Fatty Acids and Development of Atrial Fibrillation
Fatty Acids and Development of Atrial Fibrillation
This study of 179 incident cases of AF indicated a trend towards a negative association, although not statistically significant, between the content of marine n-3 PUFA in adipose tissue and the risk of incident AF (figure 2A). Similar trends, but again not statistically significant, were found for the individual marine n-3 fatty acids EPA, DHA and DPA (figure 2B–D).
Strengths of this study include the use of adipose tissue biopsies collected at baseline; detailed baseline information on participants; accurate information on vital status; and a nearly complete prospective follow-up from any hospital in the country. The adipose tissue content of marine n-3 PUFA reflects the endogenous exposure in the body and is affected by consumption, uptake, metabolism and release. In addition to dietary intakes, EPA, DHA and DPA can be synthesised endogenously from ALA although only to a very limited extent. Adipose tissue biopsies are often regarded as the preferred biomarker for the long-term intake of fatty acids, including marine n-3 PUFA. Because of a slow turnover time, adipose tissue reflects the dietary intake of marine n-3 PUFA during the previous 1–3 years. It is therefore a relevant biomarker when investigating the effects of long-term dietary exposure in relation to AF.
Limitations of our study include the fact that the study was confined to the number of participants (and relatively low number of cases) in the subcohort, and therefore it has wide CIs. The statistical model was limited in complexity according to a requirement of 15 cases per parameter. Likewise, the study was too small to allow for separate analyses for men and women. The subcohort had been randomly drawn in a sex-stratified manner for use in previous studies and, therefore, contains marginally more men than women compared with the full cohort; however, this is not likely to substantially affect the generalisability of the results. Results are reported as the proportion in % of total fatty acids, and the results thus depend on the content of other fatty acids. In terms of generalisation, the study only included middle-aged persons who had survived without a diagnosis of AF or cancer until inclusion into the study. A further limitation of the present study is that it concerned first registration of AF or AFL from a hospital visit and did not include undiagnosed persons with asymptomatic AF or patients only treated by their general practitioner or a specialist open-care outside hospitals. However, most patients with diagnosed AF are likely to have been referred for a subsequent hospital contact such as for echocardiography or cardioversion. Overall, no selection bias in terms of a systematic error related to the exposure and the outcome seems likely.
Another concern is that subjects may change dietary habits during follow-up. The participants were between 50 years and 64 years of age when entering the study, and people this age and older are likely to be relatively stable in their dietary habits. It seems unlikely that systematic changes in diet should occur that would affect the rank order of participants in regard to the adipose tissue content of n-3 PUFA and, at the same time, be related to incident AF. Outcome in this study was incident diagnosis of AF and/or AFL. The validity of this diagnosis is high with a positive predictive value above 92% in this population.
We adjusted for prespecified potential confounders, but because of the limited number of cases we could only adjust for the most important potential confounders. There were no major differences in the results of adjusted and unadjusted models, indicating that residual confounding is an unlikely explanation for our results. Since a potential mechanism of action of marine n-3 PUFA may be through an intermediate beneficial effect on ischaemic heart disease, the model was also fitted without adjustment for myocardial infarction and heart failure, but the results were not substantially affected by these adjustments (Table 2). We decided a priori not to control for potential confounding from diet, as control would make the interpretation of the fatty acid exposure unclear. In a posthoc sensitivity analysis adjusting for additional covariates, similar results were found (Table 2).
Previous studies have mainly used information on diet from food frequency questionnaires. Inherent methodological limitations of food frequency questionnaires include problems with self-reported dietary intakes; the difficulty individuals have in identifying fat sources, particularly for foods they did not prepare; and the errors in calculating food composition from database values and portion sizes. Although questionnaires and adipose tissue biopsies may serve as an estimate of the true dietary intake, they do reflect different aspects of dietary intake with questionnaires providing the best estimate of the person's knowledge on diet, whereas the content in adipose tissue serves as an objective measure of intake and the endogenous exposure.
Marine n-3 PUFA may have beneficial effects on various cardiovascular risk factors including blood pressure, heart rate, thrombosis, plasma triglycerides and inflammation, and many of these risk factors are related to the development of AF either per se or indirectly via occurrence of ischaemic heart disease. Furthermore, antiarrhythmic effects have been demonstrated in myocytes through effects on ion channels, electrical stabilising effects and fluidity of the cell membrane. In experimental animal studies on AF, marine n-3 PUFA reduce the shortening of atrial refractory period (electrical remodelling), atrial fibrosis (structural remodelling), postoperative AF and vagally induced AF. Also, in persons undergoing electrophysiological examination, long-term treatment with 1.8 g marine n-3 PUFA/day prolonged the atrial effective refractory period and made AF less inducible.
The results in the present study did not show a statistically significant association, but were suggestive of a rather strong negative association between the adipose tissue content of marine n-3 PUFA (and also the individual marine PUFA DHA, EPA and DPA) and the development of AF. The point estimates are in accordance with our a priori hypothesis and some previous studies but not others. Further studies using objective measures of the marine n-3 PUFA content in adipose tissue or other biomarkers are warranted to further explore the association between marine n-3 PUFA and AF.
