Secondary Logo

Journal Logo

Cardiovascular Disease

Risk of Acute Myocardial Infarction

Dyslipidemia More Detrimental for Men than Women

Madssen, Erika,b; Laugsand, Lars Erikc; Wiseth, Runea,b; Mørkedal, Bjørnc; Platou, Carld; Vatten, Larsc; Janszky, Imrec,e

Author Information
doi: 10.1097/EDE.0b013e31829d2632


Female sex hormones may have direct atheroprotective effects on the vasculature,1–3 and estrogens may inhibit oxidation of low-density lipoprotein (LDL) cholesterol,4–7 which is pivotal for atherosclerotic plaque formation.8 Compared with men, women may therefore be protected against the effects of dyslipidemia in middle age, and having dyslipidemia and being male may have a synergistic effect on atherosclerosis. However, a differential effect by sex in relation to cardiovascular outcomes has not yet been empirically established.

In this large prospective population study, we tested whether dyslipidemia before 60 years of age has more detrimental effects on the risk of a first acute myocardial infarction (AMI) in men than in women.


Study Population

The adult population of Nord-Trøndelag County in Norway was invited to participate in the HUNT 2 Study (the second wave of the Nord-Trøndelag Health Study) between August 1995 and June 1997. Among 94,187 invited adults 20 years of age and older, 65,215 (69%) participated. The study has been described in detail elsewhere.9 Briefly, information was collected using self-administered questionnaires and a clinical examination that included measurements of blood pressure (BP), anthropometry, and blood sampling. The study was approved by the Regional Committee for Ethics in Medical Research, the National Directorate of Health, and the Norwegian Data Inspectorate.

Since any effects of female sex hormones on AMI risk presumably decrease a few years after menopause,10 we restricted our primary analyses to participants who were younger than 60 years of age at baseline (n = 45,077). In a secondary analysis, we also examined joint effects of dyslipidemia and male sex among participants who were 60 years or older (n = 20,138) at baseline.

Serum Lipids

A nonfasting blood sample was drawn from each participant, centrifuged within 2 hours after blood draw and placed in a refrigerator at 4°C. Samples were sent to the Central Laboratory at Levanger Hospital, Norway, on the same day and analyzed using a Hitachi 911 Autoanalyzer (Hitachi, Mito, Japan). The time interval between last meal and venipuncture was recorded.

Serum concentrations of total cholesterol and high-density lipoprotein (HDL) cholesterol were analyzed using reagents from Boehringer Mannheim (Mannheim, Germany). The day-to-day coefficients of variation were 1.3–1.9% and 2.4%, respectively. Total cholesterol and HDL cholesterol were measured by an enzymatic colorimetric cholesterol esterase method. Measurements of HDL cholesterol were performed after precipitation with phosphotungsten and magnesium ions. Non-HDL cholesterol was calculated as total serum cholesterol minus HDL cholesterol.

In the analysis, lipids were dichotomized at cut-points for dyslipidemia.11 Total cholesterol, HDL cholesterol, and non-HDL cholesterol were dichotomized at ≥6.2 mmol/L (≥240 mg/dL), ≤1.03 mmol/L (≤40 mg/dL), and ≥4.9 mmol/L (≥155 mg/dL) for both sexes, respectively. Our results remained essentially the same with alternative cut-points (eTable 1–3, see Electronic Appendix,

Potentially Confounding Factors

The clinical examination was conducted by trained nurses. BP was measured using a Dinamap 845XT (Critikon, Tampa, FL) sphygomomanometer based on oscillometry. We used the average value of the second and third measurement in the analysis. Height was measured to the nearest centimeter and weight to the nearest 0.5 kg, and body mass index (BMI) was calculated as weight divided by height squared. Waist circumference was measured to the nearest centimeter at the level of the umbilicus.

Information on health, lifestyle factors, and medication was collected by a self-administered questionnaire that included self-reported history of myocardial infarction, stroke, diabetes, and a family history of coronary heart disease. Participants were asked to assess their usual level of physical activity. Light physical activity was defined as activity that does not involve sweating or feeling of breathlessness, whereas hard physical activity was defined as activity with sweating or breathlessness. Participants were classified as (1) inactive if they reported less than 1 hour of hard activity or less than 3 hours of light activity per week, (2) moderately active if they reported 1 to 3 hours of hard activity or above 3 hours of light activity per week, and (3) physically active if they reported more than 3 hours of hard physical activity per week. Responses to questions related to smoking were categorized as current, previous, or never smoking. Education was categorized as low (less than 10 years), medium (10 to 12 years), or high (more than 12 years).


