Behaviour Research and Therapy 40 (2002) 401414 www.elsevier.

com/locate/brat

Physical health correlates of overprediction of physical discomfort during exercise

Richie Poulton
a

a,*

, Judy Trevena a, Anthony I. Reeder b, Rose Richards

Dunedin Multidisciplinary Health and Development Research Unit, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, P.O. Box 913, Dunedin, New Zealand b Social and Behavioural Research in Cancer Group, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago, P.O. Box 913, Dunedin, New Zealand Accepted 28 February 2001

Abstract This study sought to determine if overprediction of physical discomfort prior to and following exercise was related to a number of self-report and objective measures of physical health status in a general population sample. Cross-sectional ndings indicated a signicant and specic relation between patterns of discomfort overprediction (vs under- or accurate prediction) and negative self-reported health status and attitudes to exercise, lower levels of physical activity, as well as poorer scores for resting heart rate, cardiorespiratory tness and body mass index (but not waist:hip ratio). Females were more likely to overpredict discomfort, were less active and had poorer physical health than males. The potential usefulness of the prediction matchmismatch paradigm for studying exercise-related appraisal processes as proximal determinants of physical activity are discussed. Information obtained from such studies could prove useful in public health education campaigns aimed at increasing levels of physical activity in the general population, particularly among the less active. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Physical activity; Cognition; Overprediction; Health; Longitudinal; Exercise

1. Introduction The benets of physical activity include better physical and mental health, a reduced risk of premature death and enhanced quality of life (e.g., Bouchard, 2000; Fentem, 1996; Blair & Brodney, 1999). As noted by the US Surgeon General (US Department of Health and Human Services,

* Corresponding author. Fax: +64-3-479-5487. E-mail address: richiep@gandalf.otago.ac.nz (R. Poulton).

0005-7967/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 5 - 7 9 6 7 ( 0 1 ) 0 0 0 1 9 - 5

402

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

1996), however, a signicant proportion of the general population do not meet recommended levels of physical activity. Low levels of activity and increasing levels of obesity in both developed and developing countries are now major public health concerns (e.g., World Health Organisation, 1998; Flegal, Carroll, Kuczmarski, & Johnson, 1998; NIH, 1996) and the promotion of physical activity has become an important goal for both health professionals and policy makers (e.g., Grundy et al., 1999). In this context, understanding the factors that inuence initiation and maintenance of participation in physical activity is particularly relevant (Snell & Mitchell, 1999). The dynamic nature of physical activity acquisition has led to the examination of stages through which individuals progress when moving from inactivity to regular participation (Hausenblas, Carron, & Mack, 1997; Godin, 1994; Prochaska & Marcus, 1994; Marcus & Simkin, 1993). Assessment of cognitive structures within the framework of stages has identied important differences between those in each stage, for example, lower self-efcacy and a greater emphasis placed on the negative aspects of activity among the inactive (Gorely & Gordon, 1995). However, comparatively less attention has been paid to cognitive appraisal processes operating immediately prior to, during and following exercise (e.g., Wing & Jakicic, 2000). The role of expectation or predictions of physical discomfort in determining levels of physical activity (and its correlates e.g., tness) are likely to be important, especially as a relationship between pain and avoidance behaviour has been reported (Philips, 1987) and overpredictions of pain are associated with avoidance (e.g., Rachman & Lopatka, 1988). Further, it appears that inaccurate predictions result in specic readjustments, which may also have long term behavioural consequences (Rachman & Arntz, 1991). These ndings highlight the potential value of studying individual differences in predictions of activity-related discomfort. Accordingly, we examined the relation between prediction of physical discomfort (prior to and following exercise conducted in a controlled setting) and a number of physical (objective) and self-reported (subjective) health parameters. Health measures included cardiorespiratory tness, body mass index (BMI), waist:hip ratio as well as self-reported physical activity levels, general health status, and attitudes to physical activity. Since sex differences are commonly found for many of these physical health measures, sex effects were also examined. 2. Method 2.1. Participants and general procedure Participants were members of the Dunedin Multidisciplinary Health and Development Study, a longitudinal investigation of the health, attitudes and behaviour of 1037 children (52% male, 48% female) born in Dunedin New Zealand, between April 1972 and March 1973 (Silva & Stanton, 1996). Following collection of perinatal data, regular assessments have been conducted every 2 years between ages 3 and 15 years, again at ages 18 and 21 and most recently at age 26 years (n=980; 96% of living cohort members, 499 males). During that assessment phase, conducted in 199899, study members spent a full day at the Research Unit for condential interviews and examinations. Typically, this occurred within 3 months of their 26th birthdays. Cohort families represent the full range of socio-economic status in the general population of New Zealands

