Blood omega-3 fatty acids and coronary artery stenosis: a postmortem study

Abstract

The role of omega-3 fatty acids in the atherosclerotic process remains unclear. Early studies demonstrated a profound anti-atherogenic effect of cod liver oil in a pig model, and autopsy reports from Alaska Natives consuming omega-3 rich diets have supported this hypothesis, but meta-analyses of omega-3 randomized trials have produced mixed results vis-à-vis cardiovascular disease. The present study was undertaken to examine the relationship between post-mortem blood omega-3 levels and maximal stenosis of the coronary arteries at autopsy. Blood was collected from the femoral artery of 112 sequential fatalities and analyzed for fatty acid composition. The erythrocyte EPA + DHA level (the omega-3 index) was found to have degraded compared to levels in 30 living controls, but the ratio of Arachidonic Acid (AA) to EPA was unaffected, thus it was used as a biomarker of omega-3 status. We found no association between the maximal coronary artery stenosis and the AA: EPA ratio. The omega-3 index estimated from the AA: EPA ratio in the cases was 2.1%, one of the lowest mean values for any population yet tested. Thus, the lack of association between omega-3 status and coronary atherosclerosis may be due to the very low omega-3 index in this cohort.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12944-025-02697-1.

Introduction

The “omega-3” story began with seminal studies by Bang and Dyerberg in Greenland Inuits in the 1970s [1]. The reduced risk for myocardial infarction among these native peoples was linked with higher dietary and blood omega-3 fatty acid (O3FA) levels, in particular eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). After EPA was found to be a competitive substrate with arachidonic acid (AA) for cyclo-oxygenase and the downstream production of prostaglandins, thromboxane and prostacyclin, the initial hypothesis put forward by these Danish investigators in 1978 was that EPA was anti-atherosclerotic because it prevented coronary artery thrombosis [2]. This hypothesis was extended to the prevention of coronary atherosclerosis by Weiner et al. who demonstrated in 1986 a remarkable inhibition of plaque formation in a pig model with cod liver oil supplementation [3]. More recently, work from Harvard researchers found in a randomized 30-month randomized trial giving 3.4 g/d of EPA + DHA slowed plaque formation [4], and the extent of progression was greatest in those with the lowest blood omega-3 levels [5].

This paper was conceived as an attempt to correlate post-mortem O3FA (and other related Fatty Acids (FAs) with coronary artery stenosis in a general population using the postmortem estimation of coronary artery stenosis cross sections. There is little information on this because, even if a systematic and random cohort of individuals could be autopsied for measurement of CAS, obtaining pre-mortem blood samples for O3FA analysis would be impossible absent knowledge of when any given individual would die.

Coronary artery angiography does correlate well (but not perfectly) [6] with postmortem evaluations but again, obtaining a random sampling of a general population to undergo this procedure would be impractical. Likewise, Coronary Artery Calcium Scores (CACs), which have been determined on millions of individuals in hospital screening programs around the country are a good predictor of coronary artery disease [7], but like all imaging methods, are subject to error and do not directly measure CAS [8, 9].

On autopsy, direct inspection of coronary arteries and direct assessment of CAS is routinely performed. Moreover, the variety of cases autopsied in forensic pathology offers the opportunity to compare the direct measurements of both CAS and FA in heterogeneous groups of people, ranging from younger subjects with no or minimal CAS, and older subjects dying of non-cardiac causes to those of various age groups dying directly due to severe and longstanding coronary artery disease. Such a group would offer a potentially good surrogate to the general population.

While post-mortem examination provides a direct and uniform assessment of CAS, it comes at a cost of potentially altered FA levels that does not represent one’s pre-mortem FA status, as already mentioned. The second thing to be considered in such analysis is that there should be an idea of general FA status in the population the autopsy group comes from, as for example, general dietary habits impact FA levels and may mask or mimic other factors.

So, in this study, we first aimed to assess the O3FA levels in healthy young non-obese adults and compare them with the O3FA levels in age and sex-matched autopsy cases (all without a severe coronary artery disease, if any), and thus identify the adequate O3 fatty acids whose levels do not significantly alter post-mortem. To the best of our knowledge, this has not been explored. Then, a prospective autopsy study on post-mortem O3FA levels and direct coronary artery stenosis assessment was conducted to analyze if there is any significant association between the two variables, with particular focus on O3 fatty acids that we previously established to be stable in post-mortem settings.

