- Research article
- Open Access
- Open Peer Review
This article has Open Peer Review reports available.
T wave abnormalities, high body mass index, current smoking and high lipoprotein (a) levels predict the development of major abnormal Q/QS patterns 20 years later. A population-based study
© Moller et al; licensee BioMed Central Ltd. 2006
Received: 23 October 2005
Accepted: 06 March 2006
Published: 06 March 2006
Most studies on risk factors for development of coronary heart disease (CHD) have been based on the clinical outcome of CHD. Our aim was to identify factors that could predict the development of ECG markers of CHD, such as abnormal Q/QS patterns, ST segment depression and T wave abnormalities, in 70-year-old men, irrespective of clinical outcome.
Predictors for development of different ECG abnormalities were identified in a population-based study using stepwise logistic regression. Anthropometrical and metabolic factors, ECG abnormalities and vital signs from a health survey of men at age 50 were related to ECG abnormalities identified in the same cohort 20 years later.
At the age of 70, 9% had developed a major abnormal Q/QS pattern, but 63% of these subjects had not been previously hospitalized due to MI, while 57% with symptomatic MI between age 50 and 70 had no major Q/QS pattern at age 70. T wave abnormalities (Odds ratio 3.11, 95% CI 1.18–8.17), high lipoprotein (a) levels, high body mass index (BMI) and smoking were identified as significant independent predictors for the development of abnormal major Q/QS patterns. T wave abnormalities and high fasting glucose levels were significant independent predictors for the development of ST segment depression without abnormal Q/QS pattern.
T wave abnormalities on resting ECG should be given special attention and correlated with clinical information. Risk factors for major Q/QS patterns need not be the same as traditional risk factors for clinically recognized CHD. High lipoprotein (a) levels may be a stronger risk factor for silent myocardial infarction (MI) compared to clinically recognized MI.
Several longitudinal studies have investigated the impact of different ECG abnormalities on future development of coronary heart disease (CHD) with morbidity, based on episodes of chest pain or mortality as end-points , excluding those with silent myocardial infarction (MI). The prevalence and prognosis of silent MI is still controversial [2, 3] and it remains uncertain whether patients with silent MI have unique characteristics that distinguish them from those with recognized events . In the present population-based study with 20 years of follow-up, our primary objective was to identify predictors for the development of abnormal major Q/QS patterns as a marker for MI, irrespective of chest pain, in order to obtain a complementary picture to previous studies. Our secondary objective was to investigate predictors for ST segment depression and T wave abnormalities. Among the possible predictors were anthropometrical and metabolic factors, as well as ECG items and vital signs.
The Uppsala Study of Adult Men (ULSAM) is a longitudinal study based on all available 50-year-old men in Uppsala, Sweden, born between 1920 and 1924. They were invited to a health survey aimed at identifying risk factors for cardiovascular disease . Of the invited subjects, 2,322 (82%) participated in 1970–1973. In 1991–1995, eligible participants were invited for re-examination at age 70. Of these, 1,221 (73%) participated. Between the 2 surveys, 219 had moved and 422 had died 128 due to acute MI. All subjects gave informed consent, and the Ethics Committee of Uppsala University approved the study. Official hospital discharge register data was available.
In the present study, all subjects hospitalized due to acute MI before the 50-year survey (n = 8) or with major Q wave (n = 6) or left bundle branch block (LBBB) (n = 5) on ECG at age 50, were excluded from further analyses, resulting in a study population of 2303 men at age 50 of which 1131 men were re-investigated at age 70. Development of ST segment depression and isolated T wave abnormality on ECG at age 70 were investigated in subgroups of 1006 men (additional exclusion of men with Minnesota codes 4.1–4.2 at age 50) and 858 men (additional exclusion of men with Minnesota codes 4.1–4.2 at age 50 and/or 70, and 5.1–5.4 at age 50), respectively.
