Next Article in Journal
Dietary Intake, Feeding Pattern, and Nutritional Status of Children with Cerebral Palsy in Rural Bangladesh
Next Article in Special Issue
Modulatory Properties of Vitamin D in Type 2 Diabetic Patients: A Focus on Inflammation and Dyslipidemia
Previous Article in Journal
Konjac Glucomannan Counteracted the Side Effects of Excessive Exercise on Gut Microbiome, Endurance, and Strength in an Overtraining Mice Model
Previous Article in Special Issue
Cholesterol Paradox in Older People with Type 2 Diabetes Mellitus Regardless of Lipid-Lowering Drug Use: A Cross-Sectional Cohort Study
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Nutrients
Volume 15
Issue 19
10.3390/nu15194208
nutrients-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Mendelian Randomization Analysis Reveals Causal Effects of Polyunsaturated Fatty Acids on Subtypes of Diabetic Retinopathy Risk

by
Shaojie Ren
,
Chen Xue
,
Manhong Xu
and
Xiaorong Li
*
Tianjin Key Laboratory of Retinal Functions and Diseases, Tianjin Branch of National Clinical Research Center for Ocular Disease, Eye Institute and School of Optometry, Tianjin Medical University Eye Hospital, Tianjin 300384, China
*
Author to whom correspondence should be addressed.
Nutrients 2023, 15(19), 4208; https://doi.org/10.3390/nu15194208
Submission received: 27 August 2023 / Revised: 18 September 2023 / Accepted: 27 September 2023 / Published: 29 September 2023
(This article belongs to the Special Issue Effects of Nutritional Interventions and Lifestyle on Diabetes and Lipid Metabolism)
Browse Figures
Review Reports Versions Notes

Abstract

:
Polyunsaturated fatty acids (PUFAs) affect several physiological processes, including visual acuity, but their relationship with diabetic retinopathy (DR) remains elusive. The aim of this study was to determine whether PUFAs have a causal effect on DR. PUFAs- (total and omega-3 [FAw3] and omega-6 [FAw6] fatty acids and their ratio) and DR-associated single nucleotide polymorphisms derived from genome-wide association studies; sample sizes were 114,999 for fatty acids and 216,666 for any DR (ADR), background DR (BDR), severe non-proliferative DR (SNPDR), and proliferative DR (PDR). We hypothesized that the intra-body levels of PUFAs have an impact on DR and conducted a two-sample Mendelian randomization (MR) study to assess the causality. Pleiotropy, heterogeneity, and sensitivity analyses were performed to verify result reliability. High levels of PUFAs were found to be associated with reduced risk of both ADR and PDR. Moreover, FAw3 was associated with a decreased risk of PDR, whereas FAw6 demonstrated an association with lowered risks of both BDR and PDR. Our findings provide genetic evidence, for the first time, for a causal relationship between PUFAs and reduced DR risk. Consequently, our comprehensive MR analysis strongly urges further investigation into the precise functions and long-term effects of PUFAs, FAw3, and FAw6 on DR.

1. Introduction

Diabetic retinopathy (DR), categorized into non-proliferative and proliferative stages (NPDR and PDR, respectively), constitutes the principal ocular sequel of diabetes mellitus, impacting approximately 30–40% of individuals with diabetes [1,2]. Globally, over 132 million individuals are affected by diabetic retinopathy (DR), a significant driver of visual impairment and blindness, especially within the working-age adult demographic [1,3]. The worldwide occurrence and impact of DR are projected to experience a substantial escalation over the next few decades, increasing from around 103 million cases in 2020 to 130 million by 2030 and further to 161 million by 2045 [4]. This sudden rise in the prevalence of DR by more than 25% in just 10 years is likely to put additional strain on the already-overburdened healthcare systems and resources. The financial expenses related to DR and its complications are substantial [5].
Polyunsaturated fatty acids (PUFAs), encompassing both omega-3 fatty acids (FAw3) and omega-6 fatty acids (FAw6), are considered to have numerous beneficial effects due to their anti-inflammatory and anti-aging properties. Consequently, the supplementation of PUFAs in the diet may serve as an effective adjunct therapy, particularly in the context of various diseases that are rooted in aging, cellular dysfunction, and prolonged exposure to light, notably eye diseases [6]. As essential constituents of the retina, PUFAs play a pivotal role in ensuring the survival of photoreceptors and actively participating in visual signal transmission, thereby maintaining the dynamic equilibrium and promoting the overall health of the retina [7,8,9]. Recently, high levels of PUFAs, particularly FAw3, have been found to be associated with a reduced risk of DR progression [10,11,12,13], whereas controversies persist regarding the effects of FAw6 [14,15]. Furthermore, recognizing that the pathological mechanisms underlying early-stage DR and advanced DR differ significantly is imperative [16]. A comprehensive understanding of these distinctions is crucial for devising tailored strategies for both prevention and treatment. However, it is worth noting that existing research on PUFAs and DR primarily focuses on a holistic perspective, lacking a refined classification based on different stages of the condition.
Serum lipid profiles are considerably altered by variables such as fasting and medication use, independent of the underlying disease, and conventional studies cannot completely rule out confounding bias or reverse causality [17,18]. Randomized controlled trials (RCTs) are typically regarded as the gold standard in clinical evidence [19], but to date, no RCTs have examined the effect of PUFAs on DR, possibly owing to difficulties in meeting the criteria of sample size and intervention duration. Mendelian randomization (MR) has become a popular and practical approach, employed mainly for epidemiological etiological inference in recent years, and can imitate RCTs without much difficulty [20,21]. It employs genetic variation as an instrumental variable (IV) to ascertain causal relationships between biomarkers and outcomes [22]. Utilizing the inherent randomization provided by genetic variation, this approach circumvents the issues associated with confounding factors and reverse causation. It serves as a suitable method to assess the connection between serum lipid levels and relevant diseases [23].
In this study, we hypothesized that the intracellular levels of PUFAs have an impact on different subtypes of DR. To prove the hypothesis, we estimated bidirectional causal effects between total PUFAs and any DR (ADR), and performed a 2-sample MR study to comprehensively investigate the potential causality of different types of PUFAs (total PUFAs, FAw3, FAw6, and FAw6/FAw3) on three DR phenotypes—background DR (BDR), severe NPDR (SNPDR), and PDR—to assess the associations between PUFAs and DR.

