Lung function is a critical predictor of various lung diseases and overall health in the general population(Reference Burney and Hooper1). The assessment of lung function is typically performed by utilising measures such as forced expiratory volume in one second (FEV1), forced vital capacity (FVC) and the ratio of FEV1 to FVC (FEV1:FVC)(Reference Agustí, Noell and Brugada2). Environmental exposures, such as smoking, air pollution and occupational hazards, have been recognised as major contributors to the deterioration of lung function(Reference Burchfiel, Marcus and Curb3–Reference Sunyer5). This decline can culminate in the clinical diagnosis of chronic obstructive pulmonary disease, which poses growing social and economic challenges as the third leading cause of morbidity globally(Reference Patel, Patel and Singh6). It is thus of great clinical and public health interest to improve lung function.
In recent years, there has been a growing interest in the health benefits of n-3 PUFA. EPA and DHA, the two major essential n-3 PUFA, have been proven to have anti-inflammatory, antioxidant and immunoregulatory properties(Reference Mickleborough and Rundell7). The potential of n-3 PUFA for the improvement of lung health is an emerging field of research. Nevertheless, the results from epidemiological studies have not been entirely consistent. While one observational cohort study found a positive correlation between dietary intake of n-3 PUFA and lung function(Reference Ng, Niti and Yap8), a cross-sectional study from a representative sample of Dutch adults showed no improvement in lung function with dietary n-3 PUFA intake(Reference McKeever, Lewis and Cassano9). Therefore, the association between n-3 PUFA and lung function remains equivocal and needs further investigation.
Of note, dietary intake of n-3 PUFA is often associated with a variety of clinical and social factors, making it challenging to determine the causal effects of diets on various outcomes. Additionally, the observed association between n-3 PUFA intake and lung function may also be influenced by confounders, such as smoking and occupational exposures, which cannot be adequately controlled for using observational study designs(Reference Kim, Moll and Qiao10). Mendelian randomisation (MR) has emerged as a powerful approach to address the limitations of observational studies by using genetic variants as instrumental variables (IV) to estimate the potential causal effect of exposures (e.g. serum n-3 PUFA) on outcomes (e.g. lung function)(Reference Davey Smith and Ebrahim11). Given genetic variants are randomly allocated at conception, MR studies are less susceptible to biases of confounding, residual bias and reverse causality(Reference Bowden, Davey Smith and Burgess12). Hence, MR studies, which are similar in concept to randomised controlled trials, can substantially improve causal inference from observational investigations(Reference Haycock, Burgess and Wade13). Our study adds to the current literature by employing MR to address these potential confounders that have challenged previous observational research. This is especially pertinent in light of previous research that has produced varying results. With this approach, our study provides new insights into the causal relationships between n-3 PUFA intake and lung function.
Therefore, in the present study, we performed a cross-sectional study utilising data from the National Health and Nutrition Examination Survey (NHANES), to assess the observational association between n-3 PUFA intake and lung function. Subsequently, we conducted the two-sample MR analysis to further estimate the causal relationship between circulating n-3 PUFA concentration and lung function.
Methods
Cross-Sectional study
Study population
NHANES is a continuous, stratified, multistage sampling, cross-sectional study designed to evaluate the health and nutritional status of the non-institutionalised civilian population of the USA(14). This study is performed by the National Center for Health Statistics (NCHS), a branch of the Centers for Disease Control and Prevention (CDC) and is approved by the NCHS Research Ethics Review Board(15). All participants provided written informed consent. The data collected from the study are publicly available through the CDC website (http://www.cdc.gov/nchs/nhanes.htm). For the present study, publicly accessible data from three NHANES cycles (2007–2008, 2009–2010 and 2011–2012) were analysed, as lung function data were only available during these cycles. A total of 29 353 participants were included in the 2007–2012 cycles, with 10 693 participants aged 20 years or older without pregnancy being screened in the analysis. Participants who had missing information on dietary n-3 PUFA intake (n 1092), had incomplete lung function test results or did not meet the American Thoracic Society data collection standards (quality grades C–F, n 4780) were excluded. Additionally, participants with missing data on BMI, marital status, drinking, smoking, hypertension, diabetes, emphysema, chronic bronchitis and asthma were also excluded (n 1653). In examining the sociodemographic characteristics of the excluded sample in comparison with the analytic sample, we observed a consistent distribution across key variables. Both samples exhibited similar profiles in terms of age, sex, race, educational attainment and income levels. After all exclusions, the final analysis sample consisted of 9378 participants (online Supplementary Fig. S1).
