© International Epidemiological Association 2001 Printed in Great Britain International Journal of Epidemiology 2001;30:363–369 Time urgency and risk of non-fatal myocardial infarction Stephen R Cole,a,b Ichiro Kawachi,b Simin Liu,a J Michael Gaziano,a JoAnn E Manson,a,c,d Julie E Buring a,b,c,d and Charles H Hennekense Background Inconsistencies in the literature linking Type A behaviour pattern (TAB) to coronary heart disease (CHD) may be due to differences in the effects of various components of TAB, namely aggressiveness, hostility, ambitiousness, competitive drive, and a chronic sense of time urgency. Methods We investigated the association between sense of time urgency/impatience and non-fatal myocardial infarction (MI) in a study of 340 cases and an equal number of age-, sex-, and community-matched controls. Results A dose-response relation was apparent among subjects who rated themselves higher on the four-item time urgency/impatience scale (P-value for trend ,0.001), with a matched odds ratio (OR) for non-fatal MI of 4.45 (95% CI : 2.20–8.99) comparing those with the highest rating to those with the lowest. After further adjustment for family history of premature MI, physical activity, body mass index, occupation, cigarette smoking, total caloric intake, per cent calories from saturated fat, alcohol intake, lipid levels, treated hypertension and diabetes, the dose-response relation remained (P-value for trend = 0.015) and the adjusted OR for MI was 3.99 (95% CI : 1.32–12.0) comparing those with the highest rating to those with the lowest. Conclusion In these data, a sense of time urgency/impatience was associated with a doseresponse increase in risk of non-fatal MI, independent of other risk factors. Prospective cohort studies of time urgency/impatience and incident CHD events are needed to confirm or refute these observations from a case-control study. Keywords Coronary heart disease, non-fatal myocardial infarction, time urgency, impatience, Type A behaviour pattern, case-control study Accepted 24 May 2000 Data concerning Type A behaviour pattern (TAB) and coronary heart disease (CHD) are inconsistent. Type A behaviour pattern is an amalgamation of several elements, including aggressiveness, hostility, ambitiousness, competitive drive, and a chronic sense of time urgency.1 Some of these, referred to as ‘toxic’ a Division of Preventive Medicine, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA. b Department of Health and Social Behavior, Harvard School of Public Health, Boston, MA, USA. c Department of Ambulatory Care and Prevention, Harvard Medical School, Boston, MA, USA. d Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA. e Visiting Professor of Medicine, and Epidemiology and Public Health, University of Miami School of Medicine, USA. Current address: 1415 West Camino Real, Boca Raton, FL 33486. Reprint requests: Dr SR Cole, Division of Preventive Medicine, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, 900 Commonwealth Avenue East, Boston, MA 02215, USA. E-mail: scole@rics. bwh.harvard.edu 363 elements, may be associated with risk of CHD, while others may not.2 Thus, simple summary scores for TAB of linear combinations of these various behavioural patterns would predict CHD only in samples where the toxic element(s) are weighted strongly in the summary score. Such a situation could arise using a TAB scale that incorporates a large number of items from the toxic relative to the non-toxic elements. This is a measurement problem where the instrument is recording several constructs instead of a single construct, i.e. lack of unidimensionality.3,4 Much of the literature on TAB has focused on hostility as the toxic component of the behaviour pattern.5 However, an inadequate number of studies have examined the cardiovascular risk associated with other components of the behaviour pattern,6,7 such as time urgency, which is characterized as a persistent preoccupation with time and need to complete tasks in a hurry. Indeed, the gold standard for measurement of TAB is the videotaped clinical examination (VCE),8 which devotes approximately equal emphasis to the measurement of time urgency and freefloating hostility. 364 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY We previously reported9 on the relation of both overall TAB and suppressed anger with non-fatal myocardial infarction (MI). It remains unclear, however, whether there is any association between the time urgency/impatience component of TAB and MI, and if so, whether any association is independent of traditional coronary risk factors. Noting that researchers may have prematurely abandoned time urgency as a potentially toxic component of TAB, we sought to determine if time urgency, measured with a self-report questionnaire, is associated with non-fatal MI using data from the Boston Area Health Study. Methods Subjects The Boston Area Health Study was a case-control study of 340 patients with first MI and an equal number of age-, sex-, and community-matched control subjects. Case subjects were selected from admissions to the coronary or intensive care units of six suburban Boston hospitals (Emerson, Framingham Union, Leonard Morse, Mount Auburn, Newton-Wellesley and Waltham) between 1 January 1982 and 31 December 1983. Those eligible for inclusion were white men and women under 76 years old living in the Boston area with no previous history of MI or angina pectoris. Using hospital records, the