2. Materials and Methods

2.1. Study Design

In this study, we conducted a bidirectional MR analysis utilizing summary genetic associations from various genome-wide association studies (GWAS) for both total PUFAs and ADR. Next, we used 2-sample MR to examine the associations of different genetically determined PUFA traits (total PUFAs, FAw3, FAw6, and ratio of FAw6 to FAw3) with three DR phenotypes (BDR, SNPDR, and PDR). A schematic diagram outlining the process of the two-sample MR study investigating the relationship between PUFAs and DR is presented in Figure 1. Furthermore, we have adhered to the MR-STROBE guidelines in reporting our findings.

2.2. Data Source

2.2.1. Exposure Data

The most recent and extensive GWAS summary statistics for PUFAs were obtained from the UK Biobank, a population-based cohort comprising around 500,000 individuals (approximately 5% of those invited). These statistics were extracted from the IEU database (https://gwas.mrcieu.ac.uk/, accessed on 1 May 2023). The GWAS for PUFAs contained 114,999 samples, all of which were from Europe, and patients ranged in age from 40 to 69 years [24]. This GWAS explains 4.8–7.9% of the variance in cyclic PUFAs.
Circulating concentrations of FAw3 and FAw6 were quantified using a targeted high-throughput nuclear magnetic resonance metabolomics platform (Nightingale Health Ltd., Helsinki, Finland; biomarker quantification version 2020) [25]. Pre-release data from a randomly selected subset of 126,846 non-fasting plasma samples collected at baseline or during the first repeat assessment were provided to early access analysts. Subsequently, 121,577 samples were retained for analysis after the removal of duplicates and observations that did not meet quality control criteria.
In this study, four PUFA traits were selected as exposure amounts: total PUFAs, circulating FAw3, FAw6, and the ratio of FAw6 to FAw3. The average levels of circulating fatty acids among participants in the UK Biobank were 0.53 mmol/L (SD: 0.22) for total FAw3 and 4.45 mmol/L (SD: 0.68) for FAw6. These concentrations corresponded to 4.4% and 38% of the total fatty acids, respectively.

2.2.2. Outcome Data

The DR GWAS data were downloaded from the FinnGen research project (https://r5.finngen.fi/, accessed on 1 May 2023) [26]. Population isolates like Finland offer advantages in genetic research, concentrating damaging alleles on low-frequency variants due to historical bottlenecks. FinnGen, studying 500,000 Finns, has enriched data for rare diseases and late-onset conditions. The project included a GWAS on 1932 clinical endpoints, identifying genome-wide significant associations at 2491 loci, including 148 putative causal coding variants, with 62 having over two-fold enrichment in Finland and a low allele frequency (<10%) [27]. DR was identified using ICD-10 codes (H36.0 or E11.3). A total of 16,962,023 variants were analyzed from 218,792 subjects through SAIGE GWAS (https://github.com/weizhouumich/SAIGE, accessed on 1 May 2023). Four DR phenotypes were selected as exposures based on different levels of DR severity: ADR (ICD 10: H36.0*; 18,097 cases and 206,364 controls), BDR (ICD 10: H36.00; 2510 cases and 242,308 controls), SNPDR (ICD 10: H36.02; 568 cases and 242,308 controls), and PDR (ICD 10: H36.03; 10,860 cases and 242,308 controls).
For the analysis, the possible confounders of age, sex, genotyping batches, genetic correlation, diabetes course, hypertension and cardiovascular history, and average daily dosage of hypoglycemic agents after cohort entry were adjusted. Further details on FinnGen are described in https://finngen.gitbook.io/documentation/v/r7/, accessed on 1 May 2023. The FinnGen study is a Finnish national meta-analysis of GWAS from nine biobanks and has limited overlap with the UK Biobank GWAS from several centers in the UK. Therefore, we believe that the limited sample overlap between the PUFA and DR data leads to minimal risk of bias.

2.3. Selection of Genetic Instruments

We selected single-nucleotide polymorphisms (SNPs), representing instrumental variables (IVs), after screening, in accordance with the three core assumptions of MR: (i) the “correlation” hypothesis, in which IVs are closely related to exposure; (ii) the assumption of “independence”, in which IVs are not related to confounders; and (iii) the “exclusion limitation” hypothesis, in which IVs do not affect results by means other than exposure, namely, excluding horizontal pleiotropy [24]. A concise outline of the MR design is presented in Figure 1.
To fulfill the three fundamental assumptions of MR, we established the following criteria to screen SNPs used as IVs:
(1)
All SNPs underwent screening at the genome-wide significance threshold (p < 5 × 10−8).
(2)
Using the “ld_clump” R package, linkage disequilibrium between SNPs (R2 < 0.001 and <10,000 from the index variant) was identified to ensure their independence [28].
(3)
We aligned effect alleles of outcome-related SNPs with those of exposure-related SNPs based on allelic letters and frequencies and removed palindromic SNP alleles [28].
(4)
We used the PhenoScanner database (https://www.phenoscanner.medschl.cam.ac.uk/, accessed on 1 May 2023) to verify whether the SNP loci were associated with other confounding factors [29].

2.4. MR Methods

Prior to conducting the MR analysis, we computed F statistics for these IVs to assess the presence of weak IV bias. In all instances, we observed that F > 10. Therefore, owing to the diminished impact of weak IV bias, the chosen SNPs were utilized in the MR study [30].
In our study, we applied three complementary approaches—inverse-variance-weighted (IVW), MR-Egger regression, and weighted median (WM)—to assess the causal effects of exposures on outcomes. The primary outcome was obtained using the IVW method. This technique involves assigning weights based on the inverse variance of each study, followed by calculating the weighted average effect size and summarizing the effect sizes from multiple independent studies. When all selected SNPs are valid IVs, this method yields the most accurate estimation results [31]. To enhance the robustness of the IVW estimates, we employed the WM and MR-Egger methods. WM is the median of the weighted empirical distribution function of individual SNP ratio estimates. Approximately 50% of the weights in the analysis come from ratio estimates smaller than or equal to the WM. This method provides a consistent effect estimate if more than 50% of the information comes from valid SNPs [32]. MR-Egger allows all genetic variants to have a pleiotropic effect but requires that the pleiotropic effects be independent of the variant–exposure association [33]. Although WM and MR-Egger methods generally offer more reliable estimates across a wider range of scenarios, these can be less efficient. Ultimately, the causal estimates were expressed as odds ratios (ORs) along with their corresponding 95% confidence intervals (CIs).