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

403

South Island, and they are primarily New Zealand European. At age 26 fewer than 8% selfidentied as New Zealand Maori. 2.2. Discomfort predictions During quality control reviews conducted as part of the age 26 assessment, it became clear that some study members were declining the submaximal exercise bike test (see below) because they anticipated discomfort. This appeared to be a non-random phenomenon, especially as a detailed explanation about the actual (relatively low) level of discomfort was provided at the beginning of the assessment day, and again at the commencement of the physical health assessment session. Based on these observations, two questions were added part-way through the assessment phase, one immediately preceding and the other immediately following the exercise bike test. The rst question asked How much discomfort do you anticipate experiencing during the bike test? (rated on an 11-point scale, with 0=no discomfort and 10=extreme levels of discomfort). The second question asked How much discomfort did you actually experience during the bike test? (rated on the same 010 scale). 2.3. Age-26 physical health correlates 2.3.1. Cardiorespiratory tness Study members underwent a 6-minute constant power submaximal exercise test on a friction braked cycle ergometer (Monark, Sweden). They rst completed a 2-minute warm-up at 50 W to gauge their heart rate (HR) response. Depending on the extent to which their HR increased in these rst 2 minutes, the workload was then adjusted to elicit a steady-state heart rate in the range of 130170 bpm for a further 56 minute constant power output stage. Maximal aerobic power was predicted from the nal submaximal heart rate of the constant power period, using a modication of the methods originally published by Astrand (1960); (for more detail see Sleivert, Hopkins, Reeder & Poulton, 2001). 2.3.2. Resting heart rate Five measures of heart rate were obtained at xed times during the cardiovascular assessment, and prior to beginning the submaximal exercise bike test. Heart rate was recorded at 5 second intervals from telemetry of the RR interval of ventricular depolarisation (Polar Accurex, Electro Oy, Finland). Ratings were made when Study members were seated and in a resting state. The ve readings were averaged to provide an overall measure of resting heart rate. 2.3.3. Anthropometric assessment Measures were obtained in the standing position, with head and eyes directed forward and upper limbs hanging by the sides with palms forward. Height was measured in stocking feet to the nearest millimetre, using a portable Harpenden Stadiometer and calculated as the maximum of the distance from the oor to the vertex of the head (stretch stature). Weight was recorded using calibrated scales (Tanita, Model No. 1609N) in light clothing to the nearest 0.1 of a kilogram. Height and weight measurements were taken twice and the two readings averaged. These values were then used to calculate body mass index (BMI) (weight [kg]/height [m]2).