Lung function measurement
Spirometry was available in NHANES 2007–2012. Participants who were receiving supplemental oxygen or exhibited symptoms such as chest pain, difficulties with forceful expiration, recent surgical procedures in the eye, chest, or abdomen, recent heart attack, stroke, tuberculosis exposure, or coughing up of blood, or a history of detached retina, collapsed lung, or aneurysm were not eligible for spirometry testing. Eligible participants underwent spirometry testing using an Ohio 822/827 dry-rolling seal volume spirometer, in accordance with the guidelines established by the American Thoracic Society and European Respiratory Society(Reference Miller, Hankinson and Brusasco16). The best FEV1, FVC and the FEV1:FVC ratio were selected for data analysis, which are commonly used clinical indicators of lung function. In addition to the primary lung function indicators, peak expiratory flow (PEF) and forced expiratory volume in 25 %–75 % of FVC (FEV25 %–75 %) were also measured. The percent predicted lung function (FEV1 and FVC) were calculated using Global Lung Initiative equations that account for age, sex, race and height(Reference Hankinson, Odencrantz and Fedan17).
Dietary intake assessment
The daily intake of n-3 PUFA in the NHANES survey was assessed through two 24-h dietary recall interviews included in the NHANES Individual Foods files. The first interview was conducted in-person by trained interviewers in the Mobile Examination Center (MEC), while the second was a follow-up telephone interview conducted from the home office 3–10 d later, but not on the same day of the week as the first interview(18). In cases where a second day of dietary recall data was available, only data from the first dietary interview were used. The total and subtypes of n-3 PUFA intake were calculated as an average of the 2 d of dietary recalls, which were then adjusted for body weight (mg/kg/d) and divided into quartiles. The n-3 PUFA analysed in this study included octadecatrienoic acid (linolenic, ALA, 18:3n-3), octadecatetraenoic acid (stearidonic, SDA, 18:4n-3), EPA (20:5n-3), docosapentaenoic acid (DPA, 22:5n-3) and DHA (22:6n-3).
Covariates of interest
The following variables were included as potential covariates in this study: age (categorised into 20–39 years, 40–59 years and ≥ 60 years), sex (male and female), ethnicity (Mexican American, other Hispanic, Non-Hispanic White, Non-Hispanic Black and other race), educational level (below high school, high school graduate/GED, and some college or above), family poverty:income ratio (< 1·0, 1·0 to < 3·0 and ≥ 3·0), marital status (married/living with a partner, and windowed/divorced/separated/never married), smoking status (never smoker, former smoker and current smoker), drinking status (none, moderate and heavy), BMI (< 25 kg/m2, 25 to < 30 kg/m2 and ≥ 30 kg/m2), emphysema, chronic bronchitis, asthma, hypertension, diabetes, total energy intake, protein intake, dietary fibre intake, fat intake, cholesterol intake, saturated fat, MUFA and PUFA. Smoking status was defined by CDC and NCHS as follows: never (has never smoked or has smoked less than 100 cigarettes in their lifetime), current smoker (has smoked 100 cigarettes in their lifetime and is currently smoking) or former (has smoked more than 100 cigarettes in their lifetime but has quit smoking)(Reference Navaneethan, Mandayam and Arrigain19). Drinkers were defined as participants who consumed ≥ 12 alcoholic drinks in a year. Moderate drinkers were defined as those who consumed < 1 drink per d (for female) or < 2 drinks per d (for male), while heavy drinkers were defined as those who consumed ≥ 1 drink per d (for female) or ≥ 2 drinks per d (for male)(Reference Chen, Du and Blumberg20). Participants with physician-diagnosed diabetes, use of oral antidiabetic agents or insulin injections, fasting plasma glucose level of 7·0 mmol/l or higher, or glycosylated Hb (HbA1c) level of 6·5 % or higher were considered to have diabetes mellitus(Reference Menke, Casagrande and Geiss21). Hypertension was defined as average systolic blood pressure ≥ 140 mmHg, average diastolic blood pressure ≥ 90 mmHg, a self-reported physician diagnosis of hypertension or use of anti-hypertensive medications(Reference Whelton, Carey and Aronow22). Emphysema was defined as a positive answer to the question: ‘Has a doctor or other health professional ever told you that you have emphysema?’. Current chronic bronchitis was defined as a positive answer to both questions ‘Has a doctor or other health professional ever told you that you have chronic bronchitis?’ and ‘Do you still have chronic bronchitis?’.
