadcolour Electronics Ltd) was used to measure heart rate. The sub-maximal test used was the one described and recommended by Astrand and Rodahl [23] Work time was measured using a stop clock and was 6 min. Subjects pedalled at 50 pedal turns per min. A heart rate of at least 130 beats per min was achieved by each subject by the end of their 6 min on the ergometric bicycle. The average heart rate over the last 2 min of the test was recorded. Maximal Oxygen Uptake per litre was predicted from the heart rate and work load on the bicycle ergometer from Astrand’s nomogram [23] The value was then corrected for age [ 231. The maximal oxygen uptake per 1 was then converted to oxygen uptake per kg of body weight per min. Finally subjects were classified into 4 fitness groups depending on their sex and age:
Participation
in exercise, Oxygen
Female
Male
mental
consumption
Age 20-29 30-39 40-49 50-65
Low < 28 < 27 < 25 < 21 N= 10
Somewhat low 29-34 28-33 26-3 1 22-28 N = 22
20-29 30-39 40-49 50-59 60-69
< 38 < 34 < 30 < 25 < 21 N=41
39-43 35-39 3 l-35 25-31 22-26 N=49
health
and mood
ml/kg
per min
659
Average 35-43 34-41 32-40 29-36 N= 39
High > 44 > 42 > 41 > 37 N=20
44-5 1 40-47 36-43 32-39 27-35 N = 40
> 52 > 48 > 44 > 40 > 36 N = 20
Blood pressure and pulse
Systolic and diastolic blood pressure and pulse were measured using a Philips electronic blood pressure meter. The cuff was inflated manually with a pump and the results displayed digitally. Four subjects did not have their blood pressure recorded, three subjects did not have their resting pulse recorded. Blood pressure was measured after the subjects had sat quietly for 5 min filling questionnaires. Lung capacity Forced Expiratory Volume in 1 set (FEVI) was measured using a Pocket Spirometer (Micro Medical Instruments Ltd. Rochester, U.K.). It was measured three. times and the maximum value selected. None of the leisure centre subjects had their lung capacity recorded. Body
fat
Skinfold measurements were taken: biceps, triceps, subscapular measurements result in an estimated fat content as a percentage age 1241.
and supra-iliac. of body weight,
The sum of these four dependent on sex and
Mental health
Every subject completed the 30 question version questionnaire was scored using the Likert method.
of the General
Health
Questionnaire
[20].
The
Mood
The Profile
of Mood
States
[ 191 was completed. STATISTICS
The data were analysed using the SPSWPC statistical package. Small cell sizes prevented the reporting of a four-way analysis of variance of GHQ and Mood scores: Activity (No activity/Some activity/ Active) x Sex (Male/Female) x Age (Under 35/0ver 35) x Fitness (Low/Somewhat low/Average/ High). As preliminary analysis showed that no interaction involved age a three way analysis of variance: Activity (No activity/Some activity/Active) x Sex (Male/Female) x Fitness (Low/Somewhat low/ Average/High) is reported. Physiological
variables
Pearson’s correlations systolic blood pressure,
were calculated between: pulse, weight and forced
Activity scores, VO, max. diastolic expiratory volume per set (FEVI).
blood pressure,
Mental health and mood
A three-way analysis of variance: Activity (No activity/Some activity/Active) X Fitness (Low/ Somewhat low/Average/High) x Sex (Male/Female) was used to analyse the GHQ scores and POMS scores. As there is no way of examining simple main effects due to the unequal cell sizes in the study, any significant interaction that resulted was further examined using one-way analysis of variance. Although it is not strictly appropriate, it is a conservative process that biases against significant findings.
K.
660
THIRLAWAY
and
D.
BENTON
RESULTS
Physiological
variables
Activity scores correlated positively with estimated VOZ max @ < 0.0001) and pulse @ < 0.042) although not with diastolic or systolic blood pressure, weight or FEVI (Table I). Estimated VO;, max correlated positively with diastolic blood pressure (p < O.OOOl), systolic blood pressure @ < 0.018), pulse (p < O.OOOl), weight @ < 0.033) and FEVI @ < 0.031) (Table I). The sample size is slightly reduced due to missing physiological data for two or three subjects, depending on the physiological parameter in question. TABLE
I.-CORRELATIONS
VO,,
DIASTOLIC
PRESSURE,
BETWEEN
BLOOD
RESTING EXPIRATORY
PULSE,
VOLUME
SCORES AND
SYSTOLIC
WEIGHT
Activity Activity
ACTIVITY
PRESSURE,
AND
BLOOD
FORCED
(FEVI) scores
vo,
max
0.2850
scores
p
of Psychosomaric Research.
Vol. 36. No. 7, pp. 657-665.