Investigations and laboratory methods
Investigations at age 50
Minnesota classification of ECGs
Q or QS pattern
Q duration ≥ 0.04s in I, II, V2–V6.
Q duration ≥ 0.05s in both aVF and III.
QS pattern through V1–V4, V5 and V6.
QS pattern when R wave is present in adjacent precordial lead to the right V2–V6.
ST segment depression
4.1: ST-depression ≥ 1 mm in I, II, aVL, aVF, V1–V6.
4.2: ST depression 0.5–0.9 mm in I, II, aVL, aVF, V1–V6.
T wave items
5.1: T amplitude ≥ 5 mm in I, II V2–V6 when R amplitude ≥ 5 mm in aVL when QRS mainly upright in aVF.
5.2: T amplitude = -1 to -5 in I, II, V2–V6 when R amplitude ≥ 5 mm in aVL when QRS mainly upright in aVF.
5.3: T wave flat or small diphasic in I, II, V2–V6 when R amplitude ≥ in aVL when QRS mainly upright in aVF.
5.4: T amplitude positive and T/R amplitude radio <1/20 in any of leads I, aVL, V6: R wave amplitude must be ≥ 10.0 mm.
Investigations at age 70
Our primary outcome, a finding of a major abnormal Q/QS pattern was defined according to the Minnesota code 1.1 [6, 7]. Minor abnormal Q/QS patterns that only fulfilled the requirements of Minnesota code 1.2 or 1.3 were not included as end-points reducing the possibility of inclusion of abnormal Q/QS patterns of other aetiologies than CHD. Our secondary end-points were ST segment depression (Minnesota 4.1–4.2) and T wave abnormalities (Minnesota 5.1–5.4).
Subjects with ECG abnormalities observed during the survey were not further assessed by a cardiologist or evaluated further by other cardiac examinations as part of the protocol.
The distribution of a continuous variable was tested for normality using Shapiro Wilk's test. Skewed variables were logarithmically transformed to reach normal distribution. Univariate regression analyes were performed on the list of candidate predictor variables at age 50, comparing those with and without ECG abnormalities at age 70 using χ2 test or ANOVA. A p < 0.05 was considered statistically significant. All candidate predictors with a univariate p < 0.15 were evaluated for independent association in a backward stepwise logistic regression model. Continuous variables were standardized to one standard deviation (SD) to determine the magnitude of the relationship to, and the statistical significance of the predictors of each of the defined outcomes in the regression analyses. The robustness of the final model was checked by including previously removed variables. As an additional analysis, the same variables were tested in the same model but substituting the end point major Q/QS pattern by death due to acute MI before the 70-year examination. The prevalence of T wave abnormalities at age 50 in the subjects who died during the 20-year follow-up period was also compared to those who survived using χ2 test.
Prevalence of ECG abnormalities at age 50 and 70
ECG abnormalities *
Age 50 (n = 2314) n (%)
Age 70 (n = 1135) n (%)
Major Q or QS pattern (1.1)
Minor Q or QS pattern (1.2/1.3)
ST seg. depression (4.1/4.2)
T wave items (5.1–5.4)
High R-amplitude (3.1–3.3)
Atrial fibrillation/flutter (8.3)
Major abnormal Q/QS pattern
Clinical characteristics at age 50 according to Q-QS pattern at age 70 (mean ± SD)
Variables, age 50
Abnormal Q-QS pattern (1.1) at age 70
No abnormal Q-QS (1.1) pattern at age 70
SBP supine (mmHg)
DBP supine (mmHg)
Fasting B-gluc (mmol/l)
Fasting insulin (μU/ml)
Serum TG (mmol/l)
Serum CHOL (mmol/l)
HDL CHOL (mmol/l)
LDL CHOL (mmol/l)
Lp (a) (U/l)
ECG findings at age 50 according to Q-QS-pattern at age 70
ECG abnormalities* at age 50
Abnormal Q-QS pattern (1.1) at age 70
(n = 102) n (%)
No abnormal Q-QS pattern (1.1) at age 70
(n = 1033) n (%)
Minor Q-QS pattern (1.2/1.3)
High amplitude R-wave (3.1/3.3)
ST-segment depression (4.1/4.2)
T-wave items (5.1–5.4)
Multivariate logistic regression analysis for development of abnormal Q-QS pattern (Minnesota 1.1) on ECG at age 70
Odds Ratio (95% - CI)
BMI (1SD = 3.3 kg/m2)
ST segment depression
Of the 1006 men free of ST segment depression at age 50, 13% had developed ST segment depression at the age of 70. In multiple logistic regression analysis T wave abnormalities (OR = 3.94, 95% CI 1.58–9.82, p = 0.003) and high fasting blood glucose levels (OR = 1.25, 95% CI 1.02–1.54, p = 0.035) were independent predictor for ST segment depression.