Statistical analysis
In this study, the normality of continuous variables was evaluated using the Kolmogorov–Smirnov test and histogram; if a variable was found to be normally distributed, it was described by the mean and standard error, otherwise, it was expressed as the median and interquartile range. Categorical data were presented in terms of counts and percentages. The participants were divided into four groups according to quartiles of n-3 PUFA intake, with the lowest quartile (Q1) serving as the reference category. The comparison of continuous variables was performed using either one-way ANOVA test if the variance was homogeneous, or Kruskal–Wallis non-parametric test if not. Meanwhile, χ 2 tests were used to compare the percentages of categorical variables. Population-weighted univariate and multivariable linear regression models were used to evaluate the association between lung function measurements and dietary n-3 PUFA intake, with the results presented as β-coefficients along with 95 % CI. The intake of n-3 PUFA was divided into quartiles (quartile 1: < 25th percentile, quartile 2: ≥ 25th to 50th percentile, quartile 3: ≥ 50th to 75th percentile and quartile 4: ≥ 75th percentile) according to their distributions. Univariate regression analysis (model 1) examined the association without adjustments. The multivariable regression analyses were performed with adjustments for potential confounders including age, sex, race, educational level, poverty:income ratio, BMI, drinking, smoking, diabetes and hypertension (model 2). Selection of these covariates for model 2 was based on their statistical significance with lung function from univariate analysis. A stepwise regression then refined these covariates, retaining those with significant contributions. To ensure model robustness, we excluded covariates with high multicollinearity. A two-sided P-value of less than 0·05 was considered statistically significant. All statistical analyses were performed using Stata 15.0 software (Stata Corporation), taking into account the weighted data and complex design of the NHANES sample.
Mendelian randomisation study
Study design
This study used a two-sample MR study design to estimate the potential causality between exposures and outcomes, using genetic variants as IV. Briefly, we analysed summary statistics from multiple genome-wide association studies (GWAS) to identify SNP that were associated with serum levels of n-3 PUFA and lung function. The identified SNP were then combined to evaluate their causal relationship. The validity of MR relies on the following assumptions(Reference VanderWeele, Tchetgen and Cornelis23), as demonstrated in online Supplementary Fig. S2: (1) the genetic IV are associated with exposure factor; (2) IV are independent of known and unknown confounders; (3) IV are associated with outcome only via exposure factors. This study's analysis is a secondary examination of publicly available data, not involving new human or animal research. All utilised GWAS datasets are openly accessible, negating the need for ethical approval or informed consent.
Selection of genetic instruments for n-3 PUFA
In the MR analysis, we extracted SNP strongly associated with circulating n-3 PUFA concentration as instruments from the most recent and largest available GWAS study, which contained 114 999 participants of European ancestry from the UK Biobank(Reference Borges, Haycock and Zheng24). To select SNP as IV, we screened the genome-wide significant SNP (P < 5 × 10–8) that showed a strong relationship with exposure factors. We performed a clustering process with a pruning threshold of R2 < 0·001 and a clumping distance of 10 000 kb to eliminate linkage disequilibrium between genetic variants. Additionally, we excluded palindrome SNP to avoid potential strand ambiguity and set a minimum allele frequency of 0·01 without the use of SNP proxies. The total coefficient of determination (R2) and mean F statistics across selected SNP were estimated to judge the strength of the selected IV, with an F statistic threshold of 10 indicating sufficient strength for MR analyses, as previously described(Reference Park, Wacholder and Gail25,Reference Burgess and Thompson26) .
