The effect of dietary components on inflammatory lung diseases – a literature review

Abstract

Anti-inflammatory treatment in chronic inflammatory lung diseases usually involves glucocorticosteroids. With patients suffering from serious side effects or becoming resistant, specific nutrients, that are suggested to positively influence disease progression, can be considered as new treatment options. The dietary inflammatory index is used to calculate effects of dietary components on inflammation and lung function to identify most potent dietary components, based on 162 articles. The positive effects of n-3 PUFAs and vitamin E on lung function can at least partially be explained by their anti-inflammatory effect. Many other dietary components showed only small or no effects on inflammation and/or lung function, although the number of weighted studies was often too small for a reliable assessment. Optimal beneficial dietary elements might reduce the required amounts of anti-inflammatory treatments, thereby decreasing both side effects and development of resistance as to improve quality of life of patients suffering from chronic inflammatory lung diseases.

Keywords:

• Chronic inflammatory lung diseases
• nutrition
• anti-inflammatory nutrition
• glucocorticosteroid resistance
• asthma
• COPD

Introduction

Lung diseases including asthma, chronic obstructive pulmonary disease (COPD) and interstitial lung diseases (ILD) such as sarcoidosis and idiopathic pulmonary fibrosis (IPF) are characterised by chronic inflammation (O’Byrne & Postma 1999; Rothkrantz-Kos et al. 2003; King 2005; Moldoveanu et al. 2009). Glucocorticosteroids are considered to be the mainstay in anti-inflammatory treatment for various inflammatory diseases, including these lung diseases (Barnes et al. 2003; Barnes & Adcock 2009). However, some patients show no response to these drugs and the use of glucocorticosteroids was shown to cause serious side effects (Barnes et al. 2003; King 2005; Barnes & Adcock 2009). Preventing such resistance and enhancing the effectiveness of glucocorticosteroids with nutrition present an interesting opportunity for treatment of these chronic inflammatory lung diseases.

Inflammation, the immediate response in tissue to injury, pathogen infiltration or irritants, is characterised by redness, swelling, heat, pain and loss of function (Coussens & Werb 2002; Calder 2006; Moldoveanu et al. 2009). When tissue is affected, chemical signals initiate activation and migration of leukocytes (neutrophils, monocytes and eosinophils) from the venous system to the damaged site and extracellular matrix. Following activation of adhesion molecules within the vascular endothelium, leukocyte integrins are triggered to be activated and upregulated. These attracted neutrophils are immobilised on the surface of the vascular endothelium and subsequently transmigrate to sites of injury (Coussens & Werb 2002). The progress of inflammation is shaped by chemotactic cytokines (chemokines) as TNF-α and TGF-β1, which stimulate monocyte migration to the injured tissue. These monocytes are involved in tissue repair, and defence mechanisms involve phagocytosis, the production of reactive oxygen species, and elimination of the pathogen as well as of cellular and tissue debris (Coussens & Werb 2002; Calder 2006; Pawelec et al. 2014). The inflammatory response may also affect healthy tissue and is therefore tightly controlled by the production of anti-inflammatory cytokines to terminate the process when the injury is repaired (Coussens & Werb 2002; Fullerton et al. 2013; Pawelec et al. 2014). When the process is not ended correctly due to either persistence of pro-inflammatory cytokines, failure or incomplete actions of anti-inflammatory cytokines or even persistence of these anti-inflammatory cytokines, inflammation can result in chronic inflammation (Coussens & Werb 2002; Moldoveanu et al. 2009; Fullerton et al. 2013; Pawelec et al. 2014).

With inflammation being a key factor in the aetiology of asthma, COPD, sarcoidosis and IPF, other aspects distinguish these diseases. COPD is characterised by a poorly reversible and usually progressive airflow limitation as a result of fixed narrowing of small airways, emphysema and luminal obstruction (Barnes et al. 2003; Rabe et al. 2007). In COPD, the parenchymal cells of the lungs are affected by inflammation, resulting in decreased airflow (Barnes et al. 2003; Moldoveanu et al. 2009). Asthma is typified by chronic inflammation of the airways only, with variable and widespread episodes of airway hyper-responsiveness leading to shortness of breath, wheezing, chest tightness and coughing due to chronic inflammation (Barnes et al. 2003; Bateman et al. 2008; Moldoveanu et al. 2009). ILD refers to a group of over 200 pulmonary diseases which are characterised by inflammation and fibrosis of the septal interstitium of the lungs, next to inflamed alveoli and distal airways (King 2005; Bourke 2006). Two of the more common diseases of this group are sarcoidosis and IPF (King 2005). Sarcoidosis is characterised by granuloma in multiple organs following inflammation, but most frequently affects the lungs and the lymphatic system (Baughman et al. 2003). IPF is a chronic disorder where fibrotic tissue accumulates in the lungs (Gross & Hunninghake 2001; King 2005; Patel et al. 2012). Following inflammation, injury in the lungs results in progressive collagen accumulation, leading to breathing difficulties and finally resulting in respiratory failure (Gross & Hunninghake 2001; Selman et al. 2001; King 2005; Patel et al. 2012).