2.5. Statistical Analysis

The “TwoSampleMR” (https://github.com/mrcieu/TwoSampleMR, accessed on 1 May 2023) and MR Pleiotropy RESidual Sum and Outlier (MR-PRESSO, version 4.2.0) packages in R (version 3.6.1; The R project for statistical computing) were used for all analyses. For the bidirectional MR of total PUFAs and ADR, a p-value < 0.05 on both sides indicates significance. The Bonferroni correction, a conservative method, was used to examine the associations between exposure (i.e., total PUFAs, FAw3, FAw6, and FAw6/FAw3) and DR outcomes (BDR, SNPDR, and PDR). For region-level analyses, a Bonferroni-corrected p-value < 4.17 × 10−3 [0.05/(3 × 4)] on both sides is regarded to indicate significance, a p-value < 0.05 on both sides is regarded to indicate nominal significance, and a p-value between 4.17 × 10−3 and 0.05 is considered as suggestive evidence.
We corrected horizontal pleiotropy by detecting and removing outliers using the MR-PRESSO test and determined whether a substantial change in causality before and after removing outliers was present [34]. MR-Egger regression was then used to evaluate the possibility of average horizontal pleiotropy among the IVs, and funnel charts were used to visualize directional pleiotropy. If the MR-Egger intercept is not statistically significant (p > 0.05), the relative symmetry of funnel plots can be considered as an indicator of horizontal pleiotropy [33,35].
Additionally, we employed Cochran’s Q test (significance set at p > 0.05 to indicate no heterogeneity) to evaluate potential heterogeneity among SNPs in IVW estimates. To ensure the robustness and consistency of the findings, we conducted a sensitivity analysis using a “leave-one-out” approach. Horizontal pleiotropy is corrected by removing outliers, which involved performing MR again with the omission of one SNP at a time [36].

3. Results

3.1. IVs for PUFAs and DR

Our rigorous screening process was based on the independence and exclusivity hypotheses as well as on the harmonization and removal of palindromic SNPs with intermediate allele frequencies. We thus meticulously selected 52 SNPs for MR analysis to assess the associations between total PUFA and ADR. Similarly, to evaluate the relationship between ADR and total PUFAs, eight SNPs were selected as IVs. In this study, we also investigated the causal effects of different types of PUFAs (total PUFA, FAw3, FAw6, and FAw6/FAw3) on three specific DR phenotypes (BDR, SNPDR, and PDR) using MR analysis. In total, 34–57 SNPs were chosen for this analysis. The F-statistics of the instruments exceeded 10 for all selected SNPs, indicating satisfactory instrument strength. Detailed information regarding the instruments is provided in Supplementary Table S1.

3.2. Bidirectional Causal Effects between Total PUFAs and ADR

The MR results revealed a substantial association between the elevated genetically predicted total PUFA levels and a reduced risk of ADR (IVW: OR, 0.90; 95% CI, 0.83–0.97; p = 5.05 × 10−3; MR-Egger: OR, 0,86; 95% CI, 0.74–1.00; p = 0.05; WM: OR, 0,90; 95% CI, 0.81–1.00; p = 0.05). The scatter plot (Figure 2) shows that the causal effects among the IVW, MR-Egger regression, and WM methods were consistent. However, no association between ADR and total PUFAs was found using any of the methods (all p > 0.05). Additional details for the MR analysis are shown in Figure 3.

3.3. Causal Effects of Different Types of PUFAs on Three DR Phenotypes

The results of MR analysis showed that genetically-predicted total PUFAs were associated with decreased risk of BDR (IVW: OR, 0.79; 95% CI, 0.65–0.98; p = 2.90 × 10−2; MR-Egger: OR, 0.97; 95% CI, 0.65–1.45; p = 0.87; WM: OR, 0.75; 95% CI, 0.57–1.00; p = 0.05) and PDR (IVW: OR, 0.86; 95% CI, 0.78–0.95; p = 2.63 × 10−3; MR-Egger: OR, 0.86; 95% CI, 0.72–1.04; p = 0.13; WM: OR, 0.88; 95% CI, 0.77–1.01; p = 0.08). FAw3 was associated with decreased risk of PDR (IVW: OR, 0.83; 95% CI, 0.74–0.93; p = 1.50 × 10−3; MR-Egger: OR, 0.82; 95% CI, 0.67–1.00; p = 0.06; WM: OR, 0,81; 95% CI, 0.70–0.94; p = 0.02) and FAw6, with decreased risk of BDR (IVW: OR, 0.69; 95% CI, 0.56–0.86; p = 9.20 × 10−4; MR-Egger: OR, 0.82; 95% CI, 0.53–1.27; p = 0.37; WM: OR, 0,65; 95% CI, 0.47–0.90; p = 9.25 × 10−3) and PDR (IVW: OR, 0.82; 95% CI, 0.74–0.91; p = 3.51 × 10−4; MR-Egger: OR, 0.89; 95% CI, 0.72–1.10; p = 0.28; WM: OR, 0.82; 95% CI, 0.69–0.96; p = 0.1). The causal effects using IVW, MR-Egger regression, and WM methods were consistent, as shown in the scatter plot (Figure 2). However, no association between any of the three fatty acids and SNPDR was found using any of the methods (all p > 0.05). In addition, no association between the combined FAw6/FAw3 ratio and any of the DR phenotypes was found using any of the methods (all p > 0.05). Further details of the MR analysis are shown in Figure 3.

3.4. Sensitivity Analysis

To bolster the credibility of the aforementioned outcomes, our assessment through MR-Egger regression analysis yielded no evidence of directional pleiotropy or heterogeneity (all p > 0.05), as depicted in Figure 3. Additionally, funnel plots were used to visually represent directional pleiotropy, displayed a symmetrical distribution in Supplementary Figure S1. Moreover, after outlier SNPs were deleted, the application of MR-PRESSO, along with the examination of leave-one-out and funnel plots (Supplementary Figure S2), did not reveal any outliers.

