Omega-3 index in patients with severe diabetic ocular complications

Diabetes mellitus (DM) can lead to various ocular complications including diabetic macular edema (DME) and/or vitreous hemorrhage (VH) due to proliferative diabetic retinopathy (DR; [1]). These complications represent one of the leading causes of legal blindness, especially in working-age adults [1]. The pathophysiology of diabetic ocular complications is complex; however, two essential factors have been identified including the hypoxia-driven upregulation of vascular endothelial growth factor (VEGF) and a chronic retinal inflammation [2‐5].

Omega‑3 fatty acids (FA) were found to play a more significant role in this context than previously thought [6]. Specifically, a high intake of omega‑3 FAs reduces the risk of developing DR in individuals with well-controlled DM (HbA_1c < 7.0%; [6]). The reason for this is assumed to be anti-VEGF and the anti-inflammatory effect of omega‑3 FAs [7‐14]. In particular, omega‑3 FAs suppress the VEGF pathway by activating the transcription factor peroxisome proliferator-activated receptor‑α [14]. Omega‑3 FAs further suppress the transcription factor nuclear factor-κB, which consecutively leads to less production of pro-inflammatory cytokines including tumor necrosis factor‑α and interleukins (IL)-1 and IL‑6 [11, 12]. In addition, omega‑3 FAs represent the main structural lipids in photoreceptor membranes and are thus essential for their regeneration and vision in general [15].

The saturation of the body with omega‑3 FAs is generally lower in the diabetic than in the non-diabetic population [16]. However, we could not find any evidence regarding omega‑3 FA saturation in patients with severe DR undergoing vitrectomy. This subset of patients is of particular interest because the surgical outcomes often do not meet expectations. Since omega‑3 FAs partly modulate the course of DR, patients with DM might generally benefit from omega‑3 FA supply. The World Health Organization recommends an intake of polyunsaturated FAs including omega‑3 FAs of 6–11% of the total energy intake to improve health in general, but the majority of the population, especially those with DM, do not follow this recommendation [17, 18]. Moreover, the amount of saturated FAs should be concomitantly decreased, since a high intake of these FAs promotes the developments of severe DR and cardiovascular diseases [6, 19].

The aim of this study was to assess the proportion of omega‑3, saturated, and other relevant FAs in the erythrocytes of patients undergoing vitrectomy due to severe diabetic ocular complications.

This study was approved by the local ethics committee of the Medical University Graz and adhered to the tenets of the Declaration of Helsinki. We recruited patients scheduled for vitrectomy for diabetic complications including chronic DME or VH secondary to proliferative DR at the Department of Ophthalmology, Medical University Graz, between March 2020 and February 2021. The patients were included in this study after they gave their informed consent. For all patients, a complete ophthalmic examination including best-corrected visual acuity (BCVA) measurement, slit-lamp examination, indirect ophthalmoscopy, and optical coherence tomography (OCT) was performed prior to surgery. Patients with VH additionally underwent ocular ultrasound to assess whether the retina was attached. On the day of surgery, a fasting venous blood sample was taken using a 2-mL ETDA tube and sent to a highly specialized laboratory for FA analysis (Omegametrix®, Martinsried, Germany). A complete assessment of the proportion of FAs was made including omega‑3, omega‑6, monounsaturated (omega-9), saturated, and trans FAs in the membranes of erythrocytes using a standardized and patented “high-sensitivity” gas chromatography with flame ionization detection described elsewhere [20]. In detail, FA methyl esters were extracted from erythrocytes by acid transesterification and analyzed with the GC-2010 gas chromatograph (Shimadzu, Duisburg, Germany) with an SP2560, 100‑m capillary column (Supelco®, Bellefonte, PA, USA) using hydrogen as carrier gas. In total, 26 FAs were identified and quantified. The results were provided as percentage of total identified FAs after response factor correction. The coefficient of variation was 5%. The quality of analyses was controlled in accordance with DIN EN ISO 15189. Details of the FAs assessed are listed in Table 1. The omega‑3 index is defined as the proportion of eicosapentaenoic and docosahexaenoic acid in the membranes of erythrocytes. The omega-6/omega‑3 ratio is defined as \(\frac{\textit{omega}\,6}{\textit{omega}\,3}\).

Overall, 12 patients (5 female; 7 male) were enrolled in the study. Their average age was 69 ± 12 years (50–87). Two patients had type 1 DM and 10 patients type 2 DM. Of those with type 2 DM, seven patients were insulin-dependent. The HbA1c averaged 7.6 ± 1.2% (5.5–9.6) in general. HbA1c ≥ 7% was noted in eight patients and HbA1c < 7% was observed in four patients.

Overall, the omega‑3 index averaged 4.5 ± 1.2% (2.8–7.2). The omega-6/omega‑3 ratio was 5.2 ± 1.4: 1. Detailed results of the FA assessment are listed in Table 1.