Since inflammation is a key factor in chronic lung diseases as described above, anti-inflammatory therapies have been widely explored. Glucocorticoids are the mainstay in anti-inflammatory therapy. Glucocorticosteroids bind to the glucocorticosteroid receptor (GR) in the cytoplasm after diffusion across the cell membrane, which is translocated to the nucleus following activation by ligand binding (Barnes & Adcock 2009). The glucocorticosteroids–GR complex binds to the glucocorticoid response element (GRE) in the promotor region of glucocorticoid-responsive genes. When this complex binds to the GRE, gene transcription of previously activated inflammatory genes is directly inactivated (Barnes & Adcock 2009). The anti-inflammatory effect of GRs is also thought to be caused by various indirect effects, such as interfering with these inflammatory transcription factors, which normally activate gene expression, through binding to their promotor sites, or by inducing inhibitors of transcription factors (O’Byrne & Postma 1999). Studies into resistance to glucocorticosteroids in mainly asthma patients have led to the identification of six possible molecular mechanisms contributing to this resistance: (i) genetic susceptibility; (ii) GR modification due to phosphorylation, nitrosylation or ubiquitination altering their binding affinity; (iii) increased expression of the GR-isoform GR-β or disruption of the isoform GRα; (iv) increased activation of pro-inflammatory transcription factors (e.g. AP1, JNK), preventing interaction of GR with GRE and other transcription factors due to binding to the GR; (v) abnormal histone acetylation leading to no transactivation of genes; or (vi) increased P-glycoprotein transporting drugs out of cells (Barnes & Adcock 2009).

Since oxidative stress is one of the factors affecting histone acetylation, it is suggested to contribute to glucocorticosteroid resistance by suppressing the anti-inflammatory effect of the drug (Barnes & Adcock 2009; Ruijters et al. 2014a). Antioxidants can prevent this resistance, as was demonstrated in vitro with cocoa-derived epicatechin which was able to reduce cortisol resistance and protect the anti-inflammatory effects of dexamethasone (Ruijters et al. 2014a, 2014b). The role of nutrition in inflammatory diseases has been the subject of research previously: a diet rich in fruit, vegetables, cereals and fish (the so called prudent dietary pattern) was associated with a lower risk on COPD whereas a higher risk was associated with consumption of a Western diet (rich in refined grains, preserved meat, potatoes and sweets) (Varraso et al. 2007, 2010). Other studies showed a protective effect on oxidative processes and inflammation by different dietary components as fruit and vegetables, flavonoids, vitamin C, vitamin E, β-carotene, fatty acids and various minerals (Romieu 2005; Geraets et al. 2009; Hazewindus et al. 2014). The link between oxidative stress and several inflammatory lung diseases suggests a pivotal role for nutrition in their treatment (Romieu 2005).

This study therefore focussed on the effects of dietary components (including non-nutrients, nutrients, food items and diets) on inflammatory and immunological markers and respiratory function in chronic inflammatory lung diseases described above. Since this study aims to direct further research into dietary components which affect inflammation and/or function, all types of research – ranging from in vitro experiments to human trials – were included. Following the dietary inflammatory index as described by Shivappa et al., an inflammatory and respiratory effect score was calculated to identify potent dietary components in the treatment of chronic inflammatory lung diseases (Shivappa et al. 2014).

Methods

The dietary inflammatory index developed by Cavicchia et al. (2009) and optimised by Shivappa et al. (2014) measures the inflammatory effect of nutrients, relating it to the total dietary intake based on food consumption data (Cavicchia et al. 2009; Shivappa et al. 2014). The dietary inflammatory index was developed to score the overall effect of diet on inflammation (Cavicchia et al. 2009; Shivappa et al. 2014). With the index, the diet of an individual can be scored on a scale from maximally anti-inflammatory (with a score of −1) up to maximally pro-inflammatory (with a score of +1) by taking into account all potential dietary components affecting inflammation (Cavicchia et al. 2009; Shivappa et al. 2014). In the optimised dietary inflammatory index developed by Shivappa et al., the individual intake was compared with referent intakes provided by different food consumption data sets (Shivappa et al. 2014).

The dietary inflammatory index and its scoring system were the starting point for this literature review focussing on the effects of dietary components on inflammatory lung diseases. With the aim of this study being to identify components which can influence chronic inflammatory lung diseases and their underlying inflammatory processes, different experimental setups (such as in vitro studies, animal studies, observational studies and intervention trials) with potential varying quality of studies were required. The dietary inflammatory index provided a tool to score the effects found in these different studies.

Literature review strategy

Various search engines were used (Google Scholar^®, Pubmed^®, Science Direct^®) to identify peer-reviewed studies published in English on the effects of single (non-)nutrients or whole diet on lung function or inflammatory markers in chronic inflammatory lung diseases. In the search strategy, the four diseases were combined with terms for nutrition (nutrients, diet, whole diet, nutrition) and “anti-inflammatory”. Cancer was explicitly excluded in the search strategy, as well as reviews, since only original research was used to calculate the effects of nutrition on these lung diseases.

In total, 1206 articles were screened (Figure 1). Studies were included when they studied one of the four diseases or models resembling these diseases and when studying either markers of lung function (forced expiratory volume in 1 s (FEV₁), forced vital capacity (FVC), FEV₁/FVC ratio, forced expiratory flow (FEF), peak expiratory flow (PEF)) or inflammatory and immunological markers (pro-inflammatory: interleukin (IL)-1β, IL-4 in asthma, IL-5, IL-6, IL-8, tumour necrosis factor (TNF)-α, leukotriene (LT) B₄ and/or C-reactive protein (CRP); anti-inflammatory: IL-4 and IL-10) and models (type 1 helper T cells (Th₁)/type 2 helper T cells (Th₂) ratio; influx of immune/inflammatory cells in lung). As IL-4 mediates pro-inflammatory functions in asthma, it is used as pro-inflammatory immunological marker, but is regarded as anti-inflammatory marker in other diseases because of its suppressing effects on pro-inflammatory cytokine production (Steinke & Borish 2001; Woodward et al. 2010).