4. Discussion

In this study, we used GWAS data to perform bidirectional MR analyses based on the genetic associations of overall PUFAs and DR and conducted a two-sample MR study to examine the causal relationship between genetically determined PUFA features (overall PUFAs, FAw3, FAw6, and FAw6/FAw3) and three DR phenotypes (BDR, SNPDR, and PDR).
Our main findings, along with the sensitivity analyses, demonstrated that a genetic predisposition to higher levels of overall PUFAs was associated with a reduced risk of DR. Reverse MR analysis did not provide any genetic evidence of ADR influencing PUFAs. The total levels of PUFAs were associated with a decreased risk of BDR and PDR. Regarding specific fatty acid levels, elevated FAw3 levels were found to be linked to decreased risk of PDR, whereas increased FAw6 levels were associated with reduced risk of both BDR and PDR.
PUFAs are indispensable fatty acids that the human body cannot synthesize and must be acquired through dietary intake, and are found abundantly in fatty fish, nuts, and seeds [37]. PUFAs can be classified into two main types: FAw3 and FAw6 [13,38]. In recent years, numerous observational studies have explored the role of PUFAs in DR development [39,40,41]. One prospective study by Sala-Vila et al. [39] provided initial evidence of a reduced risk of severe DR in middle-aged and older individuals with type 2 diabetes who consumed fish oil. Similarly, Knud Erik Alsbirk et al. [10] conducted an observational study in a coastal population that exhibited high fish oil intake, comprising 510 Norwegian individuals with diabetes, and they observed that none of the participants experienced a decline in best corrected visual acuity to 0.3 (0.48 logMAR) due to DR, suggesting a potential protective effect of PUFAs against the incidence and progression of DR microvascular complications. Animal studies have also provided convincing evidence of the protective effects of PUFAs on DR development [39,40]. However, a few studies have reported no association or even harmful effects of PUFAs on DR [42,43]. Currently, there are only a few long-term randomized studies that have investigated the relationship between PUFA intake and DR. This scarcity may be attributed to the challenges in controlling for confounding factors in lipid measurements, such as age, sex, body mass index, medication usage, fasting status, and dietary habits [41,42]. The MR approach, which we employed in this study, not only mimics RCTs but also offers a suitable method for evaluating the relationship between blood lipids and the disease of interest. MR avoids confounding factors and reverse causality without requiring interventions on humans or animals. It is commonly used in epidemiological causal inference investigations and serves as a genetic validation tool for observational studies [36,44]. Our findings align with the majority of research findings that suggest a protective effect of high levels of total PUFAs on DR.
FAw3 primarily comprise long-chain omega-3 fatty acids, including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) [45]. Accumulating evidence suggests that supplementation with FAw3 benefits patients with DR [40,41]. Nevertheless, no studies have investigated the relationship between FAw3 intake and different DR subtypes. Our MR study aimed to address this knowledge gap. The findings indicated a potential mitigating effect of FAw3 on PDR, whereas FAw3 had no protective causal relationship on BDR or SNPDR, consistent with prior research linking increased FAw3 intake to reduced risk of severe visual decline. Moreover, FAw3 intake lowers the risk of pathological retinal neovascularization and severe visual impairment [11,46,47,48]. Notably, lipotoxins, hemolysins, and protective proteins derived from FAw3 exhibit anti-angiogenic properties and hold clinical potential for preventing diabetic macular edema and retinopathy [49].
The estimates provided in this study also consider the long-term effects of FAw6, which remain an important question to be addressed. Evidence regarding the association between FAw6 levels and the risk of DR is limited and is the subject of ongoing debate. Linoleic acid (LA), an FAw6, possesses anti-inflammatory properties and can modulate inflammation through various pathways, including immune cell function, and inhibition of production of inflammatory mediators among other actions [50]. Arachidonic acid (AA), another FAw6, plays a significant role in inflammatory responses and can be metabolized by enzymes such as cyclooxyuigenase and lipoxygenase to generate various inflammatory mediators, including prostaglandins and leukotrienes [50]. In contrast to FAw3, FAw6 are considered to have pro-inflammatory properties, and thus, maintaining a lower FAw6/FAw3 ratio is crucial for achieving favorable outcomes in ocular pathology [51]. However, studies on healthy adults have demonstrated that increasing intake of AA or LA does not necessarily result in elevated concentrations of several inflammatory markers [52,53,54]. Epidemiological research has even suggested that AA and LA are associated with reduced inflammation [42,55]. Additionally, a study by Fu et al. [56] showed that a diet rich in FAw6 accelerated the maturation process of retinal neurons while inducing increased metabolism to maintain the energy balance of retinal neurons. A study conducted by Kai Wang et al. [57] revealed that LA plays a significant role in the negative correlation between FAw6 and susceptibility to age-related macular degeneration. The controversy surrounding FAw6 may arise from the differential roles of LA and AA in the eye.
Based on the results of our MR study, we suggest a consistent mitigating effect of FAw6, including LA and AA, on the risk of PDR. Of note, compared with FAw3, FAw6 also exhibited a mitigating effect on the risk of BDR, although neither type had an effect on SNPDR. This suggests an overall beneficial role for FAw6 in DR, although the specific mechanisms are yet to be studied. The interaction between FAw3 and FAw6, as well as their lipid mediators, in the context of inflammation is complex and not yet fully understood [55,58,59]. Lastly, we did not find any significant influence of the FAw6 to FAw3 ratio on any DR subtype.
Notably, dietary sources of PUFAs are limited in Western diets, which highlights the importance of dietary interventions or nutritional supplementation, especially in pregnant women, children, and older populations [60]. Therefore, implementing simple measures to enhance the production of endogenous anti-inflammatory molecules and adopting seemingly straightforward interventions throughout the entire course of DR in patients with long-term diabetes may be beneficial [37]. The establishment of a therapeutic approach for administering FAw3 to treat DR remains an important unanswered question. Therefore, further research is encouraged to determine the specific role and long-term effects of FAw3 and FAw6 in the treatment of DR.

Limitations

This study has certain limitations. First, it is important to highlight that while the GWAS summary statistics for PUFAs were obtained from the UK Biobank, the data for DR were sourced from the Finnish database. Although all selected instrumental variables (IVs) were robust, potential bias could arise due to sample overlap [61]. Second, Our MR study demonstrated a causal relationship between genetically predicted PUFAs and DR; nevertheless, it is important to note that the findings of the MR analysis are based solely on genetic evidence. Third, we considered FAw3 or FAw6 components as a single entity and analyses of the causality of individual FAw3 or FAw6 members separately should be explored in future studies. These limitations highlight the need for continued research to address the potential biases arising from sample overlap to further examine the causal relationship between PUFAs and DR through complementary experimental approaches; additionally, the analysis of individual components of FAw3 and FAw6 has the potential to provide a more comprehensive understanding of the specific roles of these fatty acids in DR.

5. Conclusions

Our findings present initial genetic evidence to support a causal association between PUFAs and a reduction in DR risk. Specifically, the presence of both FAw3 and FAw6 demonstrates favorable effects on PDR, whereas FAw6 fatty acids play a mitigating role in BDR. However, neither type exhibits a causal relationship with SNPDR. Nonetheless, establishing an effective therapeutic approach for the administration of FAw3 and FAw6 in the context of DR remains an unanswered and critical question. Consequently, our comprehensive MR analysis strongly encourages further research to ascertain the precise roles and long-term effects of PUFAs, FAw3, and FAw6 on DR treatment.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/nu15194208/s1, Supplementary Figure S1: Leave one out sensitivity tests of MR analyses of PUFAs and DR; Supplementary Figure S2: Funnel plot of MR analyses of PUFAs and DR; Supplementary Table S1: Genetic variants used to estimate the effects of DR on PUFAs in MR analysis. Reference [62] is cited in the supplementary materials.

