Am Fam Physician. 2008;78(8):2-3
Background: Anxiety and other psychological characteristics have been linked to coronary artery disease (CAD). However, because most studies have evaluated combinations of conditions, the independent role of anxiety has been relatively under-researched. Shen and colleagues studied the association between documented anxiety and myocardial infarction (MI) in a large cohort of older men.
The Study: The authors evaluated data from the Normative Aging Study, which included men (mean age 60 years) with no history or evidence of heart disease or diabetes at baseline. On entry to the study, participants completed the Minnesota Multiphasic Personality Inventory (MMPI); demographic, medical, and laboratory assessments were also performed. After baseline, the 735 active participants received medical assessments every three years. These assessments included health and health behavior inventories, physical examinations, laboratory assessments, anthropometric measurements, and comprehensive psychological assessments (e.g., a modified MMPI). The anxiety level of each participant was assessed using four anxiety variables from the MMPI (psychasthenia, social introversion, phobia, and manifest anxiety) plus an overall measure of clinically significant anxiety. Hospital records and death certificates were used to identify MIs. The relative risk of MI was calculated for each anxiety variable and for overall anxiety.
Results: Most participants were white (96 percent), the mean body mass index (BMI) was 26.63 kg per m2, and the average systolic blood pressure was 128.7 mm Hg. Smoking was reported by 43 percent of participants, and 33 percent reported drinking more than two alcoholic drinks daily.
By 2004, 75 new MI events had been documented. In univariate analysis, each of the anxiety measures predicted MI with significant relative risks (RRs) of 1.31 to 1.39. After adjusting for age, education level, marital status, fasting blood glucose level, BMI, systolic blood pressure, and cholesterol levels, the relative risk numbers remained significant (1.31 to 1.43). The predictive relationship between anxiety measures and MI persisted after further adjustment for health behaviors, such as smoking, alcohol use, and calorie intake (RR = 1.33 to 1.43). RR data are presented in the accompanying table.
| Anxiety variable | Univariate RR (95% CI); n = 735 | Multivariate RR* (95% CI); n = 735 | Multivariate RR†(95% CI); n = 638 |
|---|---|---|---|
| Psychasthenia | 1.33 (1.10 to 1.60); P= .01 | 1.37 (1.12 to 1.68); P= .01 | 1.37 (1.11 to 1.70); P= .01 |
| Social introversion | 1.31 (1.06 to 1.63); P= .05 | 1.31 (1.05 to 1.63); P= .05 | 1.33 (1.06 to 1.67); P= .05 |
| Phobia | 1.38 (1.12 to 1.70); P= .01 | 1.36 (1.10 to 1.68); P = .01 | 1.33 (1.06 to 1.67); P= .05 |
| Manifest anxiety | 1.34 (1.10 to 1.62); P= .01 | 1.42 (1.16 to 1.73); P= .001 | 1.42 (1.14 to 1.76); P= .001 |
| Overall anxiety | 1.39 (1.15 to 1.68); P= .001 | 1.43 (1.17 to 1.75); P= .001 | 1.43 (1.15 to 1.77); P= .001 |
When the participants were divided into four groups based on the level of anxiety, there was a clear correlation between increased anxiety levels and increased MI events. Only 11 events occurred in men with anxiety scores in the lowest quartile, compared with 29 events in men scoring in the highest quartile (P < .05).
Conclusion: The authors conclude that measures of anxiety independently predict likelihood of MI in healthy older men during an average follow-up period of 12.4 years. In addition, men with the highest levels of anxiety are at significantly increased risk, suggesting a dose-response effect. They postulate that severe anxiety may be a clinical risk factor for MI, although the reason behind this correlation is not clear. The authors call for further research that includes a more diverse population to determine if the association between anxiety and MI risk is more universal.