Genetic summary data for lung function
Summary statistics of lung function were extracted from the largest available GWAS meta-analysis of UK Biobank, which consists of 321 047 individuals(Reference Shrine, Guyatt and Erzurumluoglu27). Lung function measurements in this study included FEV1, FVC and FEV1:FVC. To ensure data quality, the inclusion criteria were limited to participants who had at least two measurements of FEV1 and FVC, and full details for age, sex, height, ever-smoking status, and spirometry method used were included. In addition, given the important roles of airflow limitation in diagnosing chronic obstructive pulmonary disease, we also utilised summary-level data obtained from a GWAS that analysed the risk of FEV1:FVC < 0·7 in individuals of European descent (n 353 315)(Reference Higbee, Granell and Hemani28).
Statistical analysis
We used inverse variance weighted (IVW) as the primary MR analysis approach to evaluate the causal association of genetically predicted circulating n-3 PUFA concentration and lung function(Reference Thompson, Minelli and Abrams29). We also performed multiple sensitivity analyses, including MR-Egger, weighted median and weighted mode to validate the robustness of the IVW results(Reference Burgess, Butterworth and Thompson30). In the current study, an association was deemed statistically significant when the IVW achieved a significance level of P < 0·05, and the results of other analytical methods pointed in the same direction as that of the IVW results. In addition, we used MR-Egger regression to assess the presence of potential pleiotropic effects of IV(Reference Bowden, Davey Smith and Burgess12). The absence of horizontal pleiotropy can be inferred when the MR-Egger intercept term was close to 0 and P > 0·05. The heterogeneity of SNP was examined using Cochran’s Q test in both MR-Egger and IVW methods(Reference Qian, Ye and Huang31). A leave-one-out sensitivity analysis was also adopted by removing each SNP one by one to evaluate the influence of each SNP on the overall results obtained using MR-Egger and IVW methods(Reference Zhou, Zhang and Liu32). All statistical analyses were conducted utilising R version 4.0.3 and the TwoSampleMR R package version 0.5.5(Reference Hemani, Zheng and Elsworth33).
Results
In Table 1, we present the characteristics of the study participants from the US NHANES sample. Overall, 49·5 % of the eligible participants were men and 50·5 % female, with a mean age of 46·0 years and a mean BMI of 29·08 kg/m2. Of the participants, 55·6 % reported never having smoked, 23·8 % were former smokers and 21·4 % were current smokers. The mean levels of FEV1, FVC, FEV1:FVC, PEF and FEV25 %–75 % were 3952·0 ml, 3096·1 ml, 78·3 %, 8140·6 ml/s and 2947·7 ml/s, respectively. Participants with higher n-3 PUFA intake were found to be predominantly male, younger, non-Hispanic White, with higher family income and education level, as well as lower BMI and CRP level (P < 0·001). Conversely, individuals with lower n-3 PUFA intake tended to be smokers, drinkers, and patients with hypertension and diabetes. In addition, FEV1, FVC, PEF and FEV25 %–75 % were significantly higher in participants with higher n-3 PUFA intake compared with those with lower n-3 PUFA intake.
Table 1. Baseline characteristics of study participants according to quartiles of dietary n-3 PUFA intake in the NHANES 2007–2012
NHANES, National Health and Nutrition Examination Survey; IQR, interquartile range; ALA, α-linolenic acid; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; PEF, peak expiratory flow; %, sample-weighted percentages.
Values were presented as mean ± se or median ± IQR for continuous variables and n (%) for categorical variables. All P-values are statistically significant (P < 0·05).