There was no correlation between the preoperative or postoperative BCVA and the omega 3 index (r = −0.28, p = 0.38 and r = −0.18, p = 0.63, respectively; Spearman correlation). The omega 3 index was 4.8 ± 1.3% (2.8–7.2) in patients with DME and 3.8 ± 0.8% (2.9–4.8) in those with VH. This difference was not statistically significant (p = 0.21, Mann–Whitney U test). In patients with type 1 DM (n = 2) the omega‑3 index was 5.04 ± 0.3% (4.8–5.3) and in those with type 2 DM (n = 10) it was 4.4 ± 1.3% (2.8–7.2). Among those with type 2 DM, the omega‑3 index averaged 3.8 ± 0.7% (2.9–4.5) in patients with insulin dependency (n = 7) and 5.8 ± 1.3% (4.6–7.2) in those with non-insulin-dependent DM (n = 3); this difference was statistically significant (p = 0.017, Mann–Whitney U test). There were also substantial differences in the omega‑3 index depending on the HbA1c. In general, the omega‑3 index correlated with HbA1c (Fig. 1; r = −0.51, p = 0.09; Spearman correlation). In patients with well-controlled DM, defined as HbA1c < 7%, and poorly controlled DM (HbA1c ≥ 7%), the omega‑3 index was 5.02 ± 1.6% (2.9–7.2) and 4.1 ± 0.7% (2.8–4.8), respectively (p = 0.27, Mann–Whitney U test).

The proportion of other FAs including omega‑9, saturated, and trans FAs were within the normal range (Table 1). Solely the proportion of saturated FAs was slightly below the threshold. There was a negative correlation between omega‑3 index and omega‑6 FAs (r = −0.66, p = 0.02; Spearman correlation). There were no correlations between omega‑3 index and omega‑9 FAs (r = −0.31, p = 0.33; Spearman correlation), saturated FAs (r = 0.39, p = 0.2; Spearman correlation), and trans FAs (r = 0.44, p = 0.15; Spearman correlation).

Our study shows that the omega-3-index in diabetic patients with severe ocular complications requiring vitreoretinal surgery is significantly lower than the recommended optimum of 8–11% [16]. Moreover, in patients with poorly controlled DM (HbA1c ≥ 7.0%), the omega‑3 index was even lower compared to patients with well-controlled DM (HbA1c < 7.0%). Omega‑3 FAs play a significant role in the regulation of inflammatory reactions in general [7‐17]. First, omega‑3 FAs suppress the pro-inflammatory transcription factor NF-κB and thus reduce the production and attenuate the effect of pro-inflammatory cytokines including tumor necrosis factor‑α, IL‑1, and IL‑6 [11, 12]. Especially IL‑6 plays a significant role in the development and maintenance of DME [3]. Second, omega‑3 FAs metabolites including resolvin, protectin, and maresins promote resolution of inflammation processes [8]. Third, omega‑3 FAs immanently reduce the level of omega-6-derived pro-inflammatory arachidonic acid, since both omega‑3 and omega‑6 FAs are metabolized by the same enzymes [9]. This aspect is particularly important since the diet in developed countries has become deficient in omega‑3 and excessive in omega‑6 FAs over the past few decades [9, 16]. The oversupply in omega‑6 FAs consequently leads to an accumulation of arachidonic-acid-derived pro-inflammatory prostaglandins (H2, I2, F2, D2, E2‑3, A1), leukotrienes, and lipoxins [9, 16]. The level of omega‑6 FAs was 33.2% and thus relatively high according to the reference range of 18–39.6%. In addition to the anti-inflammatory aspect of omega‑3 FAs, they also promote an anti-VEGF effect by activation of peroxisome proliferator-activated receptor‑α, which subsequently inhibits the VEGF production [14].

Other FAs including omega‑9, saturated, and trans FAs were within the reference range (Table 1). However, the proportion of saturated FAs was 41.1% and thus slightly below threshold of 43.7%. This finding is consistent with the evidence that a high intake of saturated FAs is associated with advanced DR [6]. The reason for this is the fact that a high intake of saturated FAs leads to increased levels of triglycerides and low-density-lipoprotein (LDL) cholesterol, which migrate into the retinal vasculature and oxidate and consequently contribute to the progression of DR [6, 21‐24].

The implication of our study is that patients with DM might benefit from supplementation of omega‑3 FAs. It has already been shown that an increased daily intake of omega‑3 FAs (> 500 mg) over the years reduces the risk of DR-related vison loss by 48% by slowing down the progression of DR and improving DME [25]. Recently, a randomized controlled trial showed that patients with DME undergoing intravitreal anti-VEGF injections had a significantly better outcome when supplementation of omega‑3 FAs was additionally given over 24 months compared to the control group receiving intravitreal anti-VEGF injections alone [26]. Based on this evidence, omega‑3 FA supplementation might also optimize the outcome in patients with severe diabetic ocular complications undergoing vitreoretinal surgery. In this context, it is of interest that omega‑3 FA intake did not prove to be beneficial in patients with poorly controlled DM (HbA1c ≥ 7.0%; [6]). The reason for this is not fully understood; however, it is assumed that small LDL particles, which are more prevalent in individuals with poorly controlled DM, interact with omega‑3 FAs and thus diminish their effect [6]. In our study the omega‑3 index was evidently lower in patients with poorly controlled DM. This finding might be attributable to a less balanced diet and omega-3-FA-deficient nutrition of patients with poorly controlled DM.

The saturation of the body with omega‑3 fatty acids (FAs) is considerably low in patients with severe diabetic ocular complications who require vitreoretinal surgery, especially in those with poorly controlled diabetes mellitus (DM). Conversely, the levels of omega‑6 and saturated FAs were relatively high. Our study emphasizes that this subset of patients is generally under-supplied with omega‑3 FAs and over-supplied with omega‑6 and saturated FAs, which have disadvantageous effects on diabetic retinopathy (DR) development. Consequently, patients with DM might benefit from omega‑3 FA supplementation and reduction of saturated FAs in order to prevent or at least slow down the development of DR and ultimately preserve vision.