404

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

Girths were measured using a steel tape, calibrated in centimetres with millimetre gradations. Waist girth was the perimeter at the level of the noticeable waist narrowing located approximately half way between the costal border and the iliac crest. Hip girth was taken as the perimeter at the level of the greatest protuberance and at approximately the symphysion pubis level anteriorally. Measures were repeated and the average used to calculate the waist:hip ratio an index of central adiposity. 2.3.4. Participation in physical activity Participation in vigorous physical activity was assessed with the following questions In the past four weeks, have you done any activity that caused you to breathe hard or puff a lot? This question was modelled on one used for the Life in New Zealand (LINZ) survey (Russell & Wilson, 1991). Examples were given of working out in the gym, playing sport, digging in the garden or activity at work. Those who answered afrmatively were then asked How much time have you spent doing these sorts of activities in a normal week? That includes weekend and weekdays altogether. Time was recorded in minutes. Study members were also asked During a typical week in the last year, on how many days would you spend at least 30 minutes doing at least some physical activity? e.g. walking, gardening, playing sport, being active at work. The total number of days (ranging from 0 through 7) was recorded for each Study member. 2.3.5. Self-reported health status Self-reported health status during the previous 12 months was assessed using the Australian/New Zealand adaptation of the SF-36 Health Survey (Ware, Snow, Kosinski, & Gandek, 1993). This widely-used 36-item questionnaire measures eight aspects of health: physical functioning, role limitations due to physical health, bodily pain, general health perceptions, vitality (energy/fatigue), social functioning, role limitations due to emotional health, and mental health (psychological distress/wellbeing). The questionnaire has been shown to be a valid and reliable measure of the health status of New Zealanders (Scott, Tobias, Sarfati, & Haslett, 1999) and others (see Journal of Clinical Epidemiology, 51(11), 1998). 2.3.6. Attitudes to physical activity and tness Study members were asked to indicate how much they agreed or disagreed with each of six statements about the personal consequences of more regular participation in physical activity (see Table 3). Each statement was rated on a 5-point scale where 1=strongly disagree and 5=strongly agree. Perceived tness was also rated on a 5-point scale, where 0=much less fit than my friends and 4=much more fit than my friends. 2.4. Statistical analysis Analyses include only those Study members who recorded both predicted and actual discomfort, had measurements of cardiorespiratory tness from the exercise bike test and who were not pregnant at the time or suffering from an activity restricting injury or disability. The total number included in the analyses was 574 (59% of the cohort assessed at age 26), with 254 females and 320 males. Chi-square comparisons suggested that the proportion of males to females did not differ between Study members who were and were not asked about predicted and actual dis-

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

405

comfort (p0.1). Measuring prediction of discomfort part-way through the assessment phase is unlikely to have resulted in any systematic bias in the data because Study members were typically booked and interviewed around the time of their birthdays. Data were analysed using SPSS version 10.0 for Windows (SPSS Inc., Chicago, IL). Continuous variables were analysed using t-tests and analysis of variance (ANOVA), while dichotomous variables were compared with the Pearson chi-square test. Where appropriate, Bonferroni posthoc analyses were performed. Not all Study members had scores for all dependent variables, so the degrees of freedom vary slightly across analyses: the number of cases for each dependent variable ranged from 572 to 574. 3. Results Study members were assigned to groups on the basis of their pre- and post exercise predictions of discomfort (see Table 1). The underprediction group included those whose predicted discomfort was less than that actually experienced. The accurate prediction group consisted of those whose predicted and actual discomfort ratings were exactly the same and the overprediction group comprised those who predicted more discomfort than they actually experienced. A Pearson chi-square analysis identied a signicant sex difference with females more likely to overpredict the amount of discomfort they were likely to experience (and males more likely to underpredict) [ c2(2)=8.305, p0.05]. For each physical health measure, a 32 ANOVA was performed where levels of group (under-, accurate and overprediction) and sex (male and female) were included as factors. Measures of physical health differed across the prediction groups, with most differences indicating that individuals in the overprediction group had worse physical health. As shown in Table 2, they had higher BMI and resting heart rate scores and lower cardiorespiratory tness scores (predicted VO2 max adjusted for body weight). Males had higher waist:hip ratio, cardiorespiratory tness and lower resting heart rates than females. No signicant interactions were detected between group and sex. Individuals in the overprediction group also perceived themselves to be less t than their friends
Table 1 Number and percentage of Study members in prediction groups by sex Group Underprediction n % Accurate n % Overprediction n % Total n Female 91 35.8 99 39.0 64 25.2 254 Male 145 45.3 122 38.1 53 16.6 320 Total 236 41.1 221 38.5 117 20.4 574

406

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

Table 2 Prediction group and sex differences for physical health measures Physical health Sex Female BMI Waist:hip ratio Cardiorespiratory tnessb Resting heart rate
a b

Prediction group Male 25.08 84.59** 47.27** 68.63** Under24.52 80.41 46.04 67.92 Accurate Over24.91 79.54 44.11 71.12 25.84*a 79.37 39.68** 74.59**