Table 2 displays the findings of a linear regression analysis investigating the relationship between dietary n-3 PUFA intake and lung function measurements. In the univariate analysis of the entire study population (model 1), a positive association was found between n-3 PUFA intakes and lung function. However, after adjusting for relevant confounders (model 2), participants in the highest quartile intake of n-3 PUFA showed only higher FVC and percent predicted FVC compared with the lowest quartile of intake (43·67 ml and 0·86 % for FVC and percent predicted FVC, respectively, P = 0·009 and 0·04). There was no statistically significant association identified between the intake of n-3 PUFA and FEV1 or percent predicted FEV1. In the stratified analysis by smoking status, including current, former, and never smokers, there was no statistically significant association identified between the intake of n-3 PUFA and lung function (online Supplementary Table S1). Upon excluding ALA from the n-3 PUFA intake, the association between n-3 PUFA intake and lung function measurements remained consistent with the non-significant associations observed in the initial models, suggesting that ALA does not drive the observed lack of association within our study population (online Supplementary Table S2).
Table 2. Association between n-3 PUFA intake and lung function among participants in NHANES 2007–2012
NHANES, National Health and Nutrition Examination Survey; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; β, partial regression coefficient. Model 1 is the univariate model that assesses the association without adjustments. Model 2 adjusts for potential confounders including age, sex, race, educational level, poverty:income ratio, BMI, drinking, smoking, diabetes and hypertension. Values in bold indicate statistical significance (P< 0.05).
After harmonising the SNP effects, thirty-nine SNP associated with circulating n-3 PUFA concentration were used as IV in our two-sample MR analysis, as shown in online Supplementary Table S3. Detailed summary information of these instruments is presented in Table 3. The mean F-statistic for n-3 PUFA was above 200, considerable weak instrument bias would not be expected. There was no genetic evidence of a causal relationship between circulating n-3 PUFA concentration and lung function using IVW method; MR-Egger regression, weighted median, weighted mode and simple mode methods presented similar results (Table 4). The scatter plots and forest plots of these MR results are presented in Fig. 1 and online Supplementary Fig. S3, respectively. There was evidence for heterogeneity existed, measured by Cochran’s Q test (P < 0·001) based on both IVW and MR-Egger methods (Table 4). Hence, a multiplicative random effects model (inverse variance-weighted regression) was employed to reassess the causal effects, and the findings were consistent (FEV1, P = 0·5006; FVC, P = 0·9115; FEV1:FVC, P = 0·07359; the risk of FEV1:FVC < 0·7, P = 0·2932). The MR-Egger intercept did not deviate significantly from 0 (P > 0·05), indicating no evidence of horizontal pleiotropy (Table 4). In addition, the results of the leave-one-out sensitivity analysis showed that the observed causal association was relatively credible by removing any SNP (online Supplementary Fig. S4).
Table 3. Detailed information and datasets of exposure or outcome used in the present study
PMID, PubMed Unique Identifier; GWAS, Genome-Wide Association Study; nSNP, number of SNP; FEV1, forced expiratory volume in one second; FVC, forced vital capacity.
Table 4. Summary on MR results of circulating n-3 PUFA concentration on lung function
MR, Mendelian randomization; nSNP, number of SNP; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; IVW, Inverse variance weighted.
Fig. 1. Scatter plot for MR analyses of causal associations between each circulating n-3 PUFA concentration SNP and lung function. The slope of each line represents the causal association and each approach has a different line. FEV1, forced expiratory volume in one second; FVC, forced vital capacity; MR, Mendelian randomisation.
Discussion
In this study, we sought to investigate the association between n-3 PUFA intake and lung function by conducting observational analyses based on the NHANES data and then estimate the potential causal relationship between circulating n-3 PUFA concentration and lung function through a two-sample MR analysis. Our findings revealed that n-3 PUFA intake was not associated with lung function, and that elevated circulating n-3 PUFA concentration, as genetically predicted, did not result in improvement in lung function. These results suggest that increasing n-3 PUFA intake to elevate circulating n-3 PUFA concentration is unlikely to provide a clinical benefit for enhancing lung function. Our study stands out as the first to utilise both cross-sectional and MR analysis to explore the relationship between n-3 PUFA and lung function, providing a unique contribution to the field. The incorporation of NHANES data and MR adds a novel dimension to the current understanding, particularly in a US context where dietary intake is often suboptimal.