Author Contributions

Study concept and design, S.R.; acquisition of data, S.R. and C.X.; analysis or interpretation of data, M.X.; writing of the manuscript, S.R., C.X. and M.X.; critical revision of the manuscript for important intellectual content, X.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Natural Science Foundation of China (82171085) and Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-037A).

Institutional Review Board Statement

Ethical approval was not sought for this specific project because all data came from the summary statistics of published GWAS, and no individual-level data were used.

Informed Consent Statement

Not applicable.

Data Availability Statement

The datasets analyzed during the current study are available in the https://gwas.mrcieu.ac.uk and https://msk.hugeamp.org, accessed on 1 May 2023.

Acknowledgments

All authors would like to thank the Tianjin Key Laboratory of Retinal Functions and Diseases, Tianjin Branch of National Clinical Research Center for Ocular Disease, Eye Institute and School of Optometry, Tianjin Medical University Eye Hospital.

Conflicts of Interest

All authors declare that we have no financial and personal relationship with other people or organizations that can inappropriately influence our work and that there is no professional or other personal interest of any nature or kind in any product, service, and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled “Causal Associations Between Polyunsaturated Fatty Acids and Diabetic Retinopathy: A Two-Sample Mendelian Randomization Study”.

References

  1. Tan, T.E.; Wong, T.Y. Diabetic retinopathy: Looking forward to 2030. Front. Endocrinol. 2023, 13, 1077669. [Google Scholar] [CrossRef] [PubMed]
  2. Wong, T.Y.; Cheung, C.M.; Larsen, M.; Sharma, S.; Simó, R. Diabetic retinopathy. Nat. Rev. Dis. Primers 2016, 2, 16012. [Google Scholar] [CrossRef] [PubMed]
  3. Lin, K.Y.; Hsih, W.H.; Lin, Y.B.; Wen, C.Y.; Chang, T.J. Update in the epidemiology, risk factors, screening, and treatment of diabetic retinopathy. J. Diabetes Investig. 2021, 12, 1322–1325. [Google Scholar] [CrossRef] [PubMed]
  4. Teo, Z.L.; Tham, Y.C.; Yu, M.; Chee, M.L.; Rim, T.H.; Cheung, N.; Bikbov, M.M.; Wang, Y.X.; Tang, Y.; Lu, Y.; et al. Global Prevalence of Diabetic Retinopathy and Projection of Burden through 2045: Systematic Review and Meta-analysis. Ophthalmology 2021, 128, 1580–1591. [Google Scholar] [CrossRef] [PubMed]
  5. Grauslund, J. Diabetic retinopathy screening in the emerging era of artificial intelligence. Diabetologia 2022, 65, 1415–1423. [Google Scholar] [CrossRef]
  6. Saccà, S.C.; Cutolo, C.A.; Ferrari, D.; Corazza, P.; Traverso, C.E. The Eye, Oxidative Damage and Polyunsaturated Fatty Acids. Nutrients 2018, 10, 668. [Google Scholar] [CrossRef]
  7. Layé, S.; Nadjar, A.; Joffre, C.; Bazinet, R.P. Anti-Inflammatory Effects of Omega-3 Fatty Acids in the Brain: Physiological Mechanisms and Relevance to Pharmacology. Pharmacol. Rev. 2018, 70, 12–38. [Google Scholar] [CrossRef]
  8. Duan, H.; Song, W.; Zhao, J.; Yan, W. Polyunsaturated Fatty Acids (PUFAs): Sources, Digestion, Absorption, Application and Their Potential Adjunctive Effects on Visual Fatigue. Nutrients 2023, 15, 2633. [Google Scholar] [CrossRef]
  9. Swinkels, D.; Baes, M. The essential role of docosahexaenoic acid and its derivatives for retinal integrity. Pharmacol. Ther. 2023, 247, 108440. [Google Scholar] [CrossRef]
  10. Alsbirk, K.E.; Seland, J.H.; Assmus, J. Diabetic retinopathy and visual impairment in a Norwegian diabetic coast population with a high dietary intake of fish oils. An observational study. Acta Ophthalmol. 2022, 100, e532–e538. [Google Scholar] [CrossRef]
  11. Gong, Y.; Fu, Z.; Liegl, R.; Chen, J.; Hellström, A.; Smith, L.E. ω-3 and ω-6 long-chain PUFAs and their enzymatic metabolites in neovascular eye diseases. Am. J. Clin. Nutr. 2017, 106, 16–26. [Google Scholar] [CrossRef] [PubMed]
  12. Sasaki, M.; Kawasaki, R.; Rogers, S.; Man, R.E.K.; Itakura, K.; Xie, J.; Flood, V.; Tsubota, K.; Lamoureux, E.; Wang, J.J. The Associations of Dietary Intake of Polyunsaturated Fatty Acids With Diabetic Retinopathy in Well-Controlled Diabetes. Investig. Ophthalmol. Vis. Sci. 2015, 56, 7473–7479. [Google Scholar] [CrossRef]
  13. Gorusupudi, A.; Chang, F.Y.; Nelson, K.; Hageman, G.S.; Bernstein, P.S. n-3 PUFA Supplementation Alters Retinal Very-Long-Chain-PUFA Levels and Ratios in Diabetic Animal Models. Mol. Nutr. Food Res. 2019, 63, e1801058. [Google Scholar] [CrossRef] [PubMed]
  14. Chen, W.; Jump, D.B.; Grant, M.B.; Esselman, W.J.; Busik, J.V. Dyslipidemia, but not hyperglycemia, induces inflammatory adhesion molecules in human retinal vascular endothelial cells. Investig. Ophthalmol. Vis. Sci. 2003, 44, 5016–5022. [Google Scholar] [CrossRef] [PubMed]
  15. Chen, W.; Jump, D.B.; Esselman, W.J.; Busik, J.V. Inhibition of cytokine signaling in human retinal endothelial cells through modification of caveolae/lipid rafts by docosahexaenoic acid. Investig. Ophthalmol. Vis. Sci. 2007, 48, 18–26. [Google Scholar] [CrossRef]
  16. Pitale, P.M.; Gorbatyuk, M.S. Diabetic Retinopathy: From Animal Models to Cellular Signaling. Int. J. Mol. Sci. 2022, 23, 1487. [Google Scholar] [CrossRef] [PubMed]