Methods

Study sample and case selection

This prospective study was performed on blood samples obtained from November to December 2023 from 30 living non-obese (i.e., Body Mass Index < 32) individuals aged 24 to 32 with no history of omega-fatty acid supplements intake, and from the postmortem blood samples obtained from eligible subjects autopsied at the Institute of Forensic Medicine, University of Belgrade – Faculty of Medicine, Belgrade, Serbia from June 2022 to September 2023. The study was approved by the Ethics Committee of the Faculty of Medicine University of Belgrade, Serbia (No. 17/IV-18 and No. 1322/XII-7) per the ethical guidelines of the 1975 Declaration of Helsinki. All living participants provided written consent.

During the observational period, all autopsy cases in which the postmortem period was less than 48 h and in which no macroscopically visible overt signs of putrefactive alteration were observed on external and internal examination. Assessments of CAS, performed by two forensic pathologists, were considered for inclusion in the research. The exclusion criteria were age under 15 years, death occurring in a hospital and/or emergency room (so that blood FA composition would not have been altered by intravenous fluid administration, parenteral nutrition, advanced life support), except if death was not pronounced shortly upon arrival to the ER (e.g., within 30 min).

Blood sampling and sample processing

Sampling blood of the living volunteers

The blood samples were obtained by lancing the fingertip after cleaning the skin with 70% ethanol. One drop of blood was put onto a proprietary dried blood spot (DBS) card that had been pre-treated with an antioxidant cocktail to prevent FA degradation in transit and storage. These were immediately put in the freezer at −20 °C until being shipped from Belgrade, Serbia to the OmegaQuant Laboratory in a single box with a transit time that did not exceed three days. Blood FA levels are stable under these conditions. The volunteers gave written informed consent, did not receive any financial compensation, and were informed of the test result by the laboratory.

Sampling postmortem blood on autopsy

During the autopsy, the blood sample was obtained by the needle punction of the femoral vein, a standard and routine autopsy procedure. Two drops (~ 0.05 mL) of blood were collected using the same DBS card as noted above. These cards were stored and shipped in the same manner as the sample from the living controls.

Assessing the coronary artery atherosclerosis

At autopsy, the heart was examined after evisceration and separation from the lungs and major blood vessels. After incising and opening the heart chambers, the serial-parallel transverse sections of the epicardial and superficial myocardial tissue at 0.5 cm intervals were done to make transverse sections of (1) the proximal half of the left coronary artery (LCA), including the proximal segment of the left anterior descending (LAD) (up to near the proximal 5 cm from the LCA ostium, typically initial 2–3 cm); (2) the proximal half of the right coronary artery (RCA), including the proximal segment of the posterior descending artery (PDA) (up to near the proximal 5 cm from the RCA ostium, typically initial 2–3 cm), and (3) up to 2 cm of the left circumflex artery (LCA). The sections were macroscopically examined during the autopsy directly by two forensic pathologists, and the degree (percentages) of the vessel diameter occlusion by atherosclerotic plaques was defined by assessing the section with the most pronounced narrowing of the lumen [10]. Each pathologist independently estimated the degree of stenosis, from 0 to 100%, at 5% increments (e.g., 5%, 15%, 60%, 85%). If the independent estimations were not identical, the assessment was repeated until agreement was reached.

In order to compare the FA composition of the blood from the Serbian volunteers to those of individuals in other western countries, we used data from the laboratory (OmegaQuant Analytic, LLC) which had accumulated data over many years from consumers across the US and EU.

To match the demographics of the volunteers from Belgrade with those from the west, the US and EU samples were limited to males between the ages of 20 and 30.

Analysis of dried blood spot fatty acids

The method used for measuring fatty acid content in DBS has been described previously [11]. Briefly, a punch from the DBS card was transferred to a glass vial and methanol containing 14% boron trifluoride, toluene, methanol (35:30:35 v/v/v) was added. The vial was capped, briefly vortexed and then heated in at 100˚C for 45 min. Hexane was added after cooling as was HPLC grade water. The tubes were vortexed and centrifuged to separate the aqueous from organic layers. A sample of the hexane layer was analyzed by gas chromatography (GC) using a GC-2010 Gas Chromatograph (Shimadzu Corporation, Columbia, MD) equipped with a SP-2560, 100-m fused silica capillary column (0.25 mm internal diameter, 0.2 μm film thickness; Supelco, Bellefonte, PA). Fatty acids were identified by comparison with a standard mixture (GLC OQ-A, NuCheck Prep, Elysian, MN). The following 24 fatty acids (by class) were identified: saturated (14:0, 16:0, 18:0, 20:0, 22:0 24:0); cis monounsaturated (16:1, 18:1, 20:1, 24:1); trans (t-16:1, t-18:1, t-18:2); cis n-6 polyunsaturated (18:2, 18:3, 20:2, 20:3, 20:4, 22:4, 22:5); and cis n-3 polyunsaturated (18:3, 20:5, 22:5, 22:6). Fatty acid data were reported as a percent of total fatty acids. Derived fatty acid metrics included the Omega-3 Index (i.e., red blood cell EPA + DHA), the AA: EPA ratio, and the omega-6: omega-3 fatty acid ratio.