Isolated T wave abnormality
Of the 858 men without ST segment depression (at age 50 and/or 70) and T wave abnormality at age 50, 9% had developed an isolated T wave abnormality at the age of 70. An increase in BMI by 1SD increased the risk of having a T wave abnormality at the age of 70 by 50% (OR = 1.48, 95% CI 1.14–1.92). None of the other variables tested predicted future development of a T wave abnormality. Subjects who died due to MI during the observation period also had a significantly higher prevalence of T-wave abnormalities at age 50 compared to those who survived (15.3% vs 4.8%, p < 0.01).
Death due acute MI
Multivariate logistic regression analysis for death due to myocardial infarction before 70-year examination
Odds Ratio (95% - CI)
BMI (1SD = 3.3 kg/m2)
DBP supine (mmHg)
Besides the presence of T wave abnormalities, we identified high BMI, high Lp (a) and smoking at age 50 as independent risk factors for the development of major abnormal Q/QS patterns at age 70. T wave abnormalities, together with high fasting blood glucose were also important predictors for the development of ST segment depression while participants with higher BMI had an increased risk of developing T wave abnormalities.
This is a longitudinal community based study, with long follow-up period, where all subjects from one birth cohort in the community were invited to participate. The same two nurses registered the ECGs at both surveys, contributing to consistency in electrode placement. Three physicians carried out the coding of all ECGs, whereby the coding consistency can be expected to be high. However, coding errors may confound analysis of ECG data. An obvious limitation of our study is that women were not included. There was no validation of the outcome detection system, however, the myocardial infarction diagnosis in the Swedish hospital register has been found to be correct in 93–97% of cases with less than one percent being missed .
Our study is unique in that it uses ECG variables as outcomes instead of clinical outcomes. By choosing major abnormal Q/QS patterns as end-point, silent myocardial infarctions have also been included in the analyses. According to most studies a clinically unrecognized MI is believed to be associated with the same risk as recognized symptomatic MI [1, 4, 9] although a lower risk for unrecognized MI has also been described . The prevalence of silent MI increases with age [1, 2] and varies from 21% to 68% , with an average around 30% . In our study, 63% of the subjects with a major abnormal Q/QS pattern on ECG at the 70-year survey had not been previously hospitalized due to MI. While the majority of major Q waves are believed to be due to myocardial infarction, some may be due to other causes, including myocarditis, cardiac amyloidosis, and cardiomyopathy . Also displacement of the diaphragm and the heart, as in chronic obstructive lung disease may cause the appearance of a non-infarction Q wave and non-infarction Q waves may be seen in subjects with pre-excitation syndrome and left ventricular hypertrophy (LVH) . The low specificity of Q/QS pattern to recognize silent myocardial necrosis is an additional limitation to our study.
Since some of the conventional risk factors  were not strong predictors of the development of major abnormal Q/QS patterns, our study suggests that predictors of major abnormal Q/QS patterns may be different from those predicting symptomatic MI. This is further strengthened by our additional analysis where the major Q/QS pattern end-point was substituted by death due to acute MI before the 70-year examination and where conventional risk factors such as diabetes mellitus and DBP became significant predictors. Other authors have debated whether the predisposing factors for silent MI are simply the traditional risk factors for coronary atherosclerosis, or whether silent MI patients have unique characteristics . However, no general conclusions can be drawn from our study since our major abnormal Q/QS finding will include silent MI, clinically recognized MI and some non-infarction Q waves.