Figure 1. Flowchart literature review.

Studies were excluded when they studied: (i) other inflammation-associated diseases (such as CVD); (ii) other lung diseases (e.g. cystic fibrosis, rhinitis); (iii) effects of intake or exposure to substances other than well-defined dietary components (as drugs, environmental agents, herbal extracts); and (iv) nutrition status or plasma levels of dietary components instead of intake (as vitamin D status). Last, articles focussing on disease prevalence, incidence or occurrence were excluded leading to 162 included articles.

Scoring algorithm

Included articles were scored based on (i) the outcome and (ii) the design of the study. Following the set-up of the dietary inflammatory index, the outcomes were labelled as “+1” when showing a pro-inflammatory effect or a decrease in lung function, “−1” with anti-inflammatory effects or improved lung function and “0” when no significant effect of the dietary component was reported (Shivappa et al. 2014). Scores were checked for consistency by all authors. When more effects (as independent intake of different dietary components) were described in one study or multiple methods were used (as in vitro and animal models), all effects were taken into consideration in the calculation of the final effect by weighing the study more than once. The second step was weighing the studies based on their design, which resulted in a specific amount of credits (Table 1).

Table 1. Study design weights (adapted from Shivappa et al. 2014).

Type of study	Study design	Credits
Human	Experimental	10
	Prospective cohort	8
	Case-control	7
	Ex vivo experimental	7
	Cross-sectional	6
Animal	Experimental	5
Cell culture	Experimental	3

Subsequently, the dietary component-specific effect score was calculated as follows: (i) the credits attributed to the weighted pro- and anti-inflammatory studies were divided by the weighted number of studies and (ii) the anti-inflammatory fraction was subtracted from the pro-inflammatory fraction (as exemplified in Figure 2). As previously done in calculating the dietary inflammatory index, the median of the total credits across all dietary components was used as a cut-off point to indicate the literature pool (Shivappa et al. 2014). When the number of weighted articles was below the cut-off point of 15 (inflammation) or 10 credits (lung function), the score was adjusted: the credits were divided by 15 or 10 and subsequently multiplied by the dietary component-specific effect score following from the first two steps (Shivappa et al. 2014).

Figure 2. Example of the calculation of the nutrient-specific effect score for n-3 PUFAs in two steps: (i) calculation of the pro- and anti-inflammatory fraction and (ii) the nutrient specific raw inflammatory effect score (adapted from Shivappa et al. 2014).

Results

Since some of the 162 articles measured both inflammatory markers and respiratory function and some studies focussed on multiple dietary components, 101 articles were included in calculating the inflammatory effect scores of dietary components and 88 articles were included to calculate the respiratory function effect scores.

As shown in Table 2, various dietary components showed an anti-inflammatory effect when consumption occurred during chronic inflammatory lung diseases. EGCG (epigallocatechin gallate), n-3 PUFAs (polyunsaturated fatty acids with a double bond at the third carbon atom from the ω-end of the carbon chain), probiotics and vitamin E have been studied most thoroughly and have an inflammatory effect score of −1, −0.743, −0.538 and −1, respectively. This indicates that these dietary components gave rise to an anti-inflammatory effect in most of the included studies. Other components eliciting high anti-inflammatory effects were prebiotics (−0.667), quercetin (−0.864), resveratrol (−0.889), vitamin C (−0.667) and vitamin D (−1), although the number of weighted articles studying these dietary components in chronic inflammatory lung diseases was lower and sometimes fell below the cut-off point of literature robustness of 15 credits. No significant anti-inflammatory effects were found following intake of caffeine, high fat, the Mediterranean diet and an adjusted n-6:n-3 PUFA ratio. A pro-inflammatory effect was found in chronic inflammatory lung diseases with fish consumption (0.294) and n-6 PUFA (polyunsaturated fatty acids with a double bond at the sixth carbon atom from the ω-end of the carbon chain) intake (0.677).

Table 2. Nutrient specific inflammatory effect score.

Food item	Credits	Raw inflammatory effect score	Inflammatory effect score^a	References
EGCG	34	−1.000	−1.000	(Chen et al. 2002; Wheeler et al. 2004; Bani et al. 2006; Kim et al. 2006; Qin et al. 2011; Chan et al. 2012; Lee et al. 2013; You et al. 2014)
Vitamin D	18	−1.000	−1.000	(Dimeloe et al. 2012; Kuo et al. 2012; Agrawal et al. 2013; Zhong et al. 2013)
Vitamin E^b	49	−1.000	−1.000	(Okamoto et al. 2006; Wagner et al. 2007; Wagner et al. 2008; Wiser et al. 2008; Mabalirajan et al. 2009; Geiser et al. 2013; Hernandez et al. 2013)
Resveratrol	27	−0.889	−0.889	(Culpitt 2003; Donnelly et al. 2004; Birrell 2005; Lee et al. 2009; Knobloch et al. 2010)
Quercetin	22	−0.864	−0.864	(Donnelly et al. 2004; Boots et al. 2009; Boots et al. 2011; Verma et al. 2013)
Soy isoflavone	12	−1.000	−0.800	(Kalhan et al. 2008; Bao et al. 2011)
n-3 PUFAs	109	−0.743	−0.743	(Payan et al. 1986; Arm et al. 1988; Broughton et al. 1997; Hodge et al. 1998; Okamoto et al. 2000; Mickleborough et al. 2003; Matsuyama et al. 2005; Mickleborough et al. 2006; de Batlle et al. 2012; Kunitsugu et al. 2012; Jang et al. 2014; Schuster et al. 2014)
Flavonoids	35	−0.714	−0.714	(Park et al. 2007; Geraets et al. 2009; Xie et al. 2009; Wu et al. 2011; Kim et al. 2013)
Fatty acid supplementation	15	−0.667	−0.667	(Kanwar et al. 2008; Wood et al. 2010)
Prebiotics	10	−1.000	−0.667	(Vos et al. 2007; Yasuda et al. 2010)
Vitamin C	10	−1.000	−0.667	(Tecklenburg et al. 2007)
Vitamin C + vitamin E	10	−1.000	−0.667	(Sienra-Monge et al. 2004)
Zinc	10	−1.000	−0.667	(Lang et al. 2010; Lu et al. 2012)
Probiotics	130	−0.538	−0.538	(Charng et al. 2006; Feleszko et al. 2007; Forsythe et al. 2007; Moreira et al. 2007; Karimi et al. 2009; Lim et al. 2009; Hougee et al. 2010; Lyons et al. 2010; Kukkonen et al. 2011; Jan et al. 2012; MacSharry et al. 2012; Miraglia Del Giudice et al. 2012; Zhang et al. 2012; Harb et al. 2013; Kim et al. 2013)
n-3 + n-6 PUFAs	20	−0.500	−0.500	(Surette et al. 2003; Broekhuizen et al. 2005)
Fruits and vegetables	18	−0.444	−0.444	(Romieu et al. 2009; Baldrick et al. 2012)
Synbiotics	17	−0.412	−0.412	(van de Pol et al. 2011)
Blackcurrant poly-phenolic extracts	9	−0.667	−0.400	(Hurst et al. 2010; Nyanhanda et al. 2014)
Conjugated linoleic acid	15	−0.333	−0.333	(Jaudszus et al. 2008; Kanwar et al. 2008; Böcking et al. 2014)
Dietary fibre	5	−1.000	−0.333	(Trompette et al. 2014)
Ellagic acid	5	−1.000	−0.333	(Zhou et al. 2014)
Lycopene	5	−1.000	−0.333	(Hazlewood et al. 2011)
Naringenic chalcone	5	−1.000	−0.333	(Iwamura et al. 2010)
Selenium	5	−1.000	−0.333	(Jeong et al. 2002)
Sesame	8	−0.625	−0.333	(Hsu et al. 2013; Hsieh et al. 2014)
Vitamin A	15	−0.333	−0.333	(Sauer et al. 1995; Hisada et al. 1999; Schuster et al. 2008)
Vitamin C + vitamin E + selenium	5	−1.000	−0.333	(Kirschvink et al. 2002)
Retinoic acid	8	−0.375	−0.200	(Fujita et al. 2004; Takamura et al. 2004)
High fat	5	0.000	0.000	(de Vries et al. 2009)
Mediterranean diet	18	0.000	0.000	(Romieu et al. 2009; Sexton et al. 2013)
n-6:n-3 PUFA ratio	6	0.000	0.000	(de Batlle et al. 2012)
Caffeine	3	0.000	0.000	(Geraets et al. 2006)
Creatine	18	0.111	0.111	(Nomura et al. 2003; Fuld et al. 2005; Vieira et al. 2007)
Fish	17	0.294	0.294	(Yin et al. 2009; Wood et al. 2010; Kunitsugu et al. 2012)
n-6 PUFAs	31	0.677	0.677	(Hodge et al. 1998; Okamoto et al. 2000; Muehlmann et al. 2010; de Batlle et al. 2012)

^aCorrected for literature robustness (cut off point =15).

^bIncluding studies assessing the effect of specifically α- and γ-tocopherol.

Several dietary components were shown to elicit positive effects on respiratory function in chronic inflammatory lung diseases (Table 3). Most studied were the effects of intake of caffeine (−0.455), n-3 PUFAs (−0.374), vitamin C (−0.378) and vitamin E (−0.238). Consumption of bread (−0.800), fruit (−0.758), fruit and vegetables (−0.722), n-3 and n-6 PUFAs (−0.667), probiotics (−0.714), soy isoflavone (−1), synbiotics (−1), theophylline (−1) and white wine (−0.800) was suggested to result in the most significant improvements in respiratory function, although the weighted number of studies showing these effects is only ranging from 8 up to 36 credits. No effect was found following intake of calcium, carotenoids, conjugated linoleic acid, creatine, dairy, dietary salt, flavanols, flavones, iron, margarine and oils, meat, monosodium glutamate, n-9 PUFAs, sodium and potassium, olive oil, phosphorus, potassium, potatoes, red wine, selenium, sodium, tea, total energy intake, vegetables as well as vitamin C combined with vitamin E and selenium. A decline in function was observed in studies assessing the effects of n-6 PUFAs (0.162).

Table 3. Nutrient specific respiratory effect score.