Statistical methods

Standard demographic metrics are reported as means and SDs. Comparisons between living and dead subjects were computed by t-test. Pearson’s correlations were between maximum percent stenosis and the AA/EPA ratio. Fatty acid comparisons among US, EU and Serbian subjects were assessed by ANOVA testing followed by multiple t-tests. P-values of < 0.05 were considered statistically significant.

Results

A total of 112 individuals were included in this autopsy study in Belgrade, Serbia (Table 1). Their mean (SD) age was 50.4 (19.0); 72 were men and 40 were women. The mean (SD) age of the healthy controls was 26.4 (1.9) and all were men. The blood fatty acid profiles between these groups differed markedly (Table 1) with most differences being statistically significant. The causes of death are shown in Supplemental Tables 1–3.

Table 1.

The fatty acid composition of blood samples collected post-mortem (n = 112) and from healthy volunteers (n = 30)

	Living		Dead		p-value
	Mean	SD	Mean	SD
Omega-3 Index (%)*	3.61	0.41	2.51	0.91	< 0.0001
AA: EPA Ratio	70.55	22.43	58.45	36.92	0.045
Saturated (%)	36.44	1.26	34.51	4.58	< 0.001
Monounsaturated (%)	21.77	3.05	35.56	10.39	< 0.0001
Omega-6 (%)	38.01	3.08	27.34	7.83	< 0.0001
Omega-3 (%)	2.99	0.43	1.68	1.05	< 0.0001

* Calculated from the whole blood EPA + DHA content

The mean Omega-3 Index (RBC EPA + DHA) was 2.5% in the postmortem samples and 3.6% in the living controls. The major FA classes (saturated, monounsaturated and omega-6 and omega-3 polyunsaturated) in the blood all differed significantly between cases vs. the controls, with the AA: EPA ratio being the least different.

However, the average age of the dead individuals was 50 vs. 26 in the living Serbian controls. Since omega-3 blood levels rise and the major omega-6 fatty acid, linoleic acid (18:2n6) falls with age [12], part of the explanation for the differences in fatty acid patterns between the living and the dead individuals could have been age (or possibly dietary patterns associated with age). Accordingly, to identify a FA signal that was not affected by death, we compared the FA pattern in 16 of the postmortem samples that were from people in their 20 s (mean age 23) to those in the living controls (mean age 26). As is evident from Table 2, the marked difference in FA patterns between cases and controls was still evident even in individuals of about the same age with the clear exceptions of the AA: EPA ratio and total saturated FAs. Thus, these two FA biomarkers could reasonably be assumed to be less affected by death. Since the former has been proposed to be a marker of inflammatory status and is obviously related to O3FA biology (since it includes one major O3FA’s, EPA), we chose to focus on the AA: EPA ratio as our biomarker of omega-3 status for comparisons with the extent of coronary artery stenosis.

Table 2.

Comparison of blood fatty acids in living and dead Serbians in their 20s

	Dead (n = 16)		Alive (n = 30)		P-value
	Mean	SD	Mean	SD
Omega-3 Index (%)*	2.44	0.95	3.61	0.41	< 0.0002
AA: EPA Ratio	72.02	41.22	70.55	22.43	0.82
Saturated (%)	35.56	8.50	36.44	1.26	0.50
Monounsaturated (%)	35.77	11.82	21.77	3.05	< 0.0002
Omega-6 (%)	26.14	8.32	38.01	3.08	< 0.0001
Omega-3 (%)	1.60	1.03	2.99	0.43	< 0.0002

* Calculated from the whole blood EPA + DHA content

Coronary artery stenosis and AA: FA ratio

We calculated the maximum stenosis of any of the three coronary arteries from each autopsied individual and compared it with each individual’s blood AA: EPA ratio. We found that there was no relationship between these two variables (Fig. 1). A similar analysis using the blood EPA + DHA produced essentially the same (non)-relationship (data not shown).

Fig. 1.