A total of 827 participants were excluded at baseline because they reported a previous myocardial infarction (n = 348), angina pectoris (n = 288), or stroke (n =191), leaving 44,250 people without known coronary artery disease or previous cerebrovascular disease at baseline. All participants were followed from the baseline examination until the occurrence of a first nonfatal or fatal AMI or end of follow-up at 31 December 2008, whichever occurred first. During a median follow-up of 11.8 years, 206 participants who moved out of Nord-Trøndelag County and 1068 participants who died from other causes were censored at the time of event in the statistical analysis. Hospitalizations for AMI were identified through a linkage with medical records from the two hospitals (Levanger and Namsos) in Nord-Trøndelag County. AMI was defined according to the World Health Organization criteria12 and the consensus guideline introduced in 2000 by the European Society of Cardiology and American College of Cardiology.13 Fatal AMI cases were identified by the National Cause of Death Registry according to the International Classification of Disease (ICD-9: 410; ICD-10: I21 and I22).

Statistical Analysis

We assessed to which degree dyslipidemia is more detrimental for men than women by calculating the proportion of AMI cases among men with dyslipidemia that may be attributed to a synergism between being male and having dyslipidemia according to the following formula14–16:

Attributable proportion due to synergism = (HRfor being male and having dyslipidemia – HRfor dyslipidemia – HRfor being male + 1) / HRfor being male and having dyslipidemia.

The hazard ratios (HRs) were derived from multivariable Cox proportional hazard models, where we adjusted for age (continuous), BMI (continuous), systolic and diastolic BP (continuous), diabetes mellitus (yes/no), smoking habits (never, former, current, unknown), physical activity (inactive, moderately active, physically active), and education (low, medium, high). For calculating the HRs, we compared participants with higher risks, that is, men and participants with dyslipidemia, to those with low risk, that is, women and participants without dyslipidemia. Proportionality was checked by comparing log-minus-log plots of survival and by performing tests based on Schoenfeld residuals. Missing observations were treated by list-wise deletion in the multivariable analyses.

The attributable proportion due to synergism is also referred to as the attributable proportion due to interaction. The numerator in the formula shows the relative excess risk due to synergism, that is, the risk that is additional to the risk that is expected on the basis of the independent effects of being male and having dyslipidemia. This is also called relative excess risk due to interaction or the interaction contrast ratio in the literature. These measures are recommended in assessments of whether an exposure is more detrimental in one versus another subgroup, in contrast to methods based on comparison of relative risks, that is, assessing interaction directly in Cox regression.14–17

In AMI cases with dyslipidemia, a certain proportion is attributable to dyslipidemia and another proportion is attributable to baseline risk, that is, to all other risk factors. Because men have a higher absolute risk than women, a certain proportion of AMI cases among men is also attributable to being male. If male sex and dyslipidemia have a synergistic effect on AMI risk, a proportion of cases among men with dyslipidemia will be attributable to this synergism alone. Theoretically, the synergism may also suggest that successful treatment of dyslipidemia could prevent more AMI cases among men. In contrast, a lack of synergism, that is, if dyslipidemia is equally detrimental for men and women, means that the attributable proportion due to synergism is zero.

We performed several sensitivity analyses to address the robustness of our findings. First, we restricted the analyses to AMI cases that were confirmed at hospitals (excluding cases based on death certificates alone). Furthermore, we restricted the analyses to AMI before the age of 60 years by censoring participants at the age of 60 years. In another set of analyses, we additionally controlled for family history of coronary heart disease and use of antihypertensive medication.

All statistical analyses were conducted using Stata 10.0 for Windows (Statacorp LP, TX). For calculating confidence intervals of attributable proportion due to synergism and relative excess risk due to synergism, we used the STATA codes and Excel sheet provided by Andersson et al.15


Table 1 presents baseline characteristics of study participants. Compared with men, women had lower BMI and a more favorable lipid profile but a higher prevalence of diabetes mellitus, current smoking, and physical inactivity. During 11.8 years of follow-up of participants who were younger than 60 years at baseline, 157 new cases of AMI were diagnosed among women and 525 among men. For those older than 60 years at baseline, 727 new cases of AMI occurred among women and 884 among men.

Baseline Characteristicsa for Men and Women Below 60 Years of Age

Tables 2 and 3 show that the HRs of dyslipidemia, as indicated by high total serum cholesterol, low HDL cholesterol, and high non-HDL cholesterol, were all in the expected direction, with slightly higher relative risks for high total serum cholesterol and high non-HDL cholesterol in men (Table 2), and a slightly higher relative risk for low HDL cholesterol in women (Table 3).