  17. Arah, O.A. Bias Analysis for Uncontrolled Confounding in the Health Sciences. Annu. Rev. Public Health 2017, 38, 23–38. [Google Scholar] [CrossRef]
  18. Berlin, I.; Luquiens, A.; Aubin, H.J. Smoking as a confounder of the association of suicidality with serum lipid levels. J. Psychiatry Neurosci. 2016, 41, E24. [Google Scholar] [CrossRef]
  19. Schwingshackl, L.; Balduzzi, S.; Beyerbach, J.; Bröckelmann, N.; Werner, S.S.; Zähringer, J.; Nagavci, B.; Meerpohl, J.J. Evaluating agreement between bodies of evidence from randomised controlled trials and cohort studies in nutrition research: Meta-epidemiological study. BMJ 2021, 374, n1864. [Google Scholar] [CrossRef]
  20. Kappelmann, N.; Müller-Myhsok, B.; Kopf-Beck, J. Adapting the randomised controlled trial (RCT) for precision medicine: Introducing the nested-precision RCT (npRCT). Trials 2021, 22, 13. [Google Scholar] [CrossRef]
  21. Birney, E. Mendelian Randomization. Cold Spring Harb. Perspect. Med. 2022, 12, a041302. [Google Scholar] [CrossRef] [PubMed]
  22. Fu, R.; Kim, S.J. Inferring causality from observational studies: The role of instrumental variable analysis. Kidney Int. 2021, 99, 1303–1308. [Google Scholar] [CrossRef] [PubMed]
  23. Huang, Y.; Wang, J.; Yang, H.; Lin, Z.; Xu, L. Causal associations between polyunsaturated fatty acids and kidney function: A bidirectional Mendelian randomization study. Am. J. Clin. Nutr. 2023, 117, 199–206. [Google Scholar] [CrossRef] [PubMed]
  24. Davies, N.M.; Holmes, M.V.; Smith, G.D. Reading Mendelian randomisation studies: A guide, glossary, and checklist for clinicians. BMJ 2018, 362, k601. [Google Scholar] [CrossRef] [PubMed]
  25. Julkunen, H.; Cichońska, A.; Slagboom, P.E.; Würtz, P.; Nightingale Health UK Biobank Initiative. Metabolic biomarker profiling for identification of susceptibility to severe pneumonia and COVID-19 in the general population. Elife 2021, 10, e63033. [Google Scholar] [CrossRef]
  26. Kurki, M.I.; Karjalainen, J.; Palta, P.; Sipilä, T.P.; Kristiansson, K.; Donner, K.; Reeve, M.P.; Laivuori, H.; Aavikko, M.; Kaunisto, M.A.; et al. FinnGen: Unique genetic insights from combining isolated population and national health register data. medRxiv 2022, 03, 22271360. [Google Scholar]
  27. Burgess, S.; Davies, N.M.; Thompson, S.G. Bias due to participant overlap in two-sample Mendelian randomization. Genet. Epidemiol. 2016, 40, 597–608. [Google Scholar] [CrossRef]
  28. Hemani, G.; Zheng, J.; Elsworth, B.; Wade, K.H.; Haberland, V.; Baird, D.; Laurin, C.; Burgess, S.; Bowden, J.; Langdon, R.; et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018, 7, e34408. [Google Scholar] [CrossRef]
  29. Hemani, G.; Tilling, K.; Davey Smith, G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet. 2017, 13, e1007081. [Google Scholar]
  30. Gkatzionis, A.; Burgess, S.; Newcombe, P.J. Statistical methods for cis-Mendelian randomization with two-sample summary-level data. Genet. Epidemiol. 2023, 47, 3–25. [Google Scholar] [CrossRef]
  31. Skrivankova, V.W.; Richmond, R.C.; Woolf, B.A.R.; Yarmolinsky, J.; Davies, N.M.; Swanson, S.A.; VanderWeele, T.J.; Higgins, P.T.; Timpson, N.J.; Dimou, N.K.; et al. Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement. JAMA 2021, 326, 1614–1621. [Google Scholar] [CrossRef] [PubMed]
  32. Bowden, J.; Smith, G.D.; Haycock, P.C.; Burgess, S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet. Epidemiol. 2016, 40, 304–314. [Google Scholar] [CrossRef] [PubMed]
  33. Bowden, J.; Davey Smith, G.; Burgess, S. Mendelian randomization with invalid instruments: Effect estimation and bias detection through Egger regression. Int. J. Epidemiol. 2015, 44, 512–525. [Google Scholar] [CrossRef] [PubMed]
  34. Verbanck, M.; Chen, C.-Y.; Neale, B.; Do, R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat. Genet. 2018, 50, 693–698. [Google Scholar] [CrossRef]
  35. Burgess, S.; Bowden, J.; Fall, T.; Ingelsson, E.; Thompson, S.G. Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants. Epidemiology 2017, 28, 30–42. [Google Scholar] [CrossRef]
  36. Emdin, C.A.; Khera, A.V.; Kathiresan, S. Mendelian Randomization. JAMA 2017, 318, 1925–1926. [Google Scholar] [CrossRef]
  37. Eynard, A.R.; Repossi, G. Role of ω3 polyunsaturated fatty acids in diabetic retinopathy: A morphological and metabolically cross talk among blood retina barriers damage, autoimmunity and chronic inflammation. Lipids Health Dis. 2019, 18, 114. [Google Scholar] [CrossRef]
  38. Li, J.-S.; Wang, T.; Zuo, J.-J.; Guo, C.-N.; Peng, F.; Zhao, S.-Z.; Li, H.-H.; Hou, X.-Q.; Lan, Y.; Wei, Y.-P.; et al. Association of n-6 PUFAs with the risk of diabetic retinopathy in diabetic patients. Endocr. Connect. 2020, 9, 1191–1201. [Google Scholar] [CrossRef]
  39. Sala-Vila, A.; Díaz-López, A.; Valls-Pedret, C.; Cofán, M.; García-Layana, A.; Lamuela-Raventós, R.M.; Castaner, O.; Zanon-Moreno, V.; Martinez-Gonzalez, M.A.; Toledo, E.; et al. Dietary Marine ω-3 Fatty Acids and Incident Sight-Threatening Retinopathy in Middle-Aged and Older Individuals with Type 2 Diabetes: Prospective Investigation From the PREDIMED Trial. JAMA Ophthalmol. 2016, 134, 1142–1149. [Google Scholar] [CrossRef]
