Author + information
- Susanna C. Larsson, PhD⁎ (, )
- Jarmo Virtamo, MD and
- Alicja Wolk, DMSc
- ↵⁎Division of Nutritional Epidemiology, National Institute of Environmental Medicine, Karolinska Institutet, Box 210, Stockholm SE-17177, Sweden
To the Editor:
Ample evidence indicates that chocolate may have beneficial effects on the cardiovascular system. Chocolate consumption has been shown to reduce systolic and diastolic blood pressure in short-term randomized feeding trials (1), and has been demonstrated to improve endothelial and platelet function and to ameliorate insulin resistance (2). Moreover, flavonoids in chocolate possess strong antioxidant activity and can suppress oxidation of low-density lipoprotein cholesterol (3).
We examined the association between chocolate consumption and risk of stroke in the population-based Swedish Mammography Cohort. In the autumn of 1997, 39,227 women completed a questionnaire that included approximately 350 items concerning diet and other lifestyle factors (4). We excluded women with a missing national identification number, those with implausible values for total energy intake, and those with a history of cancer, stroke, coronary heart disease, or diabetes mellitus before baseline. That left 33,372 women, age 49 to 83 years, for analysis. The study was approved by the Ethical Review Board at the Karolinska Institutet (Stockholm, Sweden). Chocolate consumption was assessed using a self-administered food-frequency questionnaire. Women were asked to indicate how often on average they had consumed chocolate and 95 other foods during the previous year. There were 8 pre-defined consumption categories ranging from never to ≥3 times a day. In the 1990s, approximately 90% of chocolate consumption in Sweden was milk chocolate, containing approximately 30% cocoa solids (5). The food-frequency questionnaire has been validated, and the Spearman correlation coefficient between chocolate intake estimated from the questionnaire and the intake estimated from 28 days of weighted diet records (4 1-week records completed approximately 3 months apart) was 0.4 (A. Wolk, unpublished data, 1992).
Incident cases of first stroke that occurred between January 1, 1998 (information on exact date of returning of the questionnaire was not available), and December 31, 2008, were ascertained by linkage with the Swedish Hospital Discharge Registry. The stroke events were classified as cerebral infarction (International Classification of Diseases-10th revision [ICD-10] code I63), intracerebral hemorrhage (I61), subarachnoid hemorrhage (I60), and unspecified stroke (I64). Information on dates of death was obtained from the Swedish Cause of Death Registry. The Swedish Hospital Discharge Registry provided information on diagnoses of atrial fibrillation (ICD-10 code I48). Cox proportional hazards models with age as the time scale was used to estimate relative risk (RR) and 95% confidence interval (CI) of stroke by exact quartiles of chocolate consumption based on the distribution in the cohort. Entry time was defined as a subject's age in months at start of follow-up, and exit time was defined as a subject's age in months at stroke diagnosis, death, or end of follow-up. The analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, North Carolina). All p values were 2-sided.
During a mean follow-up of 10.4 years, we ascertained 1,549 events of stroke, including 1,200 cerebral infarctions, 224 hemorrhagic strokes, and 125 unspecified strokes. Chocolate consumption was inversely associated with risk of total stroke, cerebral infarction, and hemorrhagic stroke (Table 1). The multivariable RRs for a 50 g/week increase of chocolate consumption were 0.86 (95% CI: 0.77 to 0.96) for total stroke, 0.88 (95% CI: 0.77 to 0.99) for cerebral infarction, and 0.73 (95% CI: 0.54 to 0.99) for hemorrhagic stroke. The difference in risk estimates for cerebral infarction and hemorrhagic stroke was not significant (p = 0.28). The association for total stroke persisted after excluding the first year of follow-up (50 g/week increase in intake RR: 0.86; 95% CI: 0.76 to 0.96). In stratified analysis by hypertension, the multivariable RRs of total stroke for 50 g per week increment of chocolate consumption were 0.89 (95% CI: 0.73 to 1.09) for women with a history of hypertension and 0.85 (95% CI: 0.74 to 0.97) for women without hypertension.
Our findings are broadly consistent with those from previous smaller studies, which observed either a statistically significant (136 stroke cases) (6) or a nonsignificant (111 or 469 stroke cases) (5,7) inverse association between chocolate consumption and total stroke. In the present study, only women in the highest quartile of chocolate consumption (median 66.5 g/week) had a significantly reduced risk of stroke, suggesting that higher intakes are necessary for a potential protective effect. The reason for the stronger association observed for hemorrhagic stroke than for cerebral infarction is unclear.
Strengths of this study include its prospective and population-based design, the nearly complete follow-up of participants through linkage with population-based Swedish registers, and the large number of stroke cases. A limitation is that chocolate consumption was self-reported, which will inevitably lead to some measurement error. Because of the prospective design, any misclassification of chocolate consumption is most likely to be nondifferential, leading to an attenuation of the true association between chocolate consumption and stroke. Although we adjusted for major risk factors for stroke, we cannot exclude the possibility that our results may have been affected by residual or unmeasured confounding. However, the similar results in the age-adjusted and multivariable models argue against the possibility of strong residual confounding.
In summary, results from this cohort of women suggest that a high chocolate consumption is associated with a lower risk of stroke.
Please note: This study was supported by research grants from the Swedish Council for Working Life and Social Research (FAS), by the Swedish Research Council/Committee for Infrastructure, and by a Research Fellow grant from Karolinska Institutet (to Dr. Larsson). The funding sources had no role in the design and conduct of the study, analysis or interpretation of the data, preparation or final approval of the manuscript, or the decision to submit the manuscript for publication. The authors have reported they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Desch S.,
- Schmidt J.,
- Kobler D.,
- et al.
- Corti R.,
- Flammer A.J.,
- Hollenberg N.K.,
- Luscher T.F.
- Engler M.B.,
- Engler M.M.
- Larsson S.C.,
- Virtamo J.,
- Wolk A.
- Mink P.J.,
- Scrafford C.G.,
- Barraj L.M.,
- et al.