Food item	Credits	Raw function effect score	Function effect score^a	References
Soy isoflavone	11	−1.000	−1.000	(Bao et al. 2011; Bime et al. 2012)
Synbiotics	10	−1.000	−1.000	(van de Pol et al. 2011)
Theophylline	10	−1.000	−1.000	(Becker et al. 1984)
Bread	8	−1.000	−0.800	(Tabak et al. 1999)
White wine	8	−1.000	−0.800	(Siedlinski et al. 2012)
Fruit	33	−0.758	−0.758	(Strachan et al. 1991; Cook et al. 1997; Carey et al. 1998; Tabak et al. 2001b; de Batlle et al. 2010)
Fruit and vegetables	36	−0.722	−0.722	(Tabak et al. 1999; Romieu et al. 2009; Keranis et al. 2010; Baldrick et al. 2012)
Probiotics	35	−0.714	−0.714	(Feleszko et al. 2007; Miraglia Del Giudice et al. 2012; Zhang et al. 2012; West et al. 2013)
n-3 + n-6 PUFAs	30	−0.667	−0.667	(Mickleborough et al. 2003; Broekhuizen et al. 2005; Mickleborough et al. 2006)
β-carotene	22	−0.636	−0.636	(Grievink et al. 1998; Tabak et al. 1999; Butland et al. 2000)
Catechins	6	−1.000	−0.600	(Tabak et al. 2001a)
Dietary fibre	6	−1.000	−0.600	(Kan et al. 2008)
Total protein	6	−1.000	−0.600	(Yazdanpanah et al. 2010)
EGCG	5	−1.000	−0.500	(Bani et al. 2006)
Ellagic acids	5	−1.000	−0.500	(Zhou et al. 2014)
Fisetin	5	−1.000	−0.500	(Wu et al. 2011)
Lycopene	20	−0.500	−0.500	(Neuman et al. 2000; Falk et al. 2005)
Prebiotics	5	−1.000	−0.500	(Vos et al. 2007)
Vitamin C + vitamin E	20	−0.500	−0.500	(Trenga et al. 2001; Hernandez et al. 2009)
Vitamin D	43	−0.488	−0.488	(Agarwal et al. 2006; Shaheen et al. 2011; Lehouck et al. 2012; Arshi et al. 2014; Ng et al. 2014)
Caffeine	110	−0.455	−0.455	(Becker et al. 1984; Crivelli et al. 1986; Gong et al. 1986; Bukowskyj & Nakatsu 1987; Colacone et al. 1990; Kivity et al. 1990; Duffy & Phillips 1991; Henderson et al. 1993; Yurach et al. 2011)
Mediterranean diet	18	−0.444	−0.444	(Romieu et al. 2009; Sexton et al. 2013)
Fish	28	−0.429	−0.429	(Schwartz & Weiss 1994; Tabak et al. 1999; Butland et al. 2000; Ng et al. 2014)
Resveratrol	13	−0.385	−0.385	(Lee et al. 2009; Siedlinski et al. 2012)
Vitamin C	185	−0.378	−0.378	(Kordansky et al. 1979; Schachter & Schlesinger 1982; Ting et al. 1983; Malo et al. 1986; Schertling 1989; Britton et al. 1995; Cohen 1997; Cook et al. 1997; Grievink et al. 1998; Hu et al. 1998; Tabak et al. 1999; Butland et al. 2000; McKeever et al. 2002; Fogarty et al. 2003; de Luis et al. 2005; Tecklenburg et al. 2007; Kan et al. 2008; de Batlle et al. 2010; Nadi et al. 2012; Ng et al. 2014)
n-3 PUFAs	115	−0.374	−0.374	(Arm et al. 1988; Kirsch et al. 1988; Duffy & Phillips 1991; Thien et al. 1993; Broughton et al. 1997; Hodge et al. 1998; Villani et al. 1998; Okamoto et al. 2000; de Luis et al. 2005; A Moreira et al. 2007; Kan et al. 2008; McKeever et al. 2008; Ng et al. 2014)
Magnesium	56	−0.357	−0.357	(Hill et al. 1997; Butland et al. 2000; McKeever et al. 2002; Fogarty et al. 2003; Gontijo-Amaral et al. 2006; Kazaks et al. 2010)
Ginger	3	−1.000	−0.300	(Ghayur et al. 2008)
Vitamin E^b	113	−0.283	−0.283	(Britton et al. 1995; Grievink et al. 1998; Tabak et al. 1999; Butland et al. 2000; McKeever et al. 2002; Daga et al. 2003; Pearson et al. 2004; de Luis et al. 2005; Okamoto et al. 2006; Kan et al. 2008; Nadeem et al. 2008; Mabalirajan et al. 2009; de Batlle et al. 2010; Ng et al. 2014)
Vitamin A	26	−0.231	−0.231	(Hisada et al. 1999; McKeever et al. 2002; de Luis et al. 2005; Ng et al. 2014)
Apples	8	0.000	0.000	(Butland et al. 2000)
Calcium	7	0.000	0.000	(Hirayama et al. 2010)
Carotenoids	6	0.000	0.000	(Kan et al. 2008)
Conjugated linoleic acid	10	0.000	0.000	(Stickford et al. 2011)
Creatine	10	0.000	0.000	(Fuld et al. 2005)
Dairy	6	0.000	0.000	(Ng et al. 2014)
Dietary salt	13	0.000	0.000	(Pistelli et al. 1993; Demissie et al. 1996)
Flavanols	6	0.000	0.000	(Tabak et al. 2001a)
Flavones	6	0.000	0.000	(Tabak et al. 2001a)
Iron	17	0.000	0.000	(Shaheen et al. 2007; Hirayama et al. 2010)
Margarine and oils	8	0.000	0.000	(Tabak et al. 1999)
Meat	6	0.000	0.000	(Kan et al. 2008)
Monosodium glutamate	10	0.000	0.000	(Yoneda et al. 2011)
n-9 PUFAs	7	0.000	0.000	(de Luis et al. 2005)
Sodium + potassium	6	0.000	0.000	(Zoia et al. 1995)
Olive oil	8	0.000	0.000	(de Batlle et al. 2010)
Phosphorus	7	0.000	0.000	(Hirayama et al. 2010)
Potassium	7	0.000	0.000	(Hirayama et al. 2010)
Potatoes	8	0.000	0.000	(Tabak et al. 1999)
Red wine	8	0.000	0.000	(Siedlinski et al. 2012)
Selenium	13	0.000	0.000	(Hirayama et al. 2010; Ng et al. 2014)
Sodium	7	0.000	0.000	(Hirayama et al. 2010)
Tea	6	0.000	0.000	(Tabak et al. 2001a)
Total energy	14	0.000	0.000	(Butland et al. 2000; Yazdanpanah et al. 2010)
Vegetables	14	0.000	0.000	(Cook et al. 1997; de Batlle et al. 2010)
Vitamin C + vitamin E + selenium	5	0.000	0.000	(Kirschvink et al. 2002)
n-6 PUFAs	43	0.162	0.162	(Hodge et al. 1998; Okamoto et al. 2000; Morris et al. 2001; de Luis et al. 2005; McKeever et al. 2008)