Loss of Q wave could explain some of the cases where diagnosis of MI was not accompanied by a visible major Q/QS pattern at age 70. Even though development of Q waves after acute MI has been considered an indicator of myocardial necrosis, studies have shown that Q wave loss may occur in 11–20% of patients after transmural MI [14, 15]. The clinical significance of such an ECG phenomenon has not been fully clarified . Functional recovery of stunned and/or hibernating myocardium , as well as regeneration of myocytes  has been proposed as possible mechanisms.
The pathophysiologic basis of ST segment depression on resting ECG remains speculative, and ST segment depression is seen both in left ventricular hypertrophy and myocardial ischemia . One of the possible etiologies for development of left ventricular hypertrophy is altered glucose metabolism with insulin resistance and hyperinsulinemia [18, 19]. On the other hand, patients with diabetes mellitus are known to have an increased risk of developing arteriosclerosis , so an impaired glucose metabolism could well be a predictor for ST segment depression due to either left ventricular hypertrophy or coronary heart disease.
While permanent negative T waves in the chronic stage of MI have indicated the presence of transmural infarction . T waves in a post-myocardial phase have been shown to be associated with the presence of viable myocardium at jeopardy . T waves should therefore be regarded as a dynamic substrate. Even though subjects not surviving the MI were not included in our main analyses, we found that the prevalence of major T wave abnormalities in those who died during the period due to acute MI was statistically significantly higher for this group compared to those who survived, suggesting that the predictive power of T wave abnormalities is not restricted to survivors of a Q wave MI.
High BMI predicted both major abnormal Q/QS pattern and abnormal T wave. One possible mechanism is horizontal displacement of the heart , but left ventricular hypertrophy may also play a role.
Lp (a) is believed to have potential pathogenic effects on atherosclerosis and thrombosis [24–26]. Some previous studies have identified elevated Lp (a) as a risk factor leading to premature MI [27, 28], while others have shown no association between Lp (a) and risk for CHD . One study has, in diabetic patients, specifically found Lp(a) to be associated with silent CHD . Our study supports the finding that Lp (a) may be important an important predictor of CHD and silent MI.
The present approach using ECG-based diagnosis should be regarded as a complement to the more traditional approach used in most previous studies. One should therefore be careful not to generalize the conclusion of this study to other population groups, and further studies for confirmation of our findings are motivated.
T wave abnormalities on ECG at age 50, was an independent risk factor for the development of major Q/QS abnormalities and ST segment depression on ECG at age 70. A finding of a T wave abnormality on ECG should be given special attention even though an isolated T wave change should be interpreted with caution and correlated with all available clinical and laboratory information. Risk factors for major abnormal Q/QS patterns may not be the same as risk factors for clinically recognized CHD and high Lp (a) levels may be a stronger risk factor for silent myocardial infarction (MI) compared to clinically recognized MI. Re-measurement of ECG after MI may give information on viability of myocardium through Q wave regression, which may be more frequent than previously believed.
This work was funded by the Medical Faculty at Uppsala University, Royal Society of Science, Swedish Council for Planning and Co-ordination of Research, Swedish National Association against Heart and Lung Disease, Swedish Medical Research Council, and the Trygg Hansa Research Fund.
None of the funders have had any role in the study design; in the collection, analysis, and interpretation of data; in the writing of the manuscript; nor in the decision to submit the manuscript for publication.