Hazard Ratios with 95% Confidence Intervals for AMI Among Men
Hazard Ratios with 95% Confidence Intervals for AMI Among Women

The results in Table 4 suggest a relatively strong synergism between dyslipidemia and being male in relation to AMI risk. The strongest synergism was for non-HDL cholesterol, followed by total serum cholesterol. For total serum cholesterol, the attributable proportion due to synergism was 0.46 (95% confidence interval = 0.35 to 0.57) and the relative excess risk due to synergism was 3.92 (2.43 to 5.42). This suggests that nearly half of the AMI cases among men with a total serum cholesterol ≥6.2 mmol/L (≥240 mg/dL) may be attributable to a synergism between being male and having high total cholesterol.

Synergism Between the Effects of Male Sex and Dyslipidemia

In contrast, hypertension (systolic ≥140 mmHg or diastolic BP ≥90 mmHg) and obesity (BMI ≥30 kg/m2) showed no synergism with male sex. The attributable proportions due to synergism between male sex and these factors were 0.02 (−0.21 to 0.25) and −0.01 (−0.27 to 0.24), respectively, and the relative excess risks due to synergism were 0.08 (−0.89 to 1.04) and −0.06 (−1.07 to 0.95), respectively.

In addition, we assessed possible synergisms between being male and having dyslipidemia among participants (8967 women and 6715 men) who were 60 years or older at baseline. There was no evidence in that age group for any synergism between being male and having dyslipidemia. The attributable proportions due to synergism were 0.03 (−0.13 to 0.20) for total cholesterol, 0.09 (−0.11 to 0.29) for HDL cholesterol, and −0.01 (−0.18 to 0.16) for non-HDL cholesterol. The corresponding relative excess risks due to synergism were 0.11 (−0.13 to 0.20), 0.24 (−0.29 to 0.77), −0.04 (−0.58 to 0.51), respectively.

We obtained essentially similar results when we restricted follow-up to AMI cases confirmed in hospitals (eTable 4, Electronic Appendix, and to AMI before 60 years of age. In these analyses, the attributable proportions due to synergism between male sex and total cholesterol, HDL cholesterol and non-HDL cholesterol were 0.44 (0.33 to 0.55), 0.20 (0.00 to 0.39), and 0.50 (0.40 to 0.61), respectively. The corresponding relative excess risks due to synergism were 3.64 (2.21 to 5.07), 1.18 (−0.06 to 2.42), and 4.45 (2.88 to 6.03), respectively. Adjustment for a family history of coronary heart disease and use of antihypertensive medication had virtually no effect on the results.


In this prospective follow-up study of a large population in Norway, we found a relatively strong synergistic effect of being male and having dyslipidemia in relation to the risk of AMI. Although dyslipidemia was associated with increased risk for AMI in both men and women, our results suggest that dyslipidemia in middle age has a more detrimental effect among men. In contrast, hypertension and obesity were equally detrimental for women as for men. Furthermore, among participants who were 60 years or older at baseline, we found no synergism between being male and having dyslipidemia in relation to AMI risk.

Current guidelines for the management of dyslipidemia do not distinguish between men and women in relation to primary prevention of AMI.11,18 In our study, approximately half of the AMI cases among middle-aged men with high total serum cholesterol or high non-HDL cholesterol may be attributable to the synergism between being male and having dyslipidemia. Thus, successful primary prevention or reduction of dyslipidemia in middle age is likely to prevent considerably more AMI cases among men than women with dyslipidemia.

Our findings may also advance our understanding of how sex differences influence the risk of AMI. The incidence of AMI is markedly higher in middle-aged men than in women, but the reasons for the disparity are not fully understood. One important cause is likely to be the higher burden of established risk factors among men in this age group, including a higher prevalence of dyslipidemia.19 Our results suggest not only that dyslipidemia is more common in men but also that dyslipidemia is considerably more detrimental for men than women in middle age.

One likely explanation for our findings is the protection that women receive from estrogens until a few years after menopause. Estrogen is a potent antioxidant that reduces oxidation of LDL cholesterol in the arterial wall,4–6 maybe preventing the formation of atheromatous plaques in middle-aged women. Estrogens may also protect against plaque rupture at advanced stages of atherosclerosis by inhibiting expression of matrix metalloproteinases induced by oxidized LDLs.7 Among participants older than 60 years, we found no excess risk of AMI due to synergism between being male and having dyslipidemia, suggesting that after menopause, endogenous estrogens may no longer provide a protection against cardiovascular consequences caused by dyslipidemia in women. However, women and men are different in many other aspects besides estrogen levels. Lifestyle, diet, social status, and participation on the labor market are sex dependent in most societies. However, it is not clear how these factors could explain our findings.