  40. Tikhonenko, M.; Lydic, T.A.; Opreanu, M.; Calzi, S.L.; Bozack, S.; McSorley, K.M.; Sochacki, A.L.; Faber, M.S.; Hazra, S.; Duclos, S.; et al. N-3 polyunsaturated Fatty acids prevent diabetic retinopathy by inhibition of retinal vascular damage and enhanced endothelial progenitor cell reparative function. PLoS ONE 2013, 8, e55177. [Google Scholar] [CrossRef]
  41. Yanai, R.; Mulki, L.; Hasegawa, E.; Takeuchi, K.; Sweigard, H.; Suzuki, J.; Gaissert, P.; Vavvas, D.G.; Sonoda, K.-H.; Rothe, M.; et al. Cytochrome P450-generated metabolites derived from ω-3 fatty acids attenuate neovascularization. Proc. Natl. Acad. Sci. USA 2014, 111, 9603–9608. [Google Scholar] [CrossRef] [PubMed]
  42. Bühler, A.D.; Bucher, F.; Augustynik, M.; Wöhrl, J.; Martin, G.; Schlunck, G.; Agostini, H.; Böhringer, D.; Pütz, G.; Stahl, A. Systemic confounders affecting serum measurements of omega-3 and -6 polyunsaturated fatty acids in patients with retinal disease. BMC Ophthalmol. 2016, 16, 159. [Google Scholar] [CrossRef] [PubMed]
  43. Hammes, H.P.; Weiss, A.; Führer, D.; Krämer, H.J.; Papavassilis, C.; Grimminger, F. Acceleration of experimental diabetic retinopathy in the rat by omega-3 fatty acids. Diabetologia 1996, 39, 251–255. [Google Scholar] [CrossRef] [PubMed]
  44. Chen, C.; Wang, P.; Zhang, R.-D.; Fang, Y.; Jiang, L.-Q.; Fang, X.; Zhao, Y.; Wang, D.-G.; Ni, J.; Pan, H.-F. Mendelian randomization as a tool to gain insights into the mosaic causes of autoimmune diseases. Autoimmun. Rev. 2022, 21, 103210. [Google Scholar] [CrossRef] [PubMed]
  45. Shahidi, F.; Ambigaipalan, P. Omega-3 Polyunsaturated Fatty Acids and Their Health Benefits. Annu. Rev. Food Sci. Technol. 2018, 9, 345–381. [Google Scholar] [CrossRef] [PubMed]
  46. Connor, K.M.; SanGiovanni, J.P.; Lofqvist, C.; Aderman, C.M.; Chen, J.; Higuchi, A.; Hong, S.; Pravda, E.A.; Majchrzak, S.; Carper, D.; et al. Increased dietary intake of omega-3-polyunsaturated fatty acids reduces pathological retinal angiogenesis. Nat. Med. 2007, 13, 868–873. [Google Scholar] [CrossRef]
  47. Busik, J.V. Lipid metabolism dysregulation in diabetic retinopathy. J. Lipid Res. 2021, 62, 100017. [Google Scholar] [CrossRef]
  48. Aranda, J.V.; Qu, J.; Valencia, G.B.; Beharry, K.D. Pharmacologic interventions for the prevention and treatment of retinopathy of prematurity. Semin. Perinatol. 2019, 43, 360–366. [Google Scholar] [CrossRef]
  49. Shen, J.; Shen, S.; Das, U.N.; Xu, G. Effect of essential fatty acids on glucose-induced cytotoxicity to retinal vascular endothelial cells. Lipids Health Dis. 2012, 11, 90. [Google Scholar] [CrossRef]
  50. D’Angelo, S.; Motti, M.L.; Meccariello, R. ω-3 and ω-6 Polyunsaturated Fatty Acids, Obesity and Cancer. Nutrients 2020, 12, 2751. [Google Scholar] [CrossRef]
  51. Liput, K.P.; Lepczyński, A.; Ogłuszka, M.; Nawrocka, A.; Poławska, E.; Grzesiak, A.; Ślaska, B.; Pareek, C.S.; Czarnik, U.; Pierzchała, M. Effects of Dietary n-3 and n-6 Polyunsaturated Fatty Acids in Inflammation and Cancerogenesis. Int. J. Mol. Sci. 2021, 22, 6965. [Google Scholar] [CrossRef]
  52. Tortosa-Caparrós, E.; Navas-Carrillo, D.; Marín, F.; Orenes-Piñero, E. Anti-inflammatory effects of omega 3 and omega 6 polyunsaturated fatty acids in cardiovascular disease and metabolic syndrome. Crit. Rev. Food Sci. Nutr. 2017, 57, 3421–3429. [Google Scholar] [CrossRef] [PubMed]
  53. Thies, F.; Miles, E.A.; Nebe-Von-Caron, G.; Powell, J.R.; Hurst, T.L.; Newsholme, E.A.; Calder, P.C. Influence of dietary supplementation with long-chain n-3 or n-6 polyunsaturated fatty acids on blood inflammatory cell populations and functions and on plasma soluble adhesion molecules in healthy adults. Lipids 2001, 36, 1183–1193. [Google Scholar] [CrossRef] [PubMed]
  54. Kakutani, S.; Ishikura, Y.; Tateishi, N.; Horikawa, C.; Tokuda, H.; Kontani, M.; Kawashima, H.; Sakakibara, Y.; Kiso, Y.; Shibata, H.; et al. Supplementation of arachidonic acid-enriched oil increases arachidonic acid contents in plasma phospholipids, but does not increase their metabolites and clinical parameters in Japanese healthy elderly individuals: A randomized controlled study. Lipids Health Dis. 2011, 10, 241. [Google Scholar] [CrossRef] [PubMed]
  55. Djuricic, I.; Calder, P.C. Beneficial Outcomes of Omega-6 and Omega-3 Polyunsaturated Fatty Acids on Human Health: An Update for 2021. Nutrients 2021, 13, 2421. [Google Scholar] [CrossRef]
  56. Fu, Z.; Yan, W.; Chen, C.T.; Nilsson, A.K.; Bull, E.; Allen, W.; Yang, J.; Ko, M.; SanGiovanni, J.P.; Akula, J.D.; et al. Omega-3/Omega-6 Long-Chain Fatty Acid Imbalance in Phase I Retinopathy of Prematurity. Nutrients 2022, 14, 1333. [Google Scholar] [CrossRef]
  57. Wang, K.; Zhong, Y.; Yang, F.; Hu, C.; Liu, X.; Zhu, Y.; Yao, K. Causal Effects of N-6 Polyunsaturated Fatty Acids on Age-related Macular Degeneration: A Mendelian Randomization Study. J. Clin. Endocrinol. Metab. 2021, 106, e3565–e3572. [Google Scholar] [CrossRef]
  58. Dyall, S.C.; Balas, L.; Bazan, N.G.; Brenna, J.T.; Chiang, N.; Souza, F.d.C.; Dalli, J.; Durand, T.; Galano, J.-M.; Lein, P.J.; et al. Polyunsaturated fatty acids and fatty acid-derived lipid mediators: Recent advances in the understanding of their biosynthesis, structures, and functions. Prog. Lipid Res. 2022, 86, 101165. [Google Scholar] [CrossRef]
  59. Simonetto, M.; Infante, M.; Sacco, R.L.; Rundek, T.; Della-Morte, D. A Novel Anti-Inflammatory Role of Omega-3 PUFAs in Prevention and Treatment of Atherosclerosis and Vascular Cognitive Impairment and Dementia. Nutrients 2019, 11, 2279. [Google Scholar] [CrossRef]