Post-hoc tests Over vs under * ** ** Over vs accurate ** ** Under vs accurate **

24.77 73.90 39.90 72.87

*p0.05; **p0.01. Calculated in ml/kg/min and adjusted for body weight.

and reported fewer days per week on which they engaged in at least 30 minutes of physical activity (see Table 3). With regard to signicant sex differences, males had greater perceived tness and amount of vigorous activity in a week and were more likely to feel that exercise would earn other peoples respect, whereas females were more likely to agree that they would feel better about themselves with increased exercise. The overprediction group also reported worse Physical Functioning and General Health as measured by these two subscales of the SF36 (see Table 4). Males had signicantly higher (better) scores on the Physical Functioning, Vitality and Mental Health subscales of the SF36 than females. We further tested the possibility that people may not exercise because they predict that it will cause them discomfort, even though they know they should. To do this we divided the Study members into three approximately equal groups (i.e., low, medium, high) according to how many minutes per week they reported spending in vigorous activity, and examined whether predictions of exercise-related discomfort differed across these activity groups. As shown in Table 5, chi square analyses revealed that males were more likely to be in the high activity group and less likely to be in the medium and low activity groups [c2(2)=12.88, p0.01], and that the overprediction group did indeed engage in less frequent vigorous physical activity than the underprediction group [c2(4)=11.79, p0.05]. 4. Discussion Distinctive patterns of predicted and actual discomfort were apparent. Almost 40% of our sample made accurate predictions, and a similar percentage underpredicted discomfort, while the remaining 20% overpredicted the level of physical discomfort they would experience during the submaximal exercise bike test. Prediction patterns were related to many of the physical health measures investigated. Specically, overprediction of physical discomfort was associated with poor self-reported health and lower levels of physical activity, as well as worse scores on all the physical health measures with the exception of waist:hip ratio. These results suggest that cognitive processes occurring in close proximity to physical exercise are important correlates of health behaviours and overall health status. They also emphasise the value of closer examination of

Table 3 Self-reported physical activity and attitudes towards physical activity by sex and prediction group Prediction group Post-hoc tests

Self-reported activity and attitudes Male 2.11**a 2.08 277.72 6.92 6.89 6.68** ** ** 257.54 161.95 2.14 1.71** ** ** 324.28** 6.85 UnderAccurate OverOver vs under Over vs accurate Under vs accurate

Sex

Female

4.07

4.13

3.97

4.25

2.18

2.06

2.14

2.35

Perceived tness 1.91 Physical activity: Minutes of vigorous 148.18 activity in week Days/week with 30+ 6.86 minutes of physical activity Attitudes towards physical activity: I would be healthier if 4.11 I took part in physical activity more regularly I would be probably 2.11 be sore and uncomfortable if I took part in physical activity If I took part in 2.28 physical activity more regularly, my family and friends would get to spend less time with me 2.42 2.40 2.39 2.21

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

(continued on next page)

407

408

Table 3 (continued) Prediction group Post-hoc tests

Self-reported activity and attitudes Male UnderAccurate OverOver vs under Over vs accurate

Sex

Female

Under vs accurate

4.03*

3.77

3.87

3.83

4.00

2.21

2.38*

2.34

2.25

2.33

I would feel better about myself if I took part in physical activity more regularly Other people would respect me more if I took part in physical activity more regularly I would feel that I was wasting my time if I took part in physical activity more regularly 1.55 1.47 1.45 1.61

1.42

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

*p0.05; **p0.01.

Table 4 Relation between sex, prediction group and self-reported general health status Prediction group Male 95.31*c 91.56 79.90 78.05 69.09** 91.13 93.85 81.53** 95.38 92.16 80.93 79.89 67.65 89.99 92.09 80.34 94.59 89.59 79.66 77.48 65.81 91.18 93.97 80.27 92.27* 91.24 77.54 73.70** 65.30 89.96 90.60 77.74 ** ** UnderAccurate OverOver vs under Post-hoc tests Over vs accurate Under vs accurate

SF36 Health Scalesa

Sexb

Female

Physical functioning Role-physical Bodily pain General health Vitality Social functioning Role-emotional Mental health

93.35 90.26 79.56 77.26 63.15 89.57 90.81 77.59

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

Scores are standardised and range from 0100, where a higher score represents better functioning. Differences between females and males had degrees of freedom (1, 565), and between groups had degrees of freedom (2, 565). *p0.05; **p0.01.