n-3 PUFA are considered as essential nutrients and primarily obtained from exogenous sources, such as seafood, due to the low efficiency of endogenous synthesis from precursors(Reference Wei, Wu and Chen34). The potent anti-inflammatory and antioxidant properties of n-3 PUFA have been demonstrated to have beneficial effects in several chronic inflammatory diseases, such as CVD, diabetes, rheumatoid arthritis and even cancer(Reference Simopoulos35). There is increasing evidence indicating that n-3 PUFA may have a protective effect on the lungs against the detrimental impacts of chronic inflammation and oxidative stress(Reference Li, Chen and Shi36). Some studies have reported that supplementation with n-3 PUFA can decrease inflammatory markers and improve asthma symptoms(Reference Farjadian, Moghtaderi and Kalani37,Reference Mickleborough, Murray and Ionescu38) . A study conducted in the USA with a sample size of 8960 found that intake of n-3 PUFA through fish consumption protect smokers from chronic obstructive pulmonary disease and deterioration of lung function(Reference Shahar, Folsom and Melnick39). The most plausible explanation for these observations is that n-3 PUFA could modulate inflammatory processes, which may have a therapeutic effect in the pathophysiology of chronic inflammatory lung diseases(Reference Tong, Zhang and Shen40). Despite these promising findings, there remains a paucity of epidemiological data on the direct relationship between n-3 PUFA and lung function.
A large cross-sectional cohort study from Honolulu Heart Program showed evidence of a protective effect of high n-3 PUFA intake on lung function among smokers(Reference Sharp, Rodriguez and Shahar41). This was one of the earliest studies conducted on a population of Japanese-American men in 1994. Higher n-3 PUFA intake was associated with improved lung function in current or former smokers according to the Atherosclerosis Risk in Communities (ARIC) study(Reference Shahar, Folsom and Melnick42). Similar results were observed in a nutritional epidemiological study conducted in the Lovelace Smokers cohort, which found that DPA was positively correlated with better average FEV1 volumes and reduced age-related FEV1 decline(Reference Leng, Picchi and Tesfaigzi43). The potential benefits of n-3 PUFA on lung function are biologically plausible, but not all observational studies have demonstrated that n-3 PUFA intake has a positive effect on lung function. The available information regarding the association between n-3 PUFA and lung function is limited and controversial. For instance, a cross-sectional study showed no association between dietary intake of n-3 PUFA and FEV1 (Reference McKeever, Lewis and Cassano9). A 25-year longitudinal study conducted in the Netherlands also failed to detect a protective role of n-3 PUFA on lung function(Reference Miedema, Feskens and Heederik44). These findings were generally consistent with our results from the nationally representative NHANES, suggesting n-3 PUFA may have no impact on lung function. This aligns with the understanding that the standard American diet is relatively deficient in n-3 PUFA, particularly EPA and DHA, which are essential for mitigating inflammation and oxidative stress. This lack of association may highlight the possibility that the intake levels in the USA fall below a threshold necessary for a detectable impact on lung function. The absence of a significant association in our findings challenges the prevailing notion that n-3 PUFA intake is universally beneficial for lung function, suggesting a need to reassess dietary guidelines and the potential for individualised nutritional recommendations based on genetic makeup and metabolic capacity. Notwithstanding this overall trend, significant results from Table 2, specifically in model 2 comparisons of the highest and lowest quartiles for FVC and % predicted FVC, warrant attention. Therefore, while our results do not support a broad role for n-3 PUFA in lung health at customary intake levels, they do suggest that substantially increased consumption, well above current averages, may be necessary to investigate potential benefits on lung function.