diagnosis of MI was confirmed based on clinical history accompanied by rise in creatine kinase enzyme. Permission was sought from each admitting physician, and informed consent was obtained from the patients in the hospital. In short, for each case patient, a control subject of the same sex and age (±5 years) was selected at random from the residents’ list of the town in which the patient resided. A detailed description of the control sampling strategy has been previously published.10 Each subject was interviewed in his or her home by one of two trained nurse interviewers approximately 8 weeks after discharge from the hospital. Of the eligible subjects contacted 84% of the cases discharged alive and 60% of controls were enrolled, yielding 340 casecontrol pairs. Assessment of time urgency and impatience We ascertained a sense of time urgency/impatience using four items from the 10-item Framingham Type A scale. This 10-item scale includes three items related to work strain, two items on competitiveness, one item on bossiness, two items on impatience, and two items related to time pressure or urgency. The first time urgency item was ‘Have you often felt very pressed for time?’ The second time urgency item was ‘Usually feel pressed for time’. We scored a positive response to each item as one point. For the two impatience items, subjects were asked to rate how well the traits ‘eating too quickly’ and ‘getting upset when you have to wait for anything’ described them, using the response options of not at all, somewhat, fairly well and very well. For these two items, we scored a response of ‘very well’ as one point. For each subject, we combined these four items into a summed rating scale, with a range of 0 to 4 points. Assessment of coronary risk factors We obtained extensive information on coronary risk factors related specifically to the time before the MI for cases and before the interview for controls, including age (years), body mass index (BMI; weight [kg]/height [m2]), cigarette smoking status (never, former, current [,1 pack/day, 1–,2 packs/day, 2+ packs/ day]), family history of premature MI (defined as positive history of MI before age 60 years in the individual’s mother, father, or siblings), history of treatment for hypertension (yes/no) and diabetes mellitus (yes/no). We recorded the ‘usual’ occupation for each subject using Edwards’ US census grouping of occupations. White-collar occupations included (1) professionals, (2) proprietors, managers and officials (including farm owners and wholesale and retail dealers), and (3) clerks and kindred workers. Blue-collar occupations included (4) skilled workers and foremen, (5) semiskilled workers, and (6) unskilled workers (farm labourers and other labourers). Additionally, we created a category for homemakers. Total caloric intake, per cent of calories from saturated fat and alcohol intake were determined from a 116-item semiquantitative food frequency questionnaire.11 We estimated energy expenditure (kJ) from physical activity with questions about walking, climbing stairs, and participation in sports or recreational activities.12 We determined lipid levels from fasting venous blood samples using the Lipid Research Clinics methods.13 Fasting blood samples were obtained from a subgroup of the cases approximately 8 weeks after hospital discharge, as well as from a subgroup of controls. Statistical analysis for lipids is restricted to the 235 matched pairs with full lipid profiles. Statistical analysis We first examined the distribution of coronary risk factors according to case-control status. For these comparisons, we employed continuity-corrected χ2 tests for discrete characteristics (sex, treated hypertension and diabetes, family history of premature MI, cigarette smoking status and occupation type), Wilcoxon rank-sum tests for skewed continuous characteristics (physical activity, alcohol intake, very low density lipoprotein [VLDL] and triglycerides) and t-tests for the normally distributed continuous characteristics (age, BMI, total calories, per cent calories from saturated fat, and the balance of the lipid values). We adjusted lipid values for age and sex before making comparisons. We then examined the distribution of these characteristics according to level of the time urgency/impatience scale. For these comparisons, we employed the Cochran-Armitage test for trend for discrete characteristics (sex, treated hypertension and diabetes, family history of premature MI, cigarette smoking status and occupation type) and the Mantel-Haenszel non-zero correlation test for continuous characteristics (age, BMI, physical activity, alcohol intake, total calories, per cent calories from saturated fat, and lipid values). We assessed the internal consistency reliability of the four-item time urgency/impatience measure using Cronbach’s coefficient alpha.14 To examine the association between our measure of time urgency/impatience and non-fatal MI, we used conditional logistic regression to estimate matched pair odds ratios (OR) and 95% CI.15 We took a log transformation for physical activity, VLDL and triglycerides and categorized alcohol intake into quintiles to adjust for the skew in each of these covariates when controlling for them in the multivariable models. Next, we fit a series of models adjusting for each covariate to assess potential confounding or mediating effects, and a full model adjusting for all covariates. In a secondary set of analyses, we explored the association of time urgency/impatience with non-fatal MI beyond the effects of 365 TIME URGENCY AND MYOCARDIAL INFARCT other components of TAB. To do so, we adjusted the association between time urgency/impatience and non-fatal MI for the six remaining items from the Framingham Type A scale. Table 1 Characteristics of cases and controls, Boston Area Health Study 1982–1983 Characteristica Cases (N = 340)b Controls (N = 340)c P-valued Results Age (years) 58 (10) 58 (10) Sex (% male) 78 78 1.00 Of the 680 subjects, 380 (56%) indicated yes to the item ‘often feel very pressed for time.’ Of the 679 subjects who responded to the remaining three items, 469 (69%) indicated that they ‘usually feel pressed for time,’ 194 (29%) responded that ‘eat too quickly’ described them ‘very well’, and 201 (30%) responded that ‘get upset when you have to wait for anything’ described them ‘very well’. The four time urgency/impatience items combined to provide a scale with a median of 2 (mean = 1.81; quartiles 1,3; range 0–4) and an internal consistency reliability of 0.50. As expected, the distribution of traditional coronary risk factors among cases differed from controls (Table 1). In Table 2, a stronger sense of time urgency/impatience was clearly associated with younger age (P-value , 0.001), male gender (P-value = 0.003), higher levels of smoking (P-value , 0.001), whitecollar occupation (P-value , 0.001), higher levels of physical activity (P-value = 0.047), and higher caloric (P-value , 0.001) and saturated fat (P-value = 0.013) intake. In crude matched-pair analysis, the measure of time urgency/ impatience was associated with a dose-response increase in risk of non-fatal MI (P-value for linear trend ,0.001; Table 3, Figure 1). The OR for non-fatal MI among those with the highest rating of time urgency/impatience was 4.45 (95% CI : 2.20–8.99) compared to those with the lowest rating of time urgency/impatience. In multivariable models controlling for age and sex by design and family history of premature MI, physical activity, BMI (continuous), occupation, cigarette smoking, total caloric intake, per cent calories from saturated fat, alcohol intake, lipids, history of treated hypertension and diabetes, the measure of time urgency/impatience remained significantly associated with non-fatal MI (P-value for linear trend = 0.015; Table 3, Figure 1). The adjusted OR for non-fatal MI among those with the highest rating of time urgency/impatience was 3.99 (95% CI : 1.32–12.0), while the adjusted OR among those with a moderate rating of time urgency/impatience (scale score = 2) was 1.99 (95% CI : 0.83–4.82), both compared to those with the lowest rating of time urgency/impatience. As smoking is the strongest potential confounder, we limited our study to the 193 subjects with no history of smoking and conducted an unmatched analysis. While results are much less precise due to the reduction in sample size, time urgency/ impatience remained associated with risk of non-fatal MI (OR comparing highest time urgency/impatience level to lowest = 2.03, 95% CI : 0.61–6.79). When we looked at each time urgency/impatience item individually, the matched pair OR were 1.43 (95% CI : 1.02–2.00) for ‘often feel very pressed for time at work,’ 1.67 (95% CI : 1.22– 2.28) for ‘usually feel pressed for time,’ 1.90 (95% CI : 1.34– 2.69) for ‘get upset when you have to wait for anything,’ and 1.36 (95% CI : 0.98–1.89) for ‘eat too quickly.’ Therefore all four items contributed to the association between the summary scale and increased risk of non-fatal MI. Finally, we adjusted for the six Framingham Type A items which measured other components of TAB (work strain, Treated hypertension (%) 35 25 0.008 Treated diabetes (%) 14 8 0.018 Family history of MIe ,60 years of age (%) 22 15 Never 25 32 Former 32 41 6 7 20–39 (cigs/day) 17 12 40+ (cigs/day) 19 8 White collar 62 70 Blue collar 30 23 Occupation (%) Homemaker Body mass index (kg/m2) Physical activity (kcal/week) 0.030 ,0.001 Cigarette smoking (%) Current: 1–19 (cigs/day) 0.952 0.108 8 7 26 (4) 26 (4) 0.483 3090 (2970) 3542 (3009) 0.007 Total calories per day 2482 (927) 2354 (773) 0.050 Per cent calories from saturated fat 13 (3) 12 (3) ,0.001 Alcohol intake (g/day) Total cholesterol (mg/dl)f Total HDL g (mg/dl)f 15 (26) 20 (28) 0.001 215 (43) 212 (42) 0.325 35 (10) 43 (12) ,0.001 HDL 2 (mg/dl)f 12 (7) 18 (10) ,0.001 HDL 3 (mg/dl)f 23 (6) 26 (7) ,0.001 LDLh (mg/dl)f 137 (37) 132 (35) 0.116 VLDLi (mg/dl)f 43 (24) 37 (26) ,0.001 175 (116) 136 (96) ,0.001 Triglycerides (mg/dl)f a All values are mean (±SD) unless otherwise indicated. b One case was missing smoking status and 10 cases were missing occupation. c One control was missing history of treatment for hypertension, 1 missing smoking status and 14 controls were missing occupation. d P-value from continuity corrected χ2, Cochran-Armitage trend test (smoking status), t-test, or Wilcoxon rank-sum test (physical activity, alcohol intake, VLDL, and triglycerides). e Myocardial infarction. f Adjusted for age and sex. g High density lipoprotein. h Low density lipoprotein. i Very low density lipoprotein. competitiveness, and bossiness). In this model, time urgency/ impatience remained independently associated with non-fatal MI (P-value for linear trend = 0.001). Here, the TAB-adjusted OR for non-fatal MI among those with the highest rating of time urgency/impatience was 4.00 (95% CI : 1.83–8.75), while the TAB-adjusted OR among those with moderate ratings of time urgency/impatience were also elevated compared to those with the lowest rating of time urgency/impatience (TABadjusted OR for scale score of 3 = 1.56 [95% CI : 0.85–2.87], TAB-adjusted OR for scale score of 2 = 1.23 [95% CI : 0.74–2.03]). 