409

410

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

Table 5 The relation between sex, prediction group and levels of self-reported vigorous physical activity Physical activity group Sex Female Male Prediction group Under70 29.7 73 30.9 93 39.4 236 Accurate 75 33.9 63 28.5 83 37.6 221 Over47 40.2 44 37.6 26 22.2 117 192 33.4 33.4 180 31.4 31.4 202 35.2 35.2 574 Total

Low (an average of 4 minutes per week) N 96 96 % by sex 37.8 30.0 % of prediction group Medium (an average of 126 minutes per week) N 89 91 % by sex 35.0 28.4 % of prediction group High (an average of 585 minutes per week) N 69 133 % by sex 27.2 41.6 % of prediction group Total N 254 320

exercise-specic cognitions. While our investigation focussed on one particular belief about activity (i.e. predicted discomfort), the ndings are consistent with stage models which predict differences in cognitive structures at different times in the process of increasing activity (e.g., Gorely & Gordon, 1995). An important caveat applies when interpreting these ndings. They reect associations between patterns of prediction and a number of physical health parameters as such no conclusions can be made about direction of causation. For example, it is possible that overprediction of physical discomfort during exercise contributes to poorer physical health via lowered levels of physical activity. Conversely, it is possible that poor physical health inuences characteristic patterns of thinking about physical activity. A third possibility also exists: that at least some of the observed association is due to unmeasured factors such as personality or affective state. In this regard, previous research has demonstrated the impact of these factors on both cognitive appraisal processes and physical health outcomes (e.g., Cohen & Rodriguez, 1995; Watson & Pennebaker, 1989). Hence, we urge against inferring cause and effect from these data until additional, prospective data have been collected. With this cautionary note in mind, it appears that studying specic thought processes and attitudes in close temporal proximity to physical activity may be worthwhile for understanding factors that lead to the uptake and maintenance of a physically active lifestyle. It is also of interest to compare our data with ndings from studies of the overprediction of pain (e.g., Rachman & Lopatka, 1988; Rachman & Arntz, 1991) and fear (e.g. Rachman & Bichard, 1988; Marks & de Silva, 1994; Arntz, 1997). Pain studies have shown that among people who overpredict pain, subsequent predictions tend to decrease with repeated exposure to pain stimuli, whereas underpredictions result in an increase in subsequent predictions (after correct predictions, future predictions remain essentially unchanged). However, these adjustments tend to occur only in the short-term (hours to days), as