Hence, studies investigating the impact of n-3 PUFA intake on lung function have reported conflicting results. Unmeasured or uncontrolled confounding factors (such as environment and selection biases) and reverse causation might partially explain the discrepancy between these observational studies(Reference Smith, Lawlor and Harbord45). Studies to date have mostly focused on the dietary intake of n-3 PUFA from different food sources, which may lead to conclusions that may be confounded by contamination or food preparation methods in different geographical regions(Reference Jiang, Li and Yang46). Moreover, the disadvantages of these studies include variations in study design, limited sample size and short follow-up periods, which may weaken the strength of the observed associations(Reference Li, Song and Shen47). It is worth mentioning that the effects of dietary intake of n-3 PUFA on lung health may be mediated through n-3 PUFA in the circulation and tissues(Reference Skilton, Raitakari and Celermajer48). A previous study found that estimated dietary intake of n-3 PUFA was associated with plasma and erythrocyte membrane levels of these fatty acids(Reference Serra-Majem, Nissensohn and Øverby49). For example, evidence from a randomised trial demonstrated that the beneficial effect of fish oil supplementation on blood pressure was associated with increased plasma phospholipid levels of n-3 PUFA(Reference Bønaa, Bjerve and Straume50). In addition to diet, there are other determinants of n-3 PUFA levels, such as genetic variation in the metabolism of these fatty acid(Reference Chilton, Murphy and Wilson51). Glaser et al. indicated that n-3 PUFA levels in plasma, breast milk and tissues are affected by the concentration of fatty acid desaturase enzymes in the liver(Reference Glaser, Lattka and Rzehak52). Therefore, lung function may be impacted by the genetic and metabolic variations that cause variability in n-3 PUFA levels. For these reasons, in addition to using the nationally representative observational study in NHANES, our study used MR approach, which is less prone to bias and provides a higher level of evidence(Reference Smith and Ebrahim53). This study selected the GWAS dataset with the most prominent circulating n-3 PUFA concentration and lung function samples to minimise bias, the results of which were consistent with each other, making our results more robust. Furthermore, the IVW, MR-Egger, weighted median and weighted mode were also employed to examine the causal relationship between the two samples. The MR findings, based on a general population, have expanded our understanding of this issue that lifelong higher circulating n-3 PUFA concentration is unlikely to have a causal relationship with lung function. Circulating concentration of n-3 PUFA is considered a biochemical marker of long-term dietary intake, providing valuable insight into patterns of n-3 PUFA consumption. Therefore, our study suggests that an increase in circulating n-3 PUFA concentration caused by n-3 PUFA intake would not result in improved lung function.
Based on the analysis of the relationship between dietary intake of n-3 PUFA intake and lung function in NHANES, and consistent with the MR analysis conducted on the entire study population, it was concluded that n-3 PUFA have no significant effect on lung function. However, we also recognise several limitations in our findings. First, the data on n-3 PUFA intake were collected from two 24-h dietary recall interviews, which may not accurately represent the long-term average intake. Despite this, some large epidemiological studies have demonstrated the validity of 24-h recall dietary assessments. Second, the effect size of the GWAS was based on circulating n-3 PUFA concentrations rather than membrane concentrations, while considering the role of fatty acid receptors in cell signalling and immune responses, the association at membrane level may be of more significance. Third, we cannot completely rule out the presence of unmeasured confounding in this cross-sectional study, even though we have already adjusted for some known factors related to lung function. Additionally, stratified analyses based on sex, age, height and ethnicity were not possible due to the limited availability of only summary-level statistics from the general population of both sexes in the MR analysis. Lastly, given the availability of data sources, the findings from the two studies were based on a multi-ethnic US population and a population of European ancestry, respectively, so may not be generalisable to other ethnic populations.
Conclusion
In summary, while our analysis generally found no association between n-3 PUFA intake and improved lung function, observed significant findings at the highest quartile point to a potential dose-dependent effect. Our findings do not support a generalised role for n-3 PUFA in lung health enhancement, yet they pave the way for future studies to explore the threshold levels of n-3 PUFA required to impact lung function, potentially leading to new insights into diet-based interventions for respiratory health.
Acknowledgements
The authors thank the participants and researchers for their participation in the NHANES and MR study.
This research was funded by the Medical Science and Technology Project of Zhejiang Province (2023RC289, 2022RC272).
Z. X. L. and J. L. L. conceived and designed the study and interpreted the results of the data analyses. J. S. and L. J. L. performed the data analyses. J. L. L., J. S. and C. Y. Z. wrote the manuscript. All authors have read and approved the final manuscript.
The authors declare no conflict of interest.
Supplementary material
For supplementary material/s referred to in this article, please visit https://doi.org/10.1017/S0007114524000266