1992. b
an Great Brimm
PARTICIPATION CARDIOVASCULAR
IN PHYSICAL FITNESS HAVE
ON MENTAL KATHRYN (Received
26 September
HEALTH
THIRLAWAY 1991; accepted
ACTIVITY DIFFERENT
0022-3999192 $5.00+.00 1992 Pergamon Press Ltd
AND EFFECTS
AND MOOD
and DAVID BENTON in revised form
12 February
1992)
Abstract-The cardiovascular fitness, blood pressure, resting pulse, lung capacity and body fat of 246 subjects were measured and each subject completed activity, mental health and mood questionnaires. The results of this study support the hypothesis that participation in physical activity, rather than cardiovascular fitness, is the factor associated with better mental health and mood. Higher levels of physical activity were associated with better mood scores unless the individuals were unfit. This research has implications for how physical activity is prescribed; the emphasis needs to be on performing physical activity rather than improving fitness and should not be beyond the physical capabilities of the individual.
INTRODUCTION THE
[ l-71
ASSOCIATION
. Participation
between physical activity and mental health is well established in exercise is associated with lower anxiety [
al. [ 141 suggested that physical activity rather than increased cardiovascular fitness was important, as moderate activity, but not high activity, was associated with better mood. Steptoe er al. [ 51 also carried out an intervention study which found moderate, but not high intensity activity, was associated with better mood. However, Goldwater and Collis [ 1 ] found that vigorous rather than less demanding cardiovascular exercise was associated with greater decreases in anxiety, and increases in well-being. The Steptoe et al. [ 51 study is one of the few intervention studies that distinguished participation and fitness changes and, despite small sample sizes, found that participation in exercise rather than increased fitness was associated with improvements in mood. The present study investigated the hypothesis that participation in physical activity, rather than cardiovascular fitness, is the factor associated with better mental health and mood. Cardiovascular fitness and activity levels were measured in a nonexperimental population. It was hypothesized that although cardiovascular fitness and physical activity would be positively associated, it would be physical activity and
Psychology
Department,
Swansea
University,
Singleton
Park,
West Glamorgan,
U.K.
K. THIRLAWAY and D. BENTON
658
not fitness psychiatric
that was associated symptoms.
with a tendency
to self report
a lower
incidence
of
METHOD Subjects
Two-hundred and eight of the 246 subjects were employees of West Glamorgan County Council. A staff bulletin was sent out by the personnel department offering a free fitness test, in return for completing three questionnaires. Two-hundred and eight out of approximately 600 white collar workers volunteered to take part. Although the subjects were volunteers they were offered the chance of a free fitness test, in the building they worked in, during office hours, so the bias towards a self selected fit sample was limited. In fact 84 subjects were classified as sedentary. The remaining 38 subjects had booked a fitness test in Swansea Leisure Centre and were requested to fill in the questionnaires after the fitness test. Every subject completed all three questionnaires. There were 155 males and 91 females aged between 18 and 63 yr. The majority of subjects were between 30 and 40 yr. PROCEDURE The subjects first completed a physical activity habits questionnaire and a Profile of Mood States (POMS) [ 191. They were weighed and their resting pulse, blood pressure, body fat and lung capacity measured. Anybody with diastolic blood pressure over 100, or systolic blood pressure over 140, was not permitted to continue with the sub-maxima1 test, one subject was excluded. As long as blood pressure was normal the subjects proceeded with the bicycle assessment. While the results of the test were being calculated the subjects completed a General Health Questionnaire (GHQ) [20]. Activity
levels
Subjects completed a questionnaire concerning their physical activity over the last fortnight. A list of activities was presented and they noted those in which they had engaged, how many times and the duration. If they mentioned running, cycling or walking they were also asked how far they ran, cycled or walked each time. Each type of exercise was assigned a value that reflected the amount of oxygen consumed by the body per min [21]. The total time spent in a fortnight on each activity was multiplied by its value per min and the scores for each activity performed were added to give a total. Having calculated an activity score for each individual they were arbitrarily split into three groups according to activity score: Group 1 None to virtually no activity - activity score O-20. - activity score 20-60. Group 2 Some activity Group 3 Active - activity score 60-high. To score 0 you must report no physical activity at all. To score 20 points in a fortnight you would have, for example, walked for 20 min one day or swum for 40 min once in the fortnight. To score 60 points you may have walked for 70 min daily or swum for 30 min eight times in a fortnight. Naturally, points are earned typically from a range of activities. These activity ratings have been previously found to relate to an estimate of VO, max [6, 221. Sub-maximal