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

411

longer term (months) predictions tend to be more strongly inuenced by an individuals original expectations than by recent experiences (Arntz, van Eck, & Heijmans, 1990). Interestingly, experimental studies examining prediction match/mismatches have shown that underpredictions of pain are most strongly associated with anticipatory anxiety and avoidance (reviewed in Arntz, 1997). These and other ndings have led to the suggestion that it may be normative (or even adaptive) to overpredict initially, and to modify these predictions (i.e., become more accurate) as new information becomes available via repeated exposure. However, it seems likely that for a subset of the population, avoidance or lack of exposure may actually protect excessive or unrealistic overpredictions against disconrmation, and thereby perpetuate avoidance behaviour and its negative consequences (Rachman & Arntz, 1991; Rachman, 1990). With regard to the fear literature, a number of researchers have investigated the relation between cognitive appraisal processes and avoidance behaviour (e.g. Rachman & Bichard, 1988; Butler & Mathews, 1983, 1987; Lucock & Salkovskis, 1988; Telch, Brouillard, Telch, Agras, & Taylor, 1989; Taylor & Rachman, 1994; Poulton & Andrews, 1994; Andrews, Freed, & Teesson, 1994; Menzies & Clarke, 1995). In a number of these studies appraisal processes were measured as the product of likelihood and cost estimates (see Carr, 1974). In general, when both are rated high, then the object or situation is likely to be avoided. Extrapolating to the physical activity data suggests that individuals who believe it is highly probable that discomfort will occur and/or that the discomfort be aversive (i.e. costly), may be more likely to avoid physical exercise and therefore miss the opportunity to disconrm erroneous expectations. Our data showing that the overprediction group had the lowest levels of self-reported vigorous physical activity and number of days per week involved in at least 30 minutes of activity is consistent with this proposition. 4.1. The prediction matchmismatch paradigm and physical activity Notwithstanding the associative nature of our ndings, they suggest that using the prediction matchmismatch paradigm for investigation of factors associated with the uptake of exercise may be useful for a number of reasons. First, it is possible that physical inactivity may partially result from specic physical activity-related cognitions, as well as other factors such as indifference, incovenience or lack of time or motivation (cf., Owen & Bauman, 1992; Booth, Bauman, Owen, & Gore, 1997). At the same time, it needs to be acknowledged that as levels of tness decrease overall, the likelihood of discomfort upon exercise increases. Accurate predictions (i.e., realistic expectations) about levels of discomfort seem particularly important if individuals are to persist in their attempts to move from a sedentary to a physically active lifestyle (Wing & Jakicic, 2000). Second, this paradigm would allow identication of those specic, unrealistic cognitions (e.g., unrealistically high cost or probability estimates) that need to be altered in order to facilitate uptake of activity. Importantly, this psychological approach can be implemented at both the individual and population level. For both, realistic expectations about the levels of discomfort during and following (both immediate and delayed) physical exercise may be of benet, especially given the typical cognitive adjustments following under-, accurate and overpredictions of pain (Rachman & Arntz, 1991). Third, this approach acknowledges individual variation in pain sensitivity (and by extension, discomfort threshold) in the population and suggests strategies for assisting individuals who score highly on these dimensions (e.g., Barsky, Orav, Delamater, Clancy, & Hartley, 1998). The corollary of the above is that among individuals predisposed to experiencing

412

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

somatic distress, a proportion are likely to overpredict the level of discomfort associated with physical exercise. This could lead to excessive avoidance of exercise, possibly at a time when physical activity forms a key part of treatment as in, for example, the case of chronic fatigue sufferers (e.g., Fulcher & White, 2000; Barsky & Borus, 1999; also see Philips, 1987, for a more general discussion of the counterproductive aspects of pain-avoidance behaviour). 4.2. Public health implications Overprediction of physical discomfort may be important from a public health perspective. Participation in small amounts of regular, moderate intensity physical activity can protect against a number of negative health outcomes and vigorous physical activity can further improve health and tness (US Department of Health and Human Services, 1996). The greatest impact on population health, however, will be achieved by increasing moderate levels of activity among those who are currently sedentary or inactive (Bauman & Egger, 2000). This group is therefore the most appropriate focus for health promotion intervention strategies (e.g. Active Australia, 1999). Our results indicate a relationship between overprediction of discomfort and a number of risk factors for physical ill health, including low levels of physical activity, though importantly, not attitudes towards increasing physical activity. Further, females were more likely to overpredict discomfort than males, and they tended to be less active, have poorer physical health and reported different attitudes towards increasing physical activity. This pattern of sex differences indicates a need for appropriate, sex-specic health promotion messages and intervention strategies. The ndings also suggest the importance of ensuring that programmes designed to increase activity among sedentary and less active people aim to increase activity gradually and in manageable steps so as to minimise negative experiences and the risk of subsequent relapse into inactivity. It is important to maximise the immediately experienced benets of increased activity and ensure that these outweigh negative consequences. This approach is comparable to graduated exposure procedures used in the treatment of phobic (and other) disorders. The goal of public health campaigns aimed at the physically inactive is for increased activity to become a non-threatening, attractive, rewarding and integral part of everyday personal and social life. Understanding activityrelated cognitions appears important in this regard. 4.3. Conclusions Different patterns of physical discomfort prediction are meaningfully related to a number of important health parameters including cardiorespiratory tness, BMI and resting heart rate as well as the frequency and type of exercise activity, self-report health status and attitudes to physical activity. Future research should determine if and how physical discomfort prediction patterns are related to changes in health risk/promoting behaviours and physical health status over time. Acknowledgements The Dunedin Multidisciplinary Health and Development Research Unit is supported by the Health Research Council of New Zealand. The Social and Behavioural Research in Cancer Group

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

413

receives support from the Cancer Society of New Zealand Inc., the Health Sponsorship Council and the University of Otago. Data collection was supported by a New Zealand National Heart Foundation grant to Dr Poulton. The authors are indebted to Air New Zealand, Dr Phil Silva, founder of the Study, and the Study members for their continued participation and support.