366 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Table 2 Risk factors by time urgency/impatience, Boston Area Health Study 1982–1983 Score on four-item time urgency/impatience measure Characteristica Low–0 (N = 106) 1 (N = 156) 2 (N = 228) 3 (N = 124) High–4 (N = 65) P-value for trendb 63 (9) 61 (9) 57 (9) 55 (10) 53 (10) ,0.001 Sex (% male) 69 77 80 81 88 0.003 Treated hypertension (%) 31 28 29 29 35 0.671 Treated diabetes (%) 16 11 8 8 12 0.164 Family history of MIc (%) 19 17 20 15 28 0.437 Never 38 29 26 26 23 Former 38 37 36 35 32 Age (years) ,0.001 Cigarette smoking (%) Current: 1–19 (cigs/day) 9 6 6 6 8 20–39 (cigs/day) 8 14 18 18 14 40+ (cigs/day) 8 14 14 15 23 White collar 51 59 73 67 79 Blue collar 35 32 23 26 18 Homemaker 14 9 4 7 3 26 (4) 26 (4) 26 (4) 26 (3) 26 (3) 0.443 3382 (2984) 2837 (2273) 3285 (2976) 3578 (3113) 4005 (4102) 0.047 2169 (814) 2297 (728) 2487 (838) 2525 (933) 2678 (985) ,0.001 12(4) 12 (3) 13 (3) 13 (3) 13 (3) 0.013 15 (30) 17 (28) 20 (28) 15 (23) 17 (25) 0.857 ,0.001 Occupation (%) Body mass index (kg/m2) Physical activity (kcal/week) Total calories per day Per cent calories from saturated fat Alcohol intake (g/day) Total cholesterol (mg/dl)d 218 (42) 215 (49) 211 (42) 215 (46) 215 (38) 0.669 Total HDLe (mg/dl)d 39 (12) 40 (13) 40 (12) 38 (11) 36 (10) 0.254 HDL 2 (mg/dl)d 16 (10) 15 (9) 15 (9) 14 (9) 13 (8) 0.153 HDL 3 (mg/dl)d 23 (7) 25 (7) 25 (7) 24 (6) 24 (6) 0.949 136 (34) 131 (39) 134 (37) 136 (38) 135 (36) 0.821 44 (29) 40 (25) 37 (24) 41 (23) 43 (27) 0.514 162 (89) 155 (97) 144 (123) 156 (92) 187 (152) 0.776 LDLf (mg/dl)d VLDLg (mg/dl)d Triglycerides (mg/dl)d a All values are mean (±SD) unless otherwise indicated. b P-value from Cochran-Armitage trend test or Mantel-Haenszel non-zero correlation test (age, body mass index, physical activity, alcohol intake, total calories, per cent calories from fat, and lipid values). c Myocardial infarction. d Adjusted for age and sex. e High density lipoprotein. f Low density lipoprotein. g Very low density lipoprotein. Discussion These data indicate that individuals reporting a greater sense of time urgency/impatience have an increased risk of non-fatal MI. This association was independent of age, sex, family history of premature MI, physical activity, BMI, occupation, cigarette smoking, total caloric intake, per cent calories from saturated fat, alcohol intake, treated hypertension and diabetes, and lipid levels. The assessment of hostility includes some aspects of time urgency (e.g. ‘It makes me angry to have people hurry me’). However, time urgency may also be related to other psychological characteristics, such as anxiety. For example, approaches to assess anxiety commonly contain a time urgency component, such as ‘I work under a great deal of tension’ or ‘I am not feeling much pressure or stress these days’. Anxiety16,17 and cognitive responses to anxiety, such as worry,18 have been implicated as risk factors for CHD. Our hypothesis that time urgency would prove a particularly ‘toxic’ element of TAB was based upon the evidence relating anxiety to CHD and the observation that time urgency and anxiety are closely related concepts. In the present study, the association between time urgency and non-fatal MI appeared to be independent of the remaining components of TAB, suggesting that time urgency is indeed an independent predictor of increased risk of non-fatal MI. This case-control study is subject to several limitations. First, as with any case-control study of psychosocial exposures, time urgency and impatience may be differentially recalled among cases and controls. There are several reasons why recall bias is an unlikely explanation for our findings. First, cases were unaware of our hypothesis, and our extensive questionnaire TIME URGENCY AND MYOCARDIAL INFARCT 367 Table 3 Odds ratios for non-fatal myocardial infarction, according to sense of time urgency/impatience, Boston Area Health Study 1982–1983 ‘Usually feeling pressed for time’ No. of caseLow–0 control pairs (N = 106) 1 (N = 156) 2 (N = 228) 3 (N = 124) High–4 P-value (N = 65) for trend Model controlling for: ,0.001 Age, sex 339 1.0 1.15 (0.70–1.90) 1.40 (0.86–2.26) 1.76 (1.02–3.03) 4.45 (2.20–8.99) Age, sex, hypertension 338 1.0 1.12 (0.68–1.86) 1.42 (0.87–2.31) 1.70 (0.98–2.94) 4.28 (2.11–8.69) ,0.001 Age, sex, diabetes 339 1.0 1.20 (0.72–1.98) 1.46 (0.90–2.38) 1.85 (1.06–3.21) 4.59 (2.26–9.32) ,0.001 Age, sex, family history of MIa 339 1.0 1.12 (0.68–1.85) 1.40 (0.86–2.28) 1.81 (1.04–3.13) 4.38 (2.16–8.87) ,0.001 Age, sex, (log) physical activity 339 1.0 1.14 (0.69–1.88) 1.36 (0.84–2.22) 1.75 (1.01–3.03) 4.21 (2.07–8.57) ,0.001 Age, sex, body mass index 338 1.0 1.18 (0.72–1.95) 1.46 (0.89–2.37) 1.80 (1.04–3.12) 4.53 (2.23–9.18) ,0.001 Age, sex, occupation 316 1.0 1.07 (0.63–1.81) 1.27 (0.76–2.13) 1.64 (0.92–2.93) 4.62 (2.16–9.91) ,0.001 Age, sex, smoking statusb 337 1.0 1.04 (0.62–1.75) 1.26 (0.76–2.07) 1.68 (0.95–2.96) 4.29 (2.06–8.93) ,0.001 Age, sex, dietary informationc 339 1.0 1.10 (0.65–1.87) 1.38 (0.82–2.32) 1.79 (1.00–3.20) 3.91 (1.87–8.16) ,0.001 Age, sex, lipidsd 235 1.0 1.88 (0.92–3.86) 2.35 (1.14–4.84) 1.89 (0.85–4.19) 4.50 (1.80–11.2) 0.004 Age, sex, all coronary risk factors 221 1.0 1.39 (0.60–3.24) 1.99 (0.83–4.82) 1.71 (0.64–4.58) 3.99 (1.32–12.0) 0.015 a Myocardial infarction. b Smoking status as: never, former, current (,1 pack/day, 1–,2 packs/day, 2+ packs/day). c Includes: total caloric intake, per cent of calories from saturated fat and quintiles of alcohol intake. d Includes: high density lipoprotein (HDL) , HDL , low density lipoprotein (LDL), log of very density lipoprotein (VLDL), and log of triglycerides. 