Flavonoid intakes, chronic obstructive pulmonary disease, adult asthma, and lung function: a cohort study in the UK Biobank

Abstract

Background: Given their antioxidative stress, anti-allergic, anti-inflammatory, and immune-modulating effects, flavonoids are hypothesized to play a role in preventing chronic obstructive pulmonary disease (COPD) and asthma.

Objectives: This cohort study aimed to examine associations between flavonoid intake and COPD, asthma, and lung function.

Methods: Among 119,466 participants of the UK Biobank, median [interquartile range] age of 60 [53, 65] y, we estimated intakes of flavonoids, flavonoid-rich foods, and a flavodiet score from 24-h diet assessments. Prospective associations with both incident COPD and asthma and cross-sectional associations with measures of lung function [%predicted forced expiratory volume in 1s (FEV₁); and FEV₁/forced vital capacity (FVC)] were examined using multivariable-adjusted Cox proportional hazards and linear regression models, respectively. We investigated mediation by inflammation--represented by the INFLA score--and stratified analyses by smoking status.

Results: Compared with low intakes, moderate intakes of total flavonoids, flavonols, theaflavins + thearubigins, and flavanones, and moderate-to-high intakes of flavanol monomers, proanthocyanidins, anthocyanins, flavones, and the flavodiet score were associated with up to an 18% lower risk of incident COPD {e.g., [hazard ratio (95% confidence interval) for total flavonoids: 0.83 (0.75, 0.92)]} but not incident asthma. Furthermore, compared with low intakes, higher intakes of all flavonoid subclasses (except theaflavins + thearubigins), and the flavodiet score were associated with better percent predicted FEV₁ baseline. Associations were most apparent in ever (current or former) smokers. Flavonoid intakes were inversely associated with the INFLA score, which appeared to mediate 11%-14% of the association between intakes of proanthocyanidins and flavones and incident COPD.

Conclusions: Moderate-to-high flavonoid intakes were associated with a lower risk of COPD and better lung function, particularly among ever smokers. Promoting intakes of healthy flavonoid-rich foods, namely, tea, apples, and berries, may improve respiratory health and lower COPD risk, particularly in individuals with a smoking history.

Keywords: COPD; UK Biobank; asthma; dietary intake; epidemiology; flavonoids; inflammation; lung; prospective cohort; respiratory health.

FIGURE 1

Cubic spline curves describing the association between total flavonoid intakes, flavonoid subclass intakes, and the flavodiet score, and incident chronic obstructive pulmonary disease (COPD) in participants of the UK Biobank (n = 117,990). Hazard ratios and 95% confidence intervals are based on Cox proportional hazards models with age as the underlying timescale, adjusted for gender, region, body mass index, income, education, ethnicity, physical activity, alcohol intake, smoking status, packyears, passive smoking status, deprivation index, prevalent asthma, diet stability, and intakes of wholegrains, refined grains, sugar-sweetened beverages, dairy, eggs, fish, and meat and are comparing the specific level of intake (horizontal axis) to the median intake of participants in the lowest intake quintile (i.e., the reference).

FIGURE 2

Graphical representation of the multivariable−adjusted dose−response relationship between total flavonoid intakes and percentage predicted forced expiratory volume in 1 s (FEV₁) and FEV₁/forced vital capacity (FVC) ratio in (A) the whole population; (B) current and former (termed “ever”) smokers; and (C) never smokers. Estimated marginal mean responses and 95% confidence intervals (CIs) were obtained from restricted cubic splines within linear regression models adjusted for age, gender, height, region, body mass index, income, education, ethnicity, physical activity, alcohol intake, smoking status∗, packyears∗, passive smoking status, deprivation index, prevalent asthma, prevalent chronic obstructive pulmonary disease, emphysema, or bronchitis, diet stability, and intakes of wholegrains, refined grains, sugar-sweetened beverages, dairy, eggs, fish, and meat. Blue shading represents 95% CIs. The rug plot along the bottom of each graph depicts each observation. P values for nonlinearity and for the effect of the exposure (when modeled as a spline) on the response were obtained using likelihood ratio tests. ∗Terms not included in the model in the never-smoker subgroup analysis.

FIGURE 3

Cubic spline curves describing the association between total flavonoid intakes, flavonoid subclass intakes, and the flavodiet score and incident chronic obstructive pulmonary disease (COPD) stratified by smoking status. Hazard ratios and 95% confidence intervals (CIs) are based on Cox proportional hazards models with age as the underlying timescale, adjusted for gender, region, body mass index, income, education, ethnicity, physical activity, alcohol intake, smoking status (current vs. former)∗, packyears∗, passive smoking status, deprivation index, prevalent asthma, diet stability, and intakes of wholegrains, refined grains, sugar-sweetened beverages, dairy, eggs, fish, and meat and are comparing the specific level of intake (horizontal axis) to the median intake of participants in the lowest intake quintile (i.e., the reference; denoted with the horizontal dotted line). ∗For “ever smokers” only.