Previously, some prospective population studies have reported sex-specific relative risks for blood lipids in middle age and cardiovascular morbidity and mortality20–22 and shown that relative risks are comparable for men and women, similar to what we found in this study. However, comparisons of relative risks and assessments of interaction terms can be misleading if subgroups of participants have different baseline risks.14,23 Men, especially in middle age, have a much higher baseline AMI risk (ie, risk without dyslipidemia) than women. Although relative risks for dyslipidemia are similar in men and women, considerably more cases are caused by dyslipidemia among men than among women, and therefore, more cases are likely to be prevented by reducing dyslipidemia in men. For example, dyslipidemia might cause an additional five AMI cases in a population of dyslipidemic women, where without dyslipidemia there would only be five cases. If, in an equal number of dyslipidemic men, dyslipidemia were responsible for 20 cases in addition to the 20 cases that would occur in absence of dyslipidemia, the relative risks among women and men would be identical, but dyslipidemia would be responsible for four times as many cases among men than among women (ie, 20 vs. five cases).

To the best of our knowledge, only one previous study has used a similar approach to assess synergism between male sex and dyslipidemia in relation to AMI. The investigators found a moderate-to-strong synergism between being male and having dyslipidemia in relation to AMI risk.24 The synergism was somewhat less pronounced than in our study, the inclusion of older participants (up to 75 years) in that study may explain the difference.


For practical reasons, strict fasting could not be required for participation. Still, normal food intake appears to have a very small effect on blood levels of total cholesterol, HDL cholesterol, and non-HDL cholesterol.25 It is unlikely that a slight misclassification in lipid values due to nonfasting would differ by sex and thereby influence the estimated synergism.

Information on the use of lipid-modifying drugs was not available. However, for primary prevention purposes, the use of statins was rare in Norway in the 1990s and the early 2000s and we excluded those reporting previous cardiovascular disease. Moreover, it is unlikely that statins were more often prescribed to women, and therefore, this limitation might have negligible effects on our estimates.

As in any observational study, unevenly distributed study characteristics could have confounding effects that may not be accounted for in the analyses. Although we adjusted for several potential confounders, we cannot exclude the possibility of uncontrolled confounding from (for example) the use of hormone replacement therapy or other medications. However, in general, observational studies and randomized controlled trials have shown comparable results in relation to potentially harmful effects of dyslipidemia on cardiovascular outcomes. It should be noted that cardiovascular prevention trials of participants with dyslipidemia have not included enough women to draw any firm conclusion with regard to lipid-lowering effects among middle-aged women with dyslipidemia. In a recent meta-analysis of controlled primary prevention trials of statin use, where the majority of participants were older than 60 years of age, the relative effects did not differ between men and women. However, synergism was not assessed and the authors included trials that were not restricted to participants with dyslipidemia.26

Lipids levels were evaluated only once at the beginning of follow-up; thus, we could not examine the possible effects of time-dependent changes. However, changes in lipid levels could not possibly explain the observed synergism between dyslipidemia and male sex.

Finally, it should be noted that the method we used to assess the degree of synergism between dyslipidemia and male sex is based on the assumption that these factors do not have protective effects in some people.23 Although we cannot test whether this assumption is satisfied, a large violation is very unlikely.

In sum, our results suggest that dyslipidemia is more detrimental for men than women, and that the prevention of dyslipidemia in middle age has a greater potential to reduce the incidence of AMI among men than among women. Our findings could therefore be important for primary prevention strategies against coronary heart disease in middle age. These results may also advance our understanding of underlying mechanisms for the much higher occurrence of AMI among men in middle age.