2 3 Figure 1 Age- and sex- and multivariable-adjusted odds ratios for non-fatal myocardial infarction by level of time urgency, compared to subjects with a time urgency score of zero on lifestyle factors was presented to them as a study of factors involved in hospitalization. Second, the time urgency hypothesis is now, and was then (in 1982–1983), not widely known to the lay public. Finally, the primary research question for this case-control study, which concerned alcohol intake and MI, was widely suspected as a factor influencing MI and no recall bias was apparent in this association.19 A definitive test of the hypothesis ruling out recall bias requires prospective measurement of time urgency/impatience. A further limitation of our study is that our results only apply to non-fatal MI. While selecting only non-fatal MI cases limits generalizability, it allowed us to ascertain measures of time urgency and impatience as well as extensive information on coronary risk factors directly from the subjects, making this study possible and reducing the possibility of confounding by unmeasured factors. Third, it is possible that a survivorship bias is at play, whereby fatal MI cases would have demonstrated lower time urgency than the non-fatal MI cases that we observed. Analysis of prospective data will be necessary to rule out this potential, but unlikely, explanation. Fourth, there may be residual confounding by measured factors. However, it is unlikely that there exists any residual positive confounding of measured factors strong enough to cancel the observed association. Fifth, random misclassification of time urgency/impatience may have occurred. Indeed, we provide only marginal evidence of the reliability of our measurement scale, with an internal consistency reliability estimate of 0.50. However, the effect of any random misclassification would tend to diminish the observed association towards the null. Finally, while information on Q-wave infarction would have been desirable, this information was not available. We previously reported9 that the unadjusted association between overall TAB score and non-fatal MI (matched OR = 1.57; 95% CI : 1.12–2.20) observed in the Boston Area Health Study attenuated markedly after control for high density lipoprotein (HDL) lipid levels (adjusted OR = 1.12; 95% CI : 0.66– 1.90), suggesting either confounding or a pathway through which TAB might affect CHD. However, our present findings indicate that lipid levels did not mediate the association of the time urgency/impatience component of TAB pattern on risk of non-fatal MI. This differing result for the time urgency/impatience component of TAB reinforces the notion that TAB is an amalgamation of disparate elements, for which simple summary scales are problematic.3,4 In 1966, Brozek, et al.20 demonstrated an increased risk of CHD among those reporting the propensity to speak, walk, write, drive, work, and eat quickly even when they did not have to do so. De Backer et al.21 reported a nearly twofold increase in 5-year CHD incidence among those in the highest tertile of the Jenkins Activity Survey subscale S, which measures time urgency, speed and impatience. However, in the VA Normative Aging Study, the Jenkins Activity Survey subscale S was not associated with CHD risk (adjusted RR = 0.86; 95% CI : 0.69–1.09), even though an association was found between Minnesota Multiphasic Personality Inventory (MMPI)-Type A score and CHD.22 BoothKewley and Friedman23 reported results from a meta-analysis of time urgency, speed and impatience and various disease endpoints. Among the nine studies of combined CHD outcomes (i.e. including angina) they observed a significant positive association 368 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY with time urgency (P-value , 0.01), however this finding was attenuated and not significant in the four studies of time urgency and MI (P-value = 0.31). More recently, Powell et al.24 reported an association between time urgency and mortality among 83 non-smoking, non-diabetic women aged 30–63, 6 months after an acute MI. Six deaths occurred during an average of 8.5 years of follow-up, with time urgency associated with a near threefold increase in risk of mortality (RR = 2.86, P-value = 0.02). Time urgency/impatience may affect CHD endpoints through one or more pathways, including decreased heart rate variability (HRV) or adverse health behaviours. Decreased HRV has been implicated as a potential pathway through which anxiety,25 psychological stress,26 and depression27 alter the incidence of CHD. However, no data are currently available on the association between time urgency and HRV. Alternatively, people with a persistent sense of time urgency/impatience may consequently exhibit other health behaviours that place them at greater risk for CHD. For example, people reporting an increased sense of time urgency/impatience may subsequently exhibit