  60. Custers Emma, E.M.; Kiliaan Amanda, J. Dietary lipids from body to brain. Prog. Lipid Res. 2022, 85, 101144. [Google Scholar] [CrossRef]
  61. Hartwig, F.P.; Borges, M.C.; Horta, B.L.; Bowden, J.; Smith, G.D. Inflammatory Biomarkers and Risk of Schizophrenia: A 2-Sample Mendelian Randomization Study. JAMA Psychiatry 2017, 74, 1226–1233. [Google Scholar] [CrossRef] [PubMed]
  62. Skrivankova, V.W.; Richmond, R.C.; Woolf, B.A.R.; Davies, N.M.; Swanson, S.A.; VanderWeele, T.J.; Timpson, N.J.; Higgins, J.P.T.; Dimou, N.; Langenberg, C.; et al. Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomisation (STROBE-MR): Explanation and Elaboration. BMJ 2021, 375, n2233. [Google Scholar] [CrossRef] [PubMed]
Nutrients 15 04208 g001
Figure 1. A study flame chart of the MR study revealing the causal relationship of PUFAs and DR. (A) The research direction in this study; (B) The principle of MR; (C) The process of MR Analysis. MR, Mendelian randomization; PUFAs, polyunsaturated fatty acids; FAw3, omega-3 fatty acids; FAw6, omega-6 fatty acids; DR, diabetic retinopathy; BDR, background diabetic retinopathy; SNPDR, severe non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy. Error symbol indicate no correlation; Asterisk indicates the most important.
Figure 1. A study flame chart of the MR study revealing the causal relationship of PUFAs and DR. (A) The research direction in this study; (B) The principle of MR; (C) The process of MR Analysis. MR, Mendelian randomization; PUFAs, polyunsaturated fatty acids; FAw3, omega-3 fatty acids; FAw6, omega-6 fatty acids; DR, diabetic retinopathy; BDR, background diabetic retinopathy; SNPDR, severe non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy. Error symbol indicate no correlation; Asterisk indicates the most important.
Nutrients 15 04208 g001
Nutrients 15 04208 g002
Figure 2. Scatter plots for MR analyses of the causal effects of PUFAs on DR. (Light blue represents IVW, dark blue represents MR-Egger regression, and green represents WM methods.) (A): ADR on total PUFAs; (B): total PUFAs on ADR; (C): total PUFAs on BDR; (D): total PUFAs on PDR; (E): FAw3 on PDR; (F): FAw6 on BDR; (G): FAw6 on PDR. MR, Mendelian randomization; PUFAs, polyunsaturated fatty acids; DR, diabetic retinopathy; ADR, any diabetic retinopathy; BDR, background diabetic retinopathy; PDR, proliferative diabetic retinopathy; FAw3, omega-3 fatty acids; FAw6, omega-6 fatty acids.
Figure 2. Scatter plots for MR analyses of the causal effects of PUFAs on DR. (Light blue represents IVW, dark blue represents MR-Egger regression, and green represents WM methods.) (A): ADR on total PUFAs; (B): total PUFAs on ADR; (C): total PUFAs on BDR; (D): total PUFAs on PDR; (E): FAw3 on PDR; (F): FAw6 on BDR; (G): FAw6 on PDR. MR, Mendelian randomization; PUFAs, polyunsaturated fatty acids; DR, diabetic retinopathy; ADR, any diabetic retinopathy; BDR, background diabetic retinopathy; PDR, proliferative diabetic retinopathy; FAw3, omega-3 fatty acids; FAw6, omega-6 fatty acids.
Nutrients 15 04208 g002
Nutrients 15 04208 g003
Figure 3. Study of association of PUFAs with DR using MR. SNP, single-nucleotide polymorphism; MR, Mendelian randomization; OR, odds ratio; CI, confidence interval; P_Heterogeneity, p-value for heterogeneity using Cochran Q test; P_Pleiotropy, p-value for MR-Egger intercept; PUFAs, polyunsaturated fatty acids; FAw3, omega-3 fatty acids; FAw6, omega-6 fatty acids; DR, diabetic retinopathy; BDR, background diabetic retinopathy; SNPDR, severe non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy.
Figure 3. Study of association of PUFAs with DR using MR. SNP, single-nucleotide polymorphism; MR, Mendelian randomization; OR, odds ratio; CI, confidence interval; P_Heterogeneity, p-value for heterogeneity using Cochran Q test; P_Pleiotropy, p-value for MR-Egger intercept; PUFAs, polyunsaturated fatty acids; FAw3, omega-3 fatty acids; FAw6, omega-6 fatty acids; DR, diabetic retinopathy; BDR, background diabetic retinopathy; SNPDR, severe non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy.
Nutrients 15 04208 g003
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ren, S.; Xue, C.; Xu, M.; Li, X. Mendelian Randomization Analysis Reveals Causal Effects of Polyunsaturated Fatty Acids on Subtypes of Diabetic Retinopathy Risk. Nutrients 2023, 15, 4208. https://doi.org/10.3390/nu15194208

AMA Style

Ren S, Xue C, Xu M, Li X. Mendelian Randomization Analysis Reveals Causal Effects of Polyunsaturated Fatty Acids on Subtypes of Diabetic Retinopathy Risk. Nutrients. 2023; 15(19):4208. https://doi.org/10.3390/nu15194208

Chicago/Turabian Style

Ren, Shaojie, Chen Xue, Manhong Xu, and Xiaorong Li. 2023. "Mendelian Randomization Analysis Reveals Causal Effects of Polyunsaturated Fatty Acids on Subtypes of Diabetic Retinopathy Risk" Nutrients 15, no. 19: 4208. https://doi.org/10.3390/nu15194208

APA Style

Ren, S., Xue, C., Xu, M., & Li, X. (2023). Mendelian Randomization Analysis Reveals Causal Effects of Polyunsaturated Fatty Acids on Subtypes of Diabetic Retinopathy Risk. Nutrients, 15(19), 4208. https://doi.org/10.3390/nu15194208

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop