Diet quality in cystic fibrosis – associations with patient reported outcome measures and enablers and barriers to eating a healthy diet: A protocol paper for a mixed methods study

Background

Cystic Fibrosis (CF) is an autosomal, recessive disorder, arising from defects in the CF transmembrane conductance regulator (CFTR) gene (Riordan et al., 1989) and affects many aspects of function and daily living for people with CF (PWCF) (Elborn, 2016; Sawicki et al., 2011). Ireland has the highest incidence of CF worldwide, and approximately 55% of the Irish CF population are in adulthood. This number is anticipated to increase to 65% by 2025 (Burgel et al., 2015; Farrell, 2008).

It has long been established that PWCF have increased energy requirements due to recurrent pulmonary infections, higher resting energy expenditure from increased work of breathing, and malabsorption (Kawchak et al., 1996). In the 1960s, a low-fat diet was advocated to manage steatorrhea, which resulted in severe malnutrition and growth failure. In a landmark study, a high-fat diet with aggressive pancreatic enzyme replacement therapy (PERT) resulted in improved growth and survival (Corey et al., 1988). This has prompted the long-standing clinical practice of prescribing unrestricted high-calorie, high-fat diets with PERT to meet gender specific body mass index (BMI) goals (Turck et al., 2016). This high energy, high fat ‘CF diet’ contributed to marked improvements in survival and prognosis (Borowitz et al., 2002; Yankaskas et al., 2004). However, in recent years and with the advancement of genetic modifiers, concerns have emerged in terms of the reliance on energy-dense, nutrient poor foods that tend to be higher in saturated fat, sugar, and salt. In the general population, poor quality diets lack essential nutrients and from a public health perspective are aetiologically linked with diet-related chronic diseases (WHO & FAO, 2003).

Dietary recommendations in CF are developed based on the best available evidence and expert consensus at that time. Substantial changes in CF treatment and consequent improvements in life expectancy require that guidelines are continuously updated. The European Guidelines advise ad libitum consumption of high-fat foods when weight gain is necessary (Turck et al., 2016). However, with remarkable improvement in survival, patients with CF may be at risk of developing cardiovascular disease, obesity, and other age-related conditions associated with a high consumption of saturated fat and sugar (Hanna & Weiner, 2015; Panagopoulou et al., 2014). In paediatric studies, it is reported that children with CF achieve elevated energy requirements through significantly higher intakes of energy-dense, nutrient poor foods (Sutherland et al., 2018). Children with CF also have micronutrient intakes significantly lower than healthy controls, which has been associated with CF related chronic conditions (e.g., impaired bone health and CF related diabetes) (Matel & Kerner Jr, 2012; Parkins et al., 2011).

CFTR modulator therapy targets the defective CFTR gene, resulting in improvements in lung function in PWCF that have specific mutations (Eckford et al., 2012). Ivacaftor (Kalydeco®; Vertex Pharmaceuticals), the first CFTR modulator therapy for individuals with the G551D mutation, along with improving chloride ion transport has been shown to improve duodenal alkalinity (Rowe et al., 2014). This effect has been documented to continue throughout the distal intestine and resembles alkalinity levels in non-CF individuals (Gelfond et al., 2013). The normalisation of gastrointestinal (GI) pH is potentially a key feature in the assimilation of food and nutrients (Rowe et al., 2014). Furthermore, CFTR modulator therapies reduce the risk of lung infections and are associated with improved lung function in PWCF (Cox et al., 2010; Zemanick et al., 2013). Improved lung function and lower risks of infection can reduce resting energy expenditure in PWCF (Matel & Milla, 2009; Milla, 2007).

The use of CFTR modulator therapies has resulted in the emergence of overweight and obesity, which will undoubtedly become a more frequent feature in CF. With overweight and obesity in PWCF more than doubling from 14.3% in 2000, to 37% in 2020 (CFF, 2021), there is an evident upward trend (1.1% increase per annum) towards the general population’s overweight and obesity prevalence rates of 39.0% and 13.0%, respectively (Statista and WHO). Furthermore, from 2018 to 2020, there was an increase of 7.0% in overweight and obesity cases in PWCF (CFF, 2019; CFF, 2021), highlighting the acceleration in recent years within this population, coinciding with increases in the use of CFTR modulator therapy (Solomon & Mallory, 2021). Higher low-density lipoprotein (LDL) cholesterol and triglyceride levels have been found in PWCF who are overweight, which may be influenced by diet related factors (Loria et al., 2014). A recent review on the historical perspective of dietary intake studies in children with CF stated that ‘the nutrition dogma has to be extended from nutrition for growth and survival to nutrition for health and well-being’ (Sutherland et al., 2017). This call for practice change needs also to extend to adults with CF.