^aCorrected for literature robustness (cut off point =10).

^bIncluding studies assessing the effect of specifically α-tocopherol.

Discussion

This study summarises the effects of intake of dietary components on chronic inflammatory lung diseases based on 162 studies in an attempt to identify components which should be studied into more detail, to review their exact effect on these diseases. The scores calculated with the dietary inflammatory index of these dietary components provide an indication of their effect on chronic inflammatory lung diseases. The scores were corrected for literature robustness by taking into account the median of credits attributed to the evidence by weighing the articles. This resulted in a cut-off point of 15 credits for inflammatory and immunological markers and a cut-off point of 10 credits for respiratory function. When the number of credits attributed to a dietary component was falling below the cut-off point, the effect found was divided by this cut-off point, as described in the “Methods” section, to correct for literature robustness. When a health benefit of a food product is reviewed however by agencies (as the European Food Safety Authority or the U.S. Food and Drug Administration) to advice on their legal status, multiple human intervention studies demonstrating the beneficial effect are required (Ellwood et al. 2010; EFSA NDA Panel 2011; U.S. Food and Drug Administration 2014). Translating this into a number of credits, a cut-off point of 20 could be recommended, below which the outcome should be corrected for literature robustness. Although this number of 20 credits is not necessarily based on two intervention studies but could also be reached by comparing seven in vitro studies (Table 1), a number of 20 credits could already establish a more reliable indication of an effect.

Effects of dietary components on markers and function

Considering the number of credits, most convincing evidence has been found for the positive effects following intake of EGCG, n-3 PUFAs, probiotics and vitamin E on inflammatory and immunological markers in chronic inflammatory lung diseases (Table 2). The calculated effects of these dietary components range from −0.538 up to −1. The effects on lung function have been studied mostly for caffeine, n-3 PUFAs, vitamin C and vitamin E (Table 3). With scores ranging from −0.238 to −0.455, their beneficial effects on respiratory function appeared to be less explicit than the effects of the most thoroughly studied dietary components on inflammatory and immune markers.

The calculated scores for the effects of n-3 PUFAs and vitamin E on chronic inflammatory lung diseases suggest that improvements in respiratory function are at least partially attributable to improvement of inflammatory and immunological markers. Intake of n-3 PUFAs may result in decreased inflammation and improved lung function by (i) competition with n-6 PUFAs for metabolism by specific enzymes (cyclooxygenase, lysyl oxidase or cytochrome P450 oxygenase), resulting in alternative, less pro-inflammatory and even anti-inflammatory eicosanoids (metabolites); (ii) binding to and activation of receptors bound to the plasma membrane or found in the cytosol as G-protein coupled receptors (mediating the anti-inflammatory effects of n-3 PUFAs) and PPAR (peroxisome proliferator-activated receptor) transcription factors (inhibiting NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) activation and thereby pro-inflammatory gene transcription); and (iii) inducing the anti-inflammatory pathways by resolvins and protectins derived from n-3 PUFAs (Yates & Calder 2014). The effect on the eicosanoid metabolisation was observed to be crucial in chronic inflammatory lung diseases: various LTs derived from arachidonic acid (n-6 PUFAs) are important mediators in asthma and are associated with inflammation in COPD (Okamoto et al. 2000; Yates & Calder 2014). N-3 PUFAs suppress the generation of LTB₄, which promotes the production of inflammatory cytokines, and reduce production of IL-1, IL-6 and TNF-α by leukocytes (Okamoto et al. 2000; Calder 2006). This mechanism of action could also explain both the pro-inflammatory effects (0.667) and negative effects on respiratory function (0.162) following n-6 PUFA intake, due to the formation of more pro-inflammatory metabolites resulting from the ingestion of n-6 PUFAs.