- Ashley EA, Raxwal VK, Froelicher VF: The prevalence and prognostic significance of electrocardiographic abnormalities. Curr Probl Cardiol. 2000, 25: 1-72. 10.1016/S0146-2806(00)70020-X.View ArticlePubMedGoogle Scholar
- Aronow WS: Silent MI. Prevalence and prognosis in older patients diagnosed by routine electrocardiograms. Geriatrics. 2003, 58: 24-6. 36–8, 40PubMedGoogle Scholar
- Cohn PF, Fox KM, Daly C: Silent myocardial ischemia. Circulation. 2003, 108: 1263-77. 10.1161/01.CIR.0000088001.59265.EE.View ArticlePubMedGoogle Scholar
- Sheifer SE, Gersh BJ, Yanez ND, Ades PA, Burke GL, Manolio TA: Prevalence, predisposing factors, and prognosis of clinically unrecognized myocardial infarction in the elderly. J Am Coll Cardiol. 2000, 35: 119-26. 10.1016/S0735-1097(99)00524-0.View ArticlePubMedGoogle Scholar
- Hedstrand H: A study of middle-aged men with particular reference to risk factors for cardiovascular disease. Ups J Med Sci Suppl. 1975, 19: 1-61.PubMedGoogle Scholar
- Blackburn H, Keys A, Simonson E, Rautaharju P, Punsar S: The electrocardiogram in population studies. A classification system. Circulation. 1960, 21: 1160-75.View ArticlePubMedGoogle Scholar
- Prineas R, Crow R, Blackburn H: The Minnesota Code Manual of Electrocardiographic Findings, Standards and Procedures for Measuring and Classification. 1982Google Scholar
- Lindblad U, Rastam L, Ranstam J, Peterson M: Validity of register data on acute myocardial infarction and acute stroke: the Skaraborg Hypertension Project. Scand J Soc Med. 1993, 21: 3-9.PubMedGoogle Scholar
- Kannel WB, Abbott RD: Incidence and prognosis of unrecognized myocardial infarction. An update on the Framingham study. N Engl J Med. 1984, 311: 1144-7.View ArticlePubMedGoogle Scholar
- Medalie JH, Kahn HA, Neufeld HN, Riss E, Goldbourt U, Perlstein T, Oron D: Myocardial infarction over a five-year period. I. Prevalence, incidence and mortality experience. J Chronic Dis. 1973, 26: 63-84. 10.1016/0021-9681(73)90006-4.View ArticlePubMedGoogle Scholar
- Nadelmann J, Frishman WH, Ooi WL, Tepper D, Greenberg S, Guzik H, Lazar EJ, Heiman M, Aronson M: Prevalence, incidence and prognosis of recognized and unrecognized myocardial infarction in persons aged 75 years or older: The Bronx Aging Study. Am J Cardiol. 1990, 66: 533-7. 10.1016/0002-9149(90)90477-I.View ArticlePubMedGoogle Scholar
- Fisch C: Heart Disease: A textbook of cardiovascular medicine, Fifth Edition. Chapter 4. 1997Google Scholar
- Wilson PW: Established risk factors and coronary artery disease: the Framingham Study. Am J Hypertens. 1994, 7: 7S-12S.PubMedGoogle Scholar
- Coll S, Betriu A, de Flores T, Roig E, Sanz G, Mont L, Magrina J, Serra A, Navarro Lopez F: Significance of Q-wave regression after transmural acute myocardial infarction. Am J Cardiol. 1988, 61: 739-42. 10.1016/0002-9149(88)91058-2.View ArticlePubMedGoogle Scholar
- Lancellotti P, Gerard PL, Kulbertus HE, Pierard LA: Persistent negative T waves in the infarct-related leads as an independent predictor of poor long-term prognosis after acute myocardial infarction. Am J Cardiol. 2002, 90: 833-7. 10.1016/S0002-9149(02)02703-0.View ArticlePubMedGoogle Scholar