reduced physical activity, poor dietary habits, increased alcohol intake and more smoking. However, our data do not suggest that present smoking status, alcohol intake or physical activity substantially mediate the association between time urgency/impatience and non-fatal MI, as adjustment for these health behaviours did not appreciably attenuate the observed association. In summary, our data demonstrate a significant increased risk of non-fatal MI associated with time urgency/impatience. Additionally, the risk of MI among those who have an increased sense of time urgency/impatience is independent of traditional coronary risk factors, and appears to be independent of other components of TAB. It is also unlikely that this association is substantially mediated by health behaviours, such as smoking, physical activity, caloric intake and alcohol intake, or by lipid levels. Similar analysis in prospective cohort studies of time urgency/impatience and incident CHD events, which are not subject to recall bias, are needed to confirm or refute these observations. 3 Holland PW, Wainer H. Differential Item Functioning. New York: Erlbaum, 1993. 4 Dean K, Salem N. Detecting measurement confounding in epidemio- logical research: construct validity in scaling risk behaviours: based on a population sample in Minnesota, USA. J Epidemiol Community Health 1998;52:195–99. 5 Miller TQ, Smith TW, Turner CW, Guijarro ML, Hallet AJ. A meta- analytic review of research on hostility and physical health. Psychol Bull 1996;119:322–48. 6 Wright L. The type A behavior pattern and coronary artery disease: quest for the active ingredients and the elusive mechanism. Am Psychologist 1988;43:2–14. 7 Thoresen CE, Powell LH. Type A behavior pattern: new perspectives on theory, assessment, and intervention. J Consult Clin Psychol 1992; 60:595–604. 8 Friedman M, Fleischmann N, Price VA. Diagnosis of Type A behavior pattern. In: Allan R, Scheidt S (eds). Heart and Mind: The Practice of Cardiac Psychology. Washington, DC: American Psychological Association, 1996, pp.179–98. 9 O’Connor NJ, Manson JE, O’Connor GT, Buring JE. Psychosocial risk factors and nonfatal myocardial infarction. Circulation 1995;92: 1458–64. 10 Buring JE, O’Connor GT, Goldhaber SZ et al. Decreased HDL2 and HDL3 cholesterol, Apo A-I and Apo A-II, and increased risk of myocardial infarction. Circulation 1992;85:22–29. 11 Willett WC, Sampson L, Stampfer MJ et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:51–65. 12 Lee IM, Hsieh CC, Paffenbarger RS Jr. Exercise intensity and longevity in men. The Harvard Alumni Health Study [see comments]. JAMA 1995;273:1179–84. 13 National Heart and Lung Institute. Manual of laboratory operations: Lipid Research Clinics Program. In: Lipid and Lipoprotein Analysis. Bethesda, MD: National Heart and Lung Institute, 1974. 14 Nunnally JC, Bernstein IH. Psychometric Theory. Third Edn. New York: McGraw-Hill, 1994. 15 Stokes ME, Davis CS, Kock GG. Categorical Data Analysis Using the SAS System. Cary, NC: SAS Institute, Inc., 1995. 16 Kawachi I, Colditz GA, Ascherio A et al. Prospective study of phobic anxiety and risk of coronary heart disease in men. Circulation 1994; 89:1992–97. 17 Kawachi I, Sparrow D, Vokonas PS, Weiss ST. Symptoms of anxiety Acknowledgements This study was supported by research grants HL-24423 and HL-21006 from the National Heart, Lung, and Blood Institute. The authors thank the six Boston area hospitals that participated in this study: Emerson Hospital (Dr Marvin H Kendrick), Framingham Union Hospital (Dr Marvin Adner and Dr Gerald Evans), Leonard Morse Hospital (Dr L Frederick Kaplan), Mount Auburn Hospital (Dr Leonard Zir), Newton-Wellesley Hospital (Dr James Sidd), and Waltham-Weston Hospital (Dr Solomon Gabbay). Dr Kawachi is supported by a Career Development Award from the National Heart, Lung, and Blood Institute. and risk of coronary heart disease. The Normative Aging Study [see comments]. Circulation 1994;90:2225–29. 18 Kubzansky LD, Kawachi I, Spiro A III, Weiss ST, Vokonas PS, Sparrow D. Is worrying bad for your heart? A prospective study of worry and coronary heart disease in the Normative Aging Study [see comments]. Circulation 1997;95:818–24. 19 Gaziano JM, Buring JE, Breslow JL et al. Moderate alcohol intake, increased levels of high-density lipoprotein and its subfractions, and decreased risk of myocardial infarction [see comments]. N Engl J Med 1993;329:1829–34. 20 Brozek J, Keys A, Blackburn H. Personality differences between potential coronary and non-coronary subjects. Ann NY Acad Sci 1966; 134:1057–64. 21 De Backer G, Kornitzer M, Kittel F, Dramaix M. Behavior, stress, and References 1 Rosenman RH, Brand RJ, Jenkins D, Friedman M, Straus R, Wurm M. psychosocial traits as risk factors. Prev Med 1983;12:32–36. 22 Kawachi I, Sparrow D, Kubzansky LD, Spiro A III, Vokonas PS, Weiss Coronary heart disease in Western Collaborative Group Study. Final follow-up experience of 81⁄ 2 years. JAMA 1975;233:872–77. ST. Prospective study of a self-report type A scale and risk of coronary heart disease: test of the MMPI-2 type A scale. Circulation 1998;98: 405–12. 