References
Active Australia (1999). National physical activity guidelines for Australians. Canberra: Commonwealth Department of Health and Aged Care. Andrews, G., Freed, S., & Teesson, M. (1994). Proximity and anticipation of a negative outcome in phobias. Behaviour Research and Therapy, 32, 643645. Arntz, A. (1997). The matchmismatch model of phobia acquisition. In G. C. L. Davey, Phobias: a handbook of theory, research and treatment. Chichester, UK: Wiley. Arntz, A., van Eck, M., & Heijmans, M. (1990). Predictions of dental pain. The fear of any expected evil is worse than the evil itself. Behaviour Research and Therapy, 28, 2942. Astrand, I. (1960). Aerobic work capacity in men and women with special reference to age. Acta Physiology Scandinavica, 49 (Suppl. 169), 192. Barsky, A. J., & Borus, J. F. (1999). Functional somatic syndromes. Annals of Internal Medicine, 130, 910921. Barsky, A. J., Orav, J. E., Delamater, B. A., Clancy, S. A., & Hartley, L. H. (1998). Cardiorespiratory symptoms in response to physiological arousal. Psychosomatic Medicine, 60, 604609. Bauman, A., & Egger, G. (2000). The dawning of a new era for physical inactivity as a health risk factor. Australia and New Zealand Journal of Medicine, 30, 6567. Blair, S. N., & Brodney, S. (1999). Effects of physical inactivity and obesity on morbidity and mortality: current evidence and research issues. Medicine and Science in Sports and Exercise, 31 (Suppl.), S646S662. Bouchard, C. (2000). Physical activity and obesity. Champaign, IL: Human Kinetics. Booth, M. L., Bauman, A., Owen, N., & Gore, C. L. (1997). Physical activity preferences, preferred sources of assistance, and perceived barriers to increased activity among physically inactive Australians. Preventive Medicine, 26, 131137. Butler, G., & Mathews, A. (1983). Cognitive processes in anxiety. Advances in Behaviour Research and Therapy, 5, 5162. Butler, G., & Mathews, A. (1987). Anticipatory anxiety and risk perception. Cognitive Therapy and Research, 11, 551565. Carr, A. T. (1974). Compulsive neurosis: a review of the literature. Psychological Bulletin, 81, 311318. Cohen, S., & Rodriguez, M. (1995). Pathways linking affective disturbances and physical disorders. Health Psychology, 14, 374380. Fentem, P. (1996). A national strategy for the promotion of physical activity. British Journal of Sports Medicine, 30, 280281. Flegal, K. M., Carroll, M. D., Kuczmarski, R. J., & Johnson, C. L. (1998). Overweight and obesity in the United States: prevalence and trends, 19601994. International Journal of Obesity, 22, 3947. Fulcher, K. Y., & White, P. D. (2000). Strength and physiological response to exercise in patients with chronic fatigue syndrome. Journal of Neurology, Neurosurgery and Psychiatry, 69, 302307. Godin, G. (1994). Theories of reasoned action and planned behaviour: usefulness for exercise promotion. Medicine and Science in Sports and Exercise, 26, 13911394. Gorely, T., & Gordon, S. (1995). An examination of the transtheoretical model and exercise behaviour in older adults. Journal of Sport and Exercise Psychology, 17, 312324. Grundy, S. M., Blackburn, G., Higgins, M., Lauer, R., Perri, M. G., & Ryan, D. (1999). Physical activity in the prevention and treatment of obesity and its comorbidites: evidence report of independent panel to assess the role of physical activity in the treatment of obesity and its comorbidities. Medicine and Science in Sports and Exercise, 31, 14931500.