For vitamin E intake, a calculated inflammatory effect score of −1 and a respiratory function score of −0.283 were found, indicating a beneficial effect on chronic inflammatory lung diseases. Vitamin E can be found in eight isoforms, with α-tocopherol and γ-tocopherol being the most abundant (Cook-Mills & Avila 2014). Human tissue concentrations of α-tocopherol are 10 times higher than γ-tocopherol concentrations due to the higher preference of HDL (high-density lipoprotein) and LDL (low-density lipoprotein) particles for α-tocopherol and a higher rate of degradation of γ-tocopherol (Cook-Mills & Avila 2014). The specific effects of the different isoforms are highly debated (van Acker et al. 1993; Wagner et al. 2007; Wagner et al. 2008; Geiser et al. 2013; Cook-Mills & Avila 2014). A total of 113 credits were taken into account for the function score, of which 108 credits were based on studies assessing total vitamin E intake (Britton et al. 1995; Grievink et al. 1998; Tabak et al. 1999; Butland et al. 2000; McKeever et al. 2002; Daga et al. 2003; Pearson et al. 2004; de Luis et al. 2005; Kan et al. 2008; Nadeem et al. 2008; Mabalirajan et al. 2009; de Batlle et al. 2010; Ng et al. 2014) and five credits focussed on the effect of α-tocopherol resulting in a significant improvement of respiratory function in mice (Okamoto et al. 2006). Inflammatory and immune responses were mainly studied with γ-tocopherol administration (39 of the 49 credits) (Wagner et al. 2007, 2008; Wiser et al. 2008; Geiser et al. 2013; Hernandez et al. 2013), where five credits originated from studies focussed on total vitamin E intake (Mabalirajan et al. 2009) and five from studies on α-tocopherol intake (Okamoto et al. 2006). Vitamin E can affect the metabolism of arachidonic acid by decreasing cyclooxygenase enzymes and lowering LT synthesis, next to decreasing the serum levels of immunoglobulin E which is typically elevated in asthmatics (Fogarty et al. 2000; Okamoto et al. 2006). Both α- and γ-tocopherol scavenge reactive oxygen species, although the clinical relevance of this antioxidant effect is questioned and the positive effects of vitamin E supplementation are attributed to other mechanisms of action such as (i) upregulating PPARγ, (ii) inhibiting cyclooxygenase and lysyl oxidase and (iii) inhibiting nitration reactions by γ-tocopherol (Wagner et al. 2008). Both isoforms show similar inhibiting capacities of protein kinase B, which induces production of pro-inflammatory cytokines and activates NF-κB (Cook-Mills & Avila 2014). But whereas α-tocopherol showed anti-inflammatory effects and reduced airway hyper-reactivity in mice, γ-tocopherol was observed in mice to induce pro-inflammatory effects and enhanced airway hyper reactivity during eosinophilic airway inflammation (Cook-Mills & Avila 2014). On the other hand, γ-tocopherol was observed to react with reactive nitrogen species (found in eosinophils and neutrophils) and blocked acute neutrophil inflammation (Patel et al. 2007; Wagner et al. 2007; Cook-Mills & Avila 2014). Therefore γ-tocopherol is suggested to exert a broader anti-inflammatory profile than α-tocopherol (Wagner et al. 2007). Consequently, both isoforms are considered to contribute to the anti-inflammatory effects of vitamin E on chronic inflammatory lung diseases.

Although the effects of probiotics on respiratory function have been assessed in relatively few studies (35 credits), the calculated anti-inflammatory effects (−0.538) are reflected in a positive effect on respiratory function (−0.714). Probiotics is the collective term for 400–500 species of live bacteria which survive digestion and are consequently colonised in the gastrointestinal tract, providing a health benefit to the host (Heller & Duchmann 2003; Osborn & Sinn 2007). Since all probiotics do not influence the immune system similarly, the elicited health benefits can differ per bacterial strain and metabolites (Frei et al. 2015). The differences in effectiveness of strains also influenced the calculated score in this study: where some tested bacteria strains showed anti-inflammatory properties (e.g. Lactobacillus rhamnosus GG and Lactobacillus reuteri) (Feleszko et al. 2007; Forsythe et al. 2007; Karimi et al. 2009; Harb et al. 2013), other strains did not affect inflammation significantly (e.g. Bifidobacterium longum and Lactobacillus casei) (Lim et al. 2009; Lyons et al. 2010). The beneficial effects found following probiotic administration appear to be triggered via different mechanisms, which are still poorly understood (Frei et al. 2015). Various aspects as dendritic cells, epithelial cells, T regulatory cells, effector lymphocytes, natural killer T cells and B cells seem to be influenced by consumption of probiotics, resulting in reduced inflammation (Strickland & Holt 2011; Jan et al. 2012; Frei et al. 2015).

A link between the calculated improved function (−0.378) and the anti-inflammatory and immunological effect (−0.667) of vitamin C is complicated to assess due to the limited number of credits. Vitamin C intake is suggested to reduce inflammation due to its antioxidant capacity, by lowering the oxidative damage in the lungs and by blocking activation of the NF-κB pathway by inhibiting TNF-α formation (Tecklenburg et al. 2007; Milan et al. 2013). Other suggested mechanisms are the antiviral properties of vitamin C or its potency to alter the arachidonic acid pathway (Cohen 1997; Milan et al. 2013).

The relationship between improvement of inflammatory and immunological markers and resulting respiratory function appeared less evident for many other studied dietary components, such as caffeine and EGCG. With a score of −0.455 based on 110 credits, caffeine intake was shown to result in an improvement in lung function. Still, only one in vitro study (assigned three credits) assessed the inflammatory and immunological effects caused by caffeine consumption, and found no significant improvements (Geraets et al. 2006). The positive effect of caffeine on lung function is attributed to its bronchodilating effect (Gong et al. 1986; Kivity et al. 1990; Duffy & Phillips 1991; Welsh et al. 2010). Caffeine, as one of the methylxanthines, can affect various cellular processes and thereby instigate bronchodilation: it is an inhibitor of cyclic nucleotide phosphodiesterases, of intracellular translocation of calcium, it can increase the intracellular accumulation of cyclic nucleotides, block the adenosine receptor as competitive antagonist and it is able to directly decrease the binding of actin to phosphorylated myosin heads of muscles (Welsh et al. 2010; Tazzeo et al. 2012; Chawla & Suleman 2013). These cellular actions lead to bronchodilation, the widening of the airways, and thereby to improved lung function (Welsh et al. 2010).