2 Hart KE. A moratorium on research using the Jenkins Activity Survey 23 Booth-Kewley S, Friedman HS. Psychological predictors of heart for Type A behavior? J Clin Psychol 1997;53:905–07. disease: a quantitative review. Psychol Bull 1987;101:343–62. TIME URGENCY AND MYOCARDIAL INFARCT 369 24 Powell LH, Shaker LA, Jones BA, Vaccarino LV, Thoresen CE, Pattillo 26 Sloan RP, Shapiro PA, Bigger JT Jr, Bagiella E, Steinman RC, Gorman JR. Psychosocial predictors of mortality in 83 women with premature acute myocardial infarction [see comments]. Psychosom Med 1993;55: 426–33. JM. Cardiac autonomic control and hostility in healthy subjects. Am J Cardiol 1994;74:298–300. 25 Kawachi I, Sparrow D, Vokonas PS, Weiss ST. Decreased heart rate variability in men with phobic anxiety (data from the Normative Aging Study). Am J Cardiol 1995;75:882–85. © International Epidemiological Association 2001 Printed in Great Britain 27 Carney RM, Saunders RD, Freedland KE, Stein P, Rich MW, Jaffe AS. Association of depression with reduced heart rate variability in coronary artery disease. Am J Cardiol 1995;76:562–64. International Journal of Epidemiology 2001;30:369–370 Commentary: Is there any validity to the Type A concept? Kenneth E Hart In 1988, the New England Journal of Medicine published an editorial commentary by Joel Dimsdale,1 who commented on the ‘topsy-turvy’ career of the putative psychosocial risk factor for coronary heart disease (CHD) originally known as the Type A behaviour pattern (TAB), an action-emotion complex observed in people who are aggressively involved in a chronic struggle with life to achieve more and more in less and less time. Since its inception in the 1950s, the quest to legitimize TAB as a genuine risk factor for CHD has indeed been turbulent, suffering numerous setbacks. Most of the controversy came from epidemiological uncertainty generated by mixed results linking TAB to CHD outcomes. At the beginning of the new millennium, the vast majority of contemporary researchers would agree that the simple model linking TAB to CHD is clearly inadequate. Does this inadequacy necessarily mean that the Type A construct has no validity, and that the entire area of investigation should be abandoned altogether? Some people agree, while others beg to differ. The article by Cole and his colleagues2 in this issue of the International Journal of Epidemiology entitled ‘Time urgency and risk of non-fatal myocardial infarction’ represents the latest in a growing chain of evidence suggesting we might be able to salvage something useful from the chequered career of the Type A concept. The Cole et al.2 paper supports the more general conclusion that researchers should not rush to abandon the use of the paper-and-pencil measures for assessing components of Type A behaviour. Briefly, Cole et al.2 culled four items from the Framingham Type A scale to examine the relationship between subjective ratings of ‘Time Urgency & Impatience’ (TUI) and non-fatal myocardial infarction (MI) in 240 cases and an equal number of matched controls. Their results showed time urgency and impatience was associated with a dose-response increase in risk of MI, independent of other risk factors. Importantly, the association remained significant after controlling for ‘global’ Type A scores. School of Psychology, University of Leeds, Leeds, UK. E-mail: kenh@psyc. leeds.ac.uk This brief commentary seeks to provide a broader context within which to interpret findings published by Cole et al. With regard to the historical backdrop, research on TAB has evolved to a point where most researchers now recognise that the construct is multidimensional in nature, and that different components have differential relevance for health, illness and medical problems resulting from a variety of causes.3 Increased recognition that we are dealing with a family of related constructs has even caused researchers to examine ‘cousin’ constructs. In particular, the field has witnessed a dramatic shift away from (global) Type A research toward research on cynical hostility (see 4 for a meta-analysis). The ‘unravelling’ of the Type A construct has left another legacy. As I have noted in a previous review article,5 this consists of the growth of an interesting and potentially fruitful body of literature relating the ‘Time Urgency & Irritability’ (TUI) subcomponent of the Type A construct to a variety of measures of general ill health. In general, this emerging literature supports the view that not all Type A behaviours are necessarily coronary-prone or even health-relevant. Evidence now suggests that some Type A components are irrelevant to health and illness, but might still predict occupational performance. Moreover, this fledgling knowledge base suggests the TUI subcomponent may be uniquely associated with a variety of health outcomes and a negative risk profile which increases susceptibility to general ill health. This risk ‘profile’ consists of the presence of vulnerability factors and the absence of health protective factors. Just to take one example, there is evidence to suggest that individuals are vulnerable to health problems resulting from (car) accidents by virtue of their TUI scores and not their global Type A scores.6 When considered in the aggregate, this growing data-base underscores the more general theme that subcomponent measures of Type A-related constructs are superior to global undifferentiated measures with respect to understanding vulnerability to different types of medical problems. It seems appropriate at this time to take stock and evaluate the state of theory and assessment in this area. A cluster of 370 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY interrelated measurement scales can readily be identified, starting with Cole et al,2 