In the general population, an emphasis on diet quality above individual macronutrients is the basis of dietary recommendations to reduce chronic disease risk (You, 2015). Certain dietary patterns are associated with beneficial health outcomes including the Mediterranean and Dietary Approaches to Stop Hypertension (DASH) diets (Bach-Faig et al., 2011; Wengreen et al., 2013). Many diet quality index tools have been developed and validated, albeit, not in the CF population. The Diet Quality Index – International (DQI-I) was constructed to provide a global tool for monitoring healthfulness of diet and for exploring aspects of diet quality. The DASH score has also been previously used within the Irish population and is based on the intake of nine food groups: wholegrains, fruit, vegetables, legumes, low-fat dairy foods, red processed meat, sweetened snacks and beverages, salty snacks, and sodium consumption (Harrington et al., 2011). Calculation of the Dietary Inflammatory Index (DII®) is another novel method for assessing the theoretical inflammatory impact of a diet (Shivappa et al., 2014). Whilst attaining disease specific nutrient targets in CF is important, a composite measure of representing whole of diet quality is preferred. However, evidence-based research is lacking to guide CF-specific nutritional recommendations or to promote an optimal dietary pattern.

Few studies have investigated perceptions and experiences of diet with adult PWCF. A recent UK study of adult PWCF (n= 9) uncovered participants had each developed a plan to regulate personal nutritional needs, which included eating well and choosing foods of higher caloric value, motivated by the opinion that better weight will certify better health (Barrett et al., 2021). Furthermore, another UK study involving adult PWCF (n= 10) sharing childhood experiences of diet, discovered while some participants embraced a high calorie, high fat diet, others pursued a diet with balance that would benefit long-term health (Cave & Milnes, 2020).

More research is needed to evaluate if a diet that meets the increased energy requirements of PWCF while maintaining the nationally recommended proportions of food groups is feasible and implementable in Ireland. Thoughts on healthy eating and willingness to change habitual dietary patterns to be more healthful whilst still maintaining overall energy targets is required from the patient population prior to any intervention or practice change being employed. Compliance with the ‘CF diet’ is generally poor with estimated adherence rates between 40-55% (O'Donohoe & Fullen, 2014). With social, behavioural, and physical factors influencing the ability of individuals with CF to follow dietary recommendations (Brown et al., 2018), exploring the role that these factors play in the consumption of typical dietary patterns and what strategies would be most effective for improving diet quality will determine next steps in research programme planning. Within this, there is growing evidence that patient-reported outcome measures (PROMs) can serve as valuable indicators for the benefit of interventions and the impact on health-related quality of life (HRQoL), as they represent direct measures of how a patient feels and functions (Solem et al., 2016).

Aims and objectives

The overall aim of the study is to provide a comprehensive evaluation of habitual dietary intakes, dietary patterns, and the overall diet quality of PWCF living in Ireland and to investigate the interrelationships between diet quality and patient reported outcome measures. Through online focus groups (OFGs), the experiences and views on nutrition and diet in this cohort, including drivers of food and dietary choices, and investigating enablers and barriers for dietary changes will also be explored.

Study design

This study is an observational, cross-sectional study with a mixed methods approach. Participant eligibility will be confirmed using a screening questionnaire. Quantitative data will be collected using a demographic and self-reported health questionnaire, three-day food diary, semiquantitative food frequency questionnaire (FFQ), the Cystic Fibrosis Questionnaire-Revised (CFQ-R), EuroQol 5-dimension 5-level questionnaire (EQ-5D-5L), Patient Assessment of Upper Gastrointestinal Symptoms questionnaire (PAGI-SYM) (Rentz et al., 2004; Revicki et al., 2004; Wyrwich et al., 2010), Patient Assessment of Constipation Symptoms questionnaire (PAC-SYM) (Frank et al., 1999) and an originally designed post-participation questionnaire (PPQ). Qualitative data will be collected via OFGs and semi-structured interviews, designed to develop discussions amongst participants on experiences and beliefs related to diet, and the impact CF and CFTR modulator therapy plays. The STROBE reporting guidelines will be adhered to for this observational study (Cuschieri, 2019; Vandenbroucke et al., 2007). The three-day food diary, PPQ, focus group script and screening form can be found as Extended data (Greaney et al., 2022).

Study population

Participant recruitment

The aim is to recruit 100-180 PWCF for this observational study using a convenient sampling method (Etikan et al., 2016). Recruitment will be carried out through Cystic Fibrosis Ireland (CFI) membership forums, CF hospital clinics, and online platforms (e.g., Instagram; Facebook; Twitter; LinkedIn). The gold standard