Relationship between the AA: EPA ratio and the Maximum Coronary Stenosis

Causes of death and FA levels

There were 21 individuals who died of probable coronary artery disease-related causes (certified as acute myocardial infarction or coronary artery disease), and 91 who died of other causes (Supplemental Tables 1–3). Therefore, it was of interest as a secondary analysis to compare the AA: EPA ratio (mean SD) between these two groups; they were not significantly different (67 (37) and 56 (37), p = 0.21).

Blood fatty acid patterns in serbian, European and US men in their 20s

As data on Serbian blood FA status is relatively lacking, it was of interest to compare, in the small cohort of 30 healthy Serbian men in their 20 s, we compared convenience samples of whole blood FAs from 189 Europeans and 1067 US men of a similar age (Table 3).

Table 3.

Comparison of blood fatty acids in living US, EU and Serbians men in their 20s

FA	EU (n = 189)		US (n = 1067)		Serbian (n = 30)
	Mean	SD	Mean	SD	Mean	SD
Omega-3 Index	5.69%a	1.63%	5.51% a	1.92%	3.61%	0.41%
AA: EPA ratio	20.75 a	12.36	21.36 a	14.06	70.55	22.43
Saturated	37.53%	2.42%	35.9% a	2.0%	36.44%	1.26%
Monounsaturated	22.44% a	2.26%	19.8%	2.9%	21.77%	3.05%
Omega-6	34.34%	3.03%	38.0% a	3.3%	38.01%	3.08%
Omega-3	5.09%	1.68%	5.5%	1.9%	2.99%	0.43%

Means within a FA group with the same superscript are not significantly different (p > 0.05); otherwise, p ≤ 0.01

The mean O3I and total omega-3 levels were significantly lower in the Serbian men than either the wider EU or US cohorts, and correspondingly, the AA: EPA ratio was higher. However, the total saturated, mono-unsaturated and poly-unsaturated FA levels were similar.

As seen above, significant mean O3I value differences were noted between antemortem Serbian data, United States data and postmortem Serbian data. However, the more pressing question is “what was the estimated Omega-3 Index in the deceased individuals?” To determine this we used a previous comparison of whole blood AA: EPA and RBC EPA + DHA (the O3I) made from 100 random blood samples sent to OmegaQuant Analytics, LLC for analysis. The correlation between the two was 0.86 (Supplemental Fig. 1). Taking this approach, the estimated premortem O3I for the deceased subjects was about 2.1%.

Discussion

The population studies done by Bang and Dyerberg five decades ago revealed an inverse relation between O3FA’s and coronary artery disease. Experimental studies in pigs pointed to a potential anti-atherosclerotic effect of cod liver oil [3], and later studies’ in Alaska Natives found an inverse associations between the omega-3 content of adipose tissue and the extent of coronary atherosclerosis [13]. Randomized trials of O3FA pharmaceutical products with heart disease endpoints have produced mixed results, with some beneficial and other neutral findings [14]. Nevertheless, there remains significant interest in defining the health effects of O3FAs, whether consumed as from seafood or from dietary supplements (primarily fish oils).

Our purpose here was to test the hypothesis that O3FA levels in the blood (i.e., the O3I) would be inversely related to the extent of coronary artery stenosis (i.e., atherosclerosis). To definitively test this hypothesis would require collecting O3I data on perhaps hundreds of individuals very close to the moment of death, and then to directly examine their coronary arteries post-mortem. Such a study would of course be logistically impossible since the moment of death cannot be predicted. An in vivo study certainly could be done with non-invasive measurements of coronary artery stenosis, such as coronary artery angiography and other radiologic procedures. Unfortunately, in vivo measurements of coronary artery stenosis don’t always reflect those seen in autopsy examination [8, 9]. In addition, they are expensive and usually are performed only in individuals with known or suspected atherosclerotic coronary vessel disease, not the general population. To our knowledge, such studies linking CAS with blood O3FA level have not been done.

To circumvent these difficulties, we compared the post-mortem O3FA levels with the extent of autopsy derived CAS. It was first critical for us to evaluate whether our exposure marker (blood O3FA levels) was itself unaffected by death, that is, that the O3I (or other markers of O3FA status) were the same pre- vs. post-mortem. Since this could obviously not be done directly, a healthy control group was recruited and tested for O3FA status.

We found first, that the O3I appeared to fall in the post-mortem period; hence we could not test our hypothesis based on correlations with the O3I. We discovered, however, that the AA: EPA ratio in whole blood was relatively unaffected by death and thus could be used as a surrogate marker of O3FA status. Others have reported that a similar metric is also resistant to oxidative degradation [15]. Using this ratio as our exposure biomarker, we found no evidence that there is a correlation between the O3FA status and maximal CAS in this cohort.