- Voon WC, Chen YW, Hsu CC, Lai WT, Sheu SH: Q-wave regression after acute myocardial infarction assessed by Tl-201 myocardial perfusion SPECT. J Nucl Cardiol. 2004, 11: 165-70. 10.1016/j.nuclcard.2003.10.009.View ArticlePubMedGoogle Scholar
- Beltrami AP, Urbanek K, Kajstura J, Yan SM, Finato N, Bussani R, Nadal-Ginard B, Silvestri F, Leri A, Beltrami CA, Anversa P: Evidence that human cardiac myocytes divide after myocardial infarction. N Engl J Med. 2001, 344: 1750-7. 10.1056/NEJM200106073442303.View ArticlePubMedGoogle Scholar
- Sundstrom J, Lind L, Nystrom N, Zethelius B, Andren B, Hales CN, Lithell HO: Left ventricular concentric remodeling rather than left ventricular hypertrophy is related to the insulin resistance syndrome in elderly men. Circulation. 2000, 101: 2595-600.View ArticlePubMedGoogle Scholar
- Felicio JS, Ferreira SR, Plavnik FL, Moises V, Kohlmann O, Ribeiro AB, Zanella MT: Effect of blood glucose on left ventricular mass in patients with hypertension and type 2 diabetes mellitus. Am J Hypertens. 2000, 13: 1149-54. 10.1016/S0895-7061(00)01200-0.View ArticlePubMedGoogle Scholar
- Koschinsky ML, Marcovina SM: The relationship between lipoprotein(a) and the complications of diabetes mellitus. Acta Diabetol. 2003, 40: 65-76.PubMedGoogle Scholar
- Maeda S, Imai T, Kuboki K, Chida K, Watanabe C, Ohkawa S: Pathologic implications of restored positive T waves and persistent negative T waves after Q wave myocardial infarction. J Am Coll Cardiol. 1996, 28: 1514-8. 10.1016/S0735-1097(96)00338-5.View ArticlePubMedGoogle Scholar
- Pierard LA, Lancellotti P: Determinants of persistent negative T waves and early versus late T wave normalisation after acute myocardial infarction. Heart. 2005, 91: 1008-12. 10.1136/hrt.2004.033936.View ArticlePubMedPubMed CentralGoogle Scholar
- Alpert MA, Terry BE, Cohen MV, Fan TM, Painter JA, Massey CV: The electrocardiogram in morbid obesity. Am J Cardiol. 2000, 85: 908-10. 10.1016/S0002-9149(99)00894-2. A10View ArticlePubMedGoogle Scholar
- Marcovina SM, Koschinsky ML: Lipoprotein(a) as a risk factor for coronary artery disease. Am J Cardiol. 1998, 82: 57U-66U. 10.1016/S0002-9149(98)00954-0. discussion 86UView ArticlePubMedGoogle Scholar
- Scott J: Lipoprotein(a). Thrombogenesis linked to atherogenesis at last?. Nature. 1989, 341: 22-3. 10.1038/341022a0.View ArticlePubMedGoogle Scholar
- Scanu AM: Lipoprotein(a) and the Atherothrombotic Process: Mechanistic Insights and Clinical Implications. Curr Atheroscler Rep. 2003, 5: 106-13.View ArticlePubMedGoogle Scholar
- Seed M, Ayres KL, Humphries SE, Miller GJ: Lipoprotein (a) as a predictor of myocardial infarction in middle-aged men. Am J Med. 2001, 110: 22-7. 10.1016/S0002-9343(00)00652-5.View ArticlePubMedGoogle Scholar
- Danesh J, Collins R, Peto R: Lipoprotein(a) and coronary heart disease. Meta-analysis of prospective studies. Circulation. 2000, 102: 1082-5.View ArticlePubMedGoogle Scholar
- Gazzaruso C, Garzaniti A, Giordanetti S, Falcone C, Fratino P: Silent coronary artery disease in type 2 diabetes mellitus: the role of Lipoprotein(a), homocysteine and apo(a) polymorphism. Cardiovasc Diabetol. 2002, 1: 5-10.1186/1475-2840-1-5.View ArticlePubMedPubMed CentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2261/6/10/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.