414

R. Poulton et al. / Behaviour Research and Therapy 40 (2002) 401414

Hausenblas, H. A., Carron, A. V., & Mack, D. E. (1997). Application of the theories of reasoned action and planned behavior: a meta-analysis. Journal of Sport and Exercise Psychology, 19, 3651. Lucock, M. P., & Salkovskis, P. M. (1988). Cognitive factors in social anxiety and its treatment. Behaviour Research and Therapy, 26, 297302. Marcus, B. H., & Simkin, L. R. (1993). The stages of exercise behavior. Journal of Sports Medicine and Physical Fitness, 33, 8388. Marks, M., & De Silva, P. (1994). The match/mismatch model of fear: empirical status and clinical implications. Behaviour Research and Therapy, 32, 759770. Menzies, R. G., & Clarke, C. (1995). Danger expectancies and insight in acrophobia. Behaviour Research and Therapy, 33, 215221. NIH (1996). Consensus development panel on physical activity and cardiovascular health. JAMA, 276, 241246. Owen, N., & Bauman, A. (1992). The descriptive epidemiology of a sedentary lifestyle in adult Australians. International Journal of Epidemiology, 21, 305310. Philips, H. C. (1987). Avoidance behaviour and its role in sustaining chronic pain. Behaviour Research and Therapy, 25, 273279. Poulton, R., & Andrews, G. (1994). Appraisal of danger and proximity in social phobics. Behaviour Research and Therapy, 32, 639642. Prochaska, J. O., & Marcus, B. H. (1994). The transtheoretical model: applications to exercise. In R. K. Dishman, Advances in exercise adherence. Champaign, IL: Human Kinetics. Rachman, S. (1990). Fear and courage. (2nd ed). New York: W.H. Freeman. Rachman, S., & Arntz, A. (1991). The overprediction and underprediction of pain. Clinical Psychology Review, 11, 339355. Rachman, S., & Bichard, S. (1988). The overprediction of fear. Clinical Psychology Review, 8, 303313. Rachman, S., & Lopatka, C. (1988). Accurate and inaccurate predictions of pain. Behaviour Research and Therapy, 26, 291296. Russell, D., & Wilson, N. (1991). Executive overview. In Life in New Zealand Commission report: vol. 1.. Otago: Hillary Commission for Recreation and Sport, University of Otago. Scott, K. M., Tobias, M. I., Sarfati, D., & Haslett, S. J. (1999). SF-36 health survey reliability, validity and norms for New Zealand. Australian and New Zealand Journal of Public Health, 23, 401406. Silva, P. A., & Stanton, W. R. (1996). From child to adult: the Dunedin multidisciplinary health and development study. Auckland: Oxford University Press. Sleivert, G., Hopkins, W., Reeder, A. I., & Poulton, R. (2001). Changes in cardiovascular tness and physical activity from adolescence to early-adulthood. Medicine and Science in Sports and Exercise (submitted for publication). Snell, P. G., & Mitchell, J. H. (1999). Physical inactivity: an easily modied risk factor? Circulation, 100, 24. Taylor, S., & Rachman, S. (1994). Stimulus estimation and the overprediction of fear. British Journal of Clinical Psychology, 33, 173181. Telch, M. J., Brouillard, M., Telch, C. F., Agras, W. S., & Taylor, C. B. (1989). Role of cognitive appraisal in panicrelated avoidance. Behaviour Research and Therapy, 27, 373383. US Department of Health and Human Services (1996). Physical activity and health: a report of the Surgeon General. Atlanta, GA: Centers for Disease Control and Prevention. Ware, J. E., Snow, K. K., Kosinski, M., & Gandek, B. (1993). SF-36 health survey manual and interpretation guide. Boston, MA: The Health Institute. Watson, D., & Pennebaker, J. (1989). Health complaints, stress, and distress: exploring the central role of negative affectivity. Psychological Review, 96, 234254. Wing, R. R., & Jakicic, J. M. (2000). Changing lifestyle: moving from sedentary to active. In C. Bouchard, Physical activity and obesity. Champaign, IL: Human Kinetics. World Health Organisation (1998). Obesity: preventing and managing the global epidemic. Report of a WHO consultation on obesity. Geneva: World Health Organisation.