1. Perez-Lopez FR, Larrad-Mur L, Kallen A, Chedraui P, Taylor HS. Gender differences in cardiovascular disease: hormonal and biochemical influences. Reprod Sci. 2010;17:511–531
2. Mercuro G, Deidda M, Piras A, Dessalvi CC, Maffei S, Rosano GM. Gender determinants of cardiovascular risk factors and diseases. J Cardiovasc Med (Hagerstown). 2010;11:207–220
3. Mendelsohn ME, Karas RH. The protective effects of estrogen on the cardiovascular system. N Engl J Med. 1999;340:1801–1811
4. Nègre-Salvayre A, Pieraggi MT, Mabile L, Salvayre R. Protective effect of 17 beta-estradiol against the cytotoxicity of minimally oxidized LDL to cultured bovine aortic endothelial cells. Atherosclerosis. 1993;99:207–217
5. Zhu X, Bonet B, Gillenwater H, Knopp RH. Opposing effects of estrogen and progestins on LDL oxidation and vascular wall cytotoxicity: implications for atherogenesis. Proc Soc Exp Biol Med. 1999;222:214–221
6. Arteaga E, Rojas A, Villaseca P, Bianchi M, Arteaga A, Duran D. In vitro effect of estradiol, progesterone, testosterone, and of combined estradiol/progestins on low density lipoprotein (LDL) oxidation in postmenopausal women. Menopause. 1998;5:16–23
7. Uzui H, Sinha SK, Rajavashisth TB. 17beta-estradiol inhibits oxidized low-density lipoprotein-induced increase in matrix metalloproteinase-9 expression in human macrophages. J Invest Med. 2011;59:1104–1108
8. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352:1685–1695
9. Holmen J, Midthjell K, Kruger Ø, et al. The Nord-Trøndelag Health Study 1995–97 (HUNT 2): objectives, contents, methods and participation. Norsk Epidemiologi. 2003;13:19–32
10. Burger H. The menopausal transition–endocrinology. J Sex Med. 2008;5:2266–2273
11. National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). . Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143–3421
12. Nomenclature and Criteria for Diagnosis of Ischemic Heart Disease. . Report of the Joint International Society and Federation of Cardiology/World Health Organization task force on standardization of clinical nomenclature. Circulation. 1979;59:607–609
13. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction redefined–a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol. 2000;36:959–969
14. Rothman K, Greenland S, Lash T Modern Epidemiology. 20083rd ed Philadelphia, PA Lippincott, Williams & Wilkins:71–83
15. Andersson T, Alfredsson L, Källberg H, Zdravkovic S, Ahlbom A. Calculating measures of biological interaction. Eur J Epidemiol. 2005;20:575–579
16. Knol MJ, VanderWeele TJ. Recommendations for presenting analyses of effect modification and interaction. Int J Epidemiol. 2012;41:514–520
17. Lundberg M, Fredlund P, Hallqvist J, Diderichsen F. A SAS program calculating three measures of interaction with confidence intervals. Epidemiology. 1996;7:655–656
18. Catapano AL, Reiner Z, De Backer G, et al. ESC/EAS guidelines for the management of dyslipidaemias: the task force for the management of dyslipidaemias of the European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS). Atherosclerosis. 2011;217(suppl 1):S1–S44
19. Lombardi M, Mercuro G, Fini M, Rosano GM. Gender-specific aspects of treatment of cardiovascular risk factors in primary and secondary prevention. Fundam Clin Pharmacol. 2010;24:699–705
20. Sharrett AR, Ballantyne CM, Coady SA, et al.Atherosclerosis Risk in Communities Study Group. Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 2001;104:1108–1113
21. Ingelsson E, Schaefer EJ, Contois JH, et al. Clinical utility of different lipid measures for prediction of coronary heart disease in men and women. JAMA. 2007;298:776–785
22. Walldius G, Jungner I, Holme I, Aastveit AH, Kolar W, Steiner E. High apolipoprotein B, low apolipoprotein A-I, and improvement in the prediction of fatal myocardial infarction (AMORIS Study): a prospective study. Lancet. 2001;358:2026–2033
23. VanderWeele TJ, Robins JM. The identification of synergism in the sufficient-component-cause framework. Epidemiology. 2007;18:329–339
24. Reuterwall C, Hallqvist J, Ahlbom A, et al. Higher relative, but lower absolute risks of myocardial infarction in women than in men: analysis of some major risk factors in the SHEEP Study. The SHEEP Study Group. J Intern Med. 1999;246:161–174
25. Langsted A, Freiberg JJ, Nordestgaard BG. Fasting and nonfasting lipid levels: influence of normal food intake on lipids, lipoproteins, apolipoproteins, and cardiovascular risk prediction. Circulation. 2008;118:2047–2056
26. Kostis WJ, Cheng JQ, Dobrzynski JM, Cabrera J, Kostis JB. Meta-analysis of statin effects in women versus men. J Am Coll Cardiol. 2012;59:572–582

Supplemental Digital Content

© 2013 by Lippincott Williams & Wilkins, Inc