test of VO, max
A sub-maximal estimate of VO, max is based on the high correlation between cardiac output during exercise and oxygen uptake. When exercise is performed on a cycle ergometer there are relatively small individual variations in mechanical efficiency, so oxygen uptake can be predicted from the external power to which the subject is subjected [23]. A Monark ergometric bicycle (Monark-Crescent AB Varberg, Sweden) was used for the sub-maximal VO, max test. A Sports Lab Computer and Heart Rate Monitor (Triadcolour Electronics Ltd) was used to measure heart rate. The sub-maximal test used was the one described and recommended by Astrand and Rodahl [23] Work time was measured using a stop clock and was 6 min. Subjects pedalled at 50 pedal turns per min. A heart rate of at least 130 beats per min was achieved by each subject by the end of their 6 min on the ergometric bicycle. The average heart rate over the last 2 min of the test was recorded. Maximal Oxygen Uptake per litre was predicted from the heart rate and work load on the bicycle ergometer from Astrand’s nomogram [23] The value was then corrected for age [ 231. The maximal oxygen uptake per 1 was then converted to oxygen uptake per kg of body weight per min. Finally subjects were classified into 4 fitness groups depending on their sex and age:
Participation
in exercise, Oxygen
Female
Male
mental
consumption
Age 20-29 30-39 40-49 50-65
Low < 28 < 27 < 25 < 21 N= 10
Somewhat low 29-34 28-33 26-3 1 22-28 N = 22
20-29 30-39 40-49 50-59 60-69
< 38 < 34 < 30 < 25 < 21 N=41
39-43 35-39 3 l-35 25-31 22-26 N=49
health
and mood
ml/kg
per min
659
Average 35-43 34-41 32-40 29-36 N= 39
High > 44 > 42 > 41 > 37 N=20
44-5 1 40-47 36-43 32-39 27-35 N = 40
> 52 > 48 > 44 > 40 > 36 N = 20
Blood pressure and pulse
Systolic and diastolic blood pressure and pulse were measured using a Philips electronic blood pressure meter. The cuff was inflated manually with a pump and the results displayed digitally. Four subjects did not have their blood pressure recorded, three subjects did not have their resting pulse recorded. Blood pressure was measured after the subjects had sat quietly for 5 min filling questionnaires. Lung capacity Forced Expiratory Volume in 1 set (FEVI) was measured using a Pocket Spirometer (Micro Medical Instruments Ltd. Rochester, U.K.). It was measured three. times and the maximum value selected. None of the leisure centre subjects had their lung capacity recorded. Body
fat
Skinfold measurements were taken: biceps, triceps, subscapular measurements result in an estimated fat content as a percentage age 1241.
and supra-iliac. of body weight,
The sum of these four dependent on sex and
Mental health
Every subject completed the 30 question version questionnaire was scored using the Likert method.
of the General
Health
Questionnaire
[20].
The
Mood
The Profile
of Mood
States
[ 191 was completed. STATISTICS
The data were analysed using the SPSWPC statistical package. Small cell sizes prevented the reporting of a four-way analysis of variance of GHQ and Mood scores: Activity (No activity/Some activity/ Active) x Sex (Male/Female) x Age (Under 35/0ver 35) x Fitness (Low/Somewhat low/Average/ High). As preliminary analysis showed that no interaction involved age a three way analysis of variance: Activity (No activity/Some activity/Active) x Sex (Male/Female) x Fitness (Low/Somewhat low/ Average/High) is reported. Physiological
variables
Pearson’s correlations systolic blood pressure,
were calculated between: pulse, weight and forced
Activity scores, VO, max. diastolic expiratory volume per set (FEVI).
blood pressure,
Mental health and mood
A three-way analysis of variance: Activity (No activity/Some activity/Active) X Fitness (Low/ Somewhat low/Average/High) x Sex (Male/Female) was used to analyse the GHQ scores and POMS scores. As there is no way of examining simple main effects due to the unequal cell sizes in the study, any significant interaction that resulted was further examined using one-way analysis of variance. Although it is not strictly appropriate, it is a conservative process that biases against significant findings.
K.
660
THIRLAWAY
and
D.
BENTON
RESULTS
Physiological
variables
Activity scores correlated positively with estimated VOZ max @ < 0.0001) and pulse @ < 0.042) although not with diastolic or systolic blood pressure, weight or FEVI (Table I). Estimated VO;, max correlated positively with diastolic blood pressure (p < O.OOOl), systolic blood pressure @ < 0.018), pulse (p < O.OOOl), weight @ < 0.033) and FEVI @ < 0.031) (Table I). The sample size is slightly reduced due to missing physiological data for two or three subjects, depending on the physiological parameter in question. TABLE
I.-CORRELATIONS
VO,,
DIASTOLIC
PRESSURE,
BETWEEN
BLOOD
RESTING EXPIRATORY
PULSE,
VOLUME
SCORES AND
SYSTOLIC
WEIGHT
Activity Activity
ACTIVITY
PRESSURE,
AND
BLOOD
FORCED
(FEVI) scores
vo,
max
0.2850
scores
p