There are several possible reasons for our findings of no association. First and most obvious is that there is, in fact, no relationship between O3FA status and the extent of coronary atherosclerosis. Another reason for this lack of relationship could be that the omega-3 levels were simply too low to affect the atherosclerotic process. The O3I in the living subjects was 3.6%, and the estimated (based on the observed AA/EPA ratio) of the deceased individuals was 2.1%, one of the lowest group means yet recorded (e.g., US vegans and US military deployed in Iraq both had mean O3I values of 3.4% [16, 17]). Given that an optimal O3I is ~ 8% [18] and that virtually none of the autopsy cases had cardioprotective O3I levels [19], there is little hope that omega-3 FAs could afford any anti-atherosclerotic protection.

Omega-3 FAs are believed to exert their beneficial health effects via a generalized anti-inflammatory action. Indeed, ten circulating biomarkers of inflammation were inversely related with the O3I in the Framingham Offspring study [20]. Thus, the extremely low O3I values in this Serbian cohort would be expected to exert little if any anti-inflammatory actions. EPA and DHA also inhibit platelet function resulting in lower risk for thrombosis [21, 22]. Again, the lack of these FAs could contribute to pro-thrombotic effects in the vascular endothelium. Both of the above would make it difficult to demonstrate an inverse relationship between a very low O3I and coronary artery stenosis.

Are postmortem AA: EPA levels clinically useable as a marker of antemortem O3IFA status? The AA: EPA ratio between antemortem and postmortem cohorts appears to be relatively stable, but to confirm that would require the logistically and ethically challenging process of collecting blood immediately before and soon after death. It could be explored in animal models, but exactly what the postmortem conditions would be to reflect the human situation are unknown. It is well known that there are chemical and physical changes that occur in a regular and semi-predictable fashion after death such as potassium levels in vitreous humor and rigor/livor/algor mortis. O3I values appear to fall from the antemortem values.

Study strengths and limitations

This study contributes with novel information on Serbian O3I levels that may inform future updates to the recently reported “O3I world map” [23]. It also provides direct evidence on potential postmortem alterations of fatty acid levels in contrast to clinical data but also on potential patterns of some alterations. This study indicates that there may be no overt difference between postmortem and clinical specimens regarding the AA: EPA ratio, so this can be useful in other similar research on postmortem samples.

Also, there are several limitations to this study. The first is the method of stenosis assessment, that was limited to evaluation only of the proximal segments of main coronary vessels – this does not exclude that significant stenosis was not present in distal segments of the coronary bed, which was not considered. However, macroscopic evaluation of CAS remains the common measure of documenting stenosis to be found on autopsy reports worldwide [10]. We do not believe that estimates of CAS found macroscopically, versus microscopically with or without image analysis, would have been significantly different had other methods been used. Furthermore, we believe that the macroscopic estimation of stenosis used in this study may improve ‘precision’ of the results given that the degree of stenosis was consistently applied by the two pathologists throughout the study. Although the sample size of deceased individuals was reasonable, the autopsy group was also not truly representative of a general population and had a male gender bias. Lastly, this sample size limited study power, which could also explain our null findings.

Conclusion

In conclusion, our study showed no relation between postmortem O3FA status and coronary artery stenosis. This is most likely due to the very low O3FA status in this cohort and the minimal range of O3FA levels with virtually no individuals having cardioprotective levels. Although our results have no direct impact on patient care, two things should be noted: 1) the general population from which the autopsy and control sample was derived may be at risk of altered O3FA levels and low O3I status, and 2) the AA: EPA ratio may be a useable postmortem marker for indicating antemortem O3I status, if it was unknown prior to death. Clinically unknown, i.e., antemortem O3l status could be better assessed by postmortem analysis by determining the degradation dynamics of these FA. Such estimation would be best if the samples are obtained as soon as possible after death onset with precise information on postmortem interval: this should be considered in further similar autopsy studies. This may even be interesting and useful from the perspective of forensic pathology, as it is possible that the rate of degradation of omega-3 fatty acids in the blood post-mortem could, in theory, be used to determine the death interval.

Authors’ contributions

The authors are solely responsible for the design and conduct of the study; collection, management, analysis and interpretation of the data; as well as preparation of the manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

Dr. Harris holds an interest in OmegaQuant Analytics, LLC, a laboratory that offers blood fatty acid testing to researchers, clinicians and the public. Dr. Randall serves as the Medical Director of OmegaQuant. The other authors have no conflicts to disclose.