who culled four items from the Framingham Type A scale to develop a scale they believe measures ‘Time Urgency & Impatience’ (TUI). Wright et al7 have developed a 42-item scale to assess a construct they labelled ‘Time Urgency & Perpetual Activation,’ (TUPA), and have provided preliminary evidence for construct validity in research that pitted the TUPA against other measures that ostensibly measure the same construct. Spence et al8 factor analysed the student version of the Jenkins Activity Survey (JAS-Form T) to develop a five-item subscale they labelled the Impatience-Irritability (II) scale, finding the scale was uniquely related to physical health. Landy et al.9 have also studied the construct of ‘time urgency’ in relation to health outcomes by developing a new multidimensional scale called the Behaviourally Anchored Rating Scale (BARS), which ostensibly measures seven dimensions of time urgency. Menon et al.10 have recent factor analysed the BARS, finding two separate interpretable subscales, which they labelled ‘strategic time urgency,’ and ‘obsessive time urgency.’ Furthermore, they reported a differential pattern of correlations to physical/ medical symptoms implicating obsessional time urgency as the sole health risk. By far the most widely used scale in this area is the 21-item ‘Speed & Impatience’ (SI) subscale of the Jenkins Activity Survey (JAS) measure of the TAB. Hart5 has reviewed evidence to suggest the SI subscale from the JAS is uniquely predictive (relative to global A-B scores) of generalized health problems and health-risk processes. Clearly, work in this area seems to be growing at a healthy pace. But, a word of caution is in order, lest history repeat itself. Ironically, one of the potential pitfalls in the burgeoning area of Type A subcomponent research is the very rapid proliferation of these different measurement scales. Because research to adequately support the reliability or validity of the various scales (i.e. TUI, TUPA, II, Strategic & Obsessional Time Urgency, SI) has yet to be conducted, it cannot be assumed they are all measuring the same theoretical construct. Also, we still do not know if this ambiguously defined construct is sufficiently stable over time to make it a viable CHD risk factor. Thus, generalizability of findings across different measurement scales might be quite limited. Clearly, in order to extrapolate from the operational level of analysis to the theoretical and conceptual level (and thereby move the field ahead), we need to quickly develop a programme of studies to evaluate the construct validity and reliability of the available assessment instruments. Without independent evidence of construct validity, empirical associations to health indicators derived from epidemiological research remain sterile, telling us little about the nature of the putative causal agent. Also, in order for us to develop conceptual models of the mediating mechanisms that might explain how this poorly understood personality construct (‘time urgency & impatience’?) confers increased health risk, we need to have a much richer theoretical understanding of the attributes of the trait(s). Finally, efficient and effective therapeutic and preventative interventions would also depend on more precise theoretical specification of the nature of the causal agent. Thus, future studies that seek to extend and refine work by Cole2 and others might wish to address the question ‘Is there any validity to the Time Urgency & Irritability concept?’ Developmentally, the Time Urgency & Irritability concept is now beginning to metamorphose from infancy to early childhood. I sincerely hope it can avoid following in the topsy-turvy footsteps of its Type A predecessor, which left in its wake a rather dubious career trajectory. Something seems to be emerging from the literature, but what—precisely—is it? References 1 Dimsdale JE. A perspective on Type A behavior and coronary disease. N Engl J Med 1988;318:110–12. 2 Cole SR, Kawachi I, Liu S et al. Time urgency and risk of non-fatal myocardial infarction. Int J Epidemiol 2001;30:363–69. 3 Booth-Kewley S, Friedman HS. Psychosocial predictors of heart disease: a quantitative review. Psychol Bull 1987;101:342–62. 4 Miller TQ, Smith TW, Turner CW, Guijarro ML, Hallet AJ. Meta- analytic review of research on hostility and physical health. Psychol Bull 1996;119:322–48. 5 Hart KE. A moratorium on research using the Jenkins Activity Survey for Type A behaviour? J Clin Psychol 1997;53:905–07. 6 Karlberg L, Unden AL, Elofsson, S, Krakau I. Is there a connection between car accidents and Type A drivers? Behav Med 1998;24:99–106. 7 Wright L, Nielsen BA, Abbanato KR, Jackson T, Lancaster C, Son J. The relationship of various measures of time urgency to indices of physical health. J Clin Psychol 1995;51:610–14. 8 Spence JT, Helmreich RL, Pred RS. Impatience versus achievement strivings in the Type A pattern: differential effects on students’ health and academic achievement. J Appl Psychol 1987;72:522–28. 9 Landy FL, Rastegary H, Thayer J, Colvin, C. Time urgency: the con- struct and its measurement. J Appl Psychol 1991;76:644–57. 10 Menon S, Narayanan L, Spector PE. Time urgency and its relation to occupational stressors and health outcomes for health care professionals. In: Spielberger CR, Sarason CD (eds). Stress and Emotion: Anxiety, Anger and Curiosity. Vol. 16. San Francisco: Taylor and Francis, 1996, pp.127–42.
© Copyright 2026 Paperzz