Discussion
This study of 179 incident cases of AF indicated a trend towards a negative association, although not statistically significant, between the content of marine n-3 PUFA in adipose tissue and the risk of incident AF (figure 2A). Similar trends, but again not statistically significant, were found for the individual marine n-3 fatty acids EPA, DHA and DPA (figure 2B–D).
Strengths of this study include the use of adipose tissue biopsies collected at baseline; detailed baseline information on participants; accurate information on vital status; and a nearly complete prospective follow-up from any hospital in the country. The adipose tissue content of marine n-3 PUFA reflects the endogenous exposure in the body and is affected by consumption, uptake, metabolism and release. In addition to dietary intakes, EPA, DHA and DPA can be synthesised endogenously from ALA although only to a very limited extent. Adipose tissue biopsies are often regarded as the preferred biomarker for the long-term intake of fatty acids, including marine n-3 PUFA. Because of a slow turnover time, adipose tissue reflects the dietary intake of marine n-3 PUFA during the previous 1–3 years. It is therefore a relevant biomarker when investigating the effects of long-term dietary exposure in relation to AF.
Limitations of our study include the fact that the study was confined to the number of participants (and relatively low number of cases) in the subcohort, and therefore it has wide CIs. The statistical model was limited in complexity according to a requirement of 15 cases per parameter. Likewise, the study was too small to allow for separate analyses for men and women. The subcohort had been randomly drawn in a sex-stratified manner for use in previous studies and, therefore, contains marginally more men than women compared with the full cohort; however, this is not likely to substantially affect the generalisability of the results. Results are reported as the proportion in % of total fatty acids, and the results thus depend on the content of other fatty acids. In terms of generalisation, the study only included middle-aged persons who had survived without a diagnosis of AF or cancer until inclusion into the study. A further limitation of the present study is that it concerned first registration of AF or AFL from a hospital visit and did not include undiagnosed persons with asymptomatic AF or patients only treated by their general practitioner or a specialist open-care outside hospitals. However, most patients with diagnosed AF are likely to have been referred for a subsequent hospital contact such as for echocardiography or cardioversion. Overall, no selection bias in terms of a systematic error related to the exposure and the outcome seems likely.
Another concern is that subjects may change dietary habits during follow-up. The participants were between 50 years and 64 years of age when entering the study, and people this age and older are likely to be relatively stable in their dietary habits. It seems unlikely that systematic changes in diet should occur that would affect the rank order of participants in regard to the adipose tissue content of n-3 PUFA and, at the same time, be related to incident AF. Outcome in this study was incident diagnosis of AF and/or AFL. The validity of this diagnosis is high with a positive predictive value above 92% in this population.
We adjusted for prespecified potential confounders, but because of the limited number of cases we could only adjust for the most important potential confounders. There were no major differences in the results of adjusted and unadjusted models, indicating that residual confounding is an unlikely explanation for our results. Since a potential mechanism of action of marine n-3 PUFA may be through an intermediate beneficial effect on ischaemic heart disease, the model was also fitted without adjustment for myocardial infarction and heart failure, but the results were not substantially affected by these adjustments (Table 2). We decided a priori not to control for potential confounding from diet, as control would make the interpretation of the fatty acid exposure unclear. In a posthoc sensitivity analysis adjusting for additional covariates, similar results were found (Table 2).
Previous studies have mainly used information on diet from food frequency questionnaires. Inherent methodological limitations of food frequency questionnaires include problems with self-reported dietary intakes; the difficulty individuals have in identifying fat sources, particularly for foods they did not prepare; and the errors in calculating food composition from database values and portion sizes. Although questionnaires and adipose tissue biopsies may serve as an estimate of the true dietary intake, they do reflect different aspects of dietary intake with questionnaires providing the best estimate of the person's knowledge on diet, whereas the content in adipose tissue serves as an objective measure of intake and the endogenous exposure.
Marine n-3 PUFA may have beneficial effects on various cardiovascular risk factors including blood pressure, heart rate, thrombosis, plasma triglycerides and inflammation, and many of these risk factors are related to the development of AF either per se or indirectly via occurrence of ischaemic heart disease. Furthermore, antiarrhythmic effects have been demonstrated in myocytes through effects on ion channels, electrical stabilising effects and fluidity of the cell membrane. In experimental animal studies on AF, marine n-3 PUFA reduce the shortening of atrial refractory period (electrical remodelling), atrial fibrosis (structural remodelling), postoperative AF and vagally induced AF. Also, in persons undergoing electrophysiological examination, long-term treatment with 1.8 g marine n-3 PUFA/day prolonged the atrial effective refractory period and made AF less inducible.
The results in the present study did not show a statistically significant association, but were suggestive of a rather strong negative association between the adipose tissue content of marine n-3 PUFA (and also the individual marine PUFA DHA, EPA and DPA) and the development of AF. The point estimates are in accordance with our a priori hypothesis and some previous studies but not others. Further studies using objective measures of the marine n-3 PUFA content in adipose tissue or other biomarkers are warranted to further explore the association between marine n-3 PUFA and AF.
Source...