The inflammatory effect score of EGCG of −1.000 indicated an anti-inflammatory potential, although the positive effect on respiratory function (−0.500) has only been established in one animal study so far (Bani et al. 2006). The polyphenol EGCG is not only an antioxidant but can also enhance endothelial-type and neuronal-type nitric oxide synthase enzymes (Bani et al. 2006). A decline in these enzymes is a feature of inflammation, leading to decreased nitric oxide, which is required to prevent activation of NF-κB and thereby pro-inflammatory gene transcription (Chen et al. 2002; Wheeler et al. 2004; Bani et al. 2006). In different animal and in vitro studies, a negative effect of EGCG on chemokines as well as matrix metalloproteinases (extracellular matrix proteins) has been observed, suppressing collagen production in fibroblasts (Kim et al. 2006; Qin et al. 2011; Chan et al. 2012; Lee et al. 2013).

Still, many of the studied dietary components showed no effect on either inflammatory markers and/or respiratory function. With using the median as cut-off point, the number of weighted articles fell below this established cut-off point. However, the calculated effects of many dietary components were only just above the cut-off points not reaching the “ideal amount” of 20 credits. This shows the need for more well-designed studies to assess the effects of dietary intake on patients suffering from chronic inflammatory lung diseases, with determining not only inflammatory and immunological markers but also the effect on respiratory function.

Limitations

The number of studies exploring the effects of the different dietary components and food items on chronic inflammatory lung diseases was limited and conflicting results were found. Ideally, the number of weighed articles would be above 20 credits. In our study, the cut-off points were established by calculating the median of credits assigned to the group of dietary components, in total 15 credits and 10 credits. For many dietary components with credits ranging between the cut-off point and the ideal number of 20 credits, the effects might be more reliable when they would be corrected for literature robustness as well. To interpret the results of this study that showed possible health enhancing effects for specific components, some further limitations have to be taken into account.

The calculated inflammatory and respiratory effect scores give an indication of the effect of consuming specific dietary components when suffering from chronic inflammatory lung diseases. The calculation is based on the dietary inflammatory index, to measure the inflammatory effect of dietary components (Cavicchia et al. 2009; Shivappa et al. 2014). Shivappa et al. (2014) related the calculated effect to the total dietary intake based on food consumption data. This was not done in the present study since our main interest was to identify potent nutrients or food items affecting chronic inflammatory lung diseases. Not the amounts consumed by the general population, but intake resulting in a positive effect for patients was the main point of interest. Including a variety of experimental setups and abstaining from a quality control of these included studies could introduce an additional bias to this research. However, with the main aim being to identify dietary compounds potent to affect inflammation and/or function, the scores described in this study should serve as a starting point for future research rather than provide a calculation of an effect that can be reached.

By studying the effect of intake of single dietary components, the components most capable of positively influencing the progress of chronic inflammatory lung diseases were identified. However, methods to measure nutrient intake such as food questionnaires can result in measurement errors and bias by under- or overestimation of consumption (Freedman et al. 2011). Conflicting results between studies on similar ingredients can also be explained by heterogeneity, resulting from: (i) variations in the intake of the active ingredient (due to different compositions of the matrix of the active ingredient, the dosage regimen or duration of intake); (ii) set-up and quality of the study; and (iii) variations between patients (in genetics, dietary status, physical activity and dietary intake).

Following food consumption, not only single dietary components potentially affect health but all bioactive components are able to influence disease progress. Therefore the whole diet should be taken into consideration: absorption of bioactive ingredients can be influenced, but more importantly the intake of various bioactive ingredients could result in synergistic effects as suggested with adherence to the Mediterranean diet, a diet high in fruit, vegetable and fibre intake (Jacobs & Steffen 2003; Liu 2003; Jacobs & Tapsell 2007; Romieu et al. 2009).

Finally, publication and selection bias could have influenced the inclusion of 162 articles from over 1000 articles uncovered by the literature review strategy. With many studies showing significant anti-inflammatory or positive function effects, the question can be raised whether studies showing no or negative effects on inflammation or respiratory function are published to a similar extent as studies showing these positive effects, the so-called publication bias. Furthermore, the search strategy and inclusion criteria for this study could have led to selection bias of dietary components or models, although the possibility of this bias was reduced by conducting various searches by different authors and thorough discussion of inclusion and exclusion criteria.

Conclusions

The calculated inflammatory and respiratory function effect scores showed predominantly beneficial influences of various dietary components and food items on inflammatory and immunological responses as well as on lung function in patients suffering from chronic inflammatory lung diseases. Although inflammatory and immunological markers are not the only factors influencing disease progress, the positive effects of n-3 PUFAs and vitamin E on these markers were accompanied by improved lung function scores. Therefore, the consumption of these components could improve quality of life of patients and reduce the need for pharmaceutical anti-inflammatory therapies, thereby reducing both side effects and development of resistance.

Disclosure statement

The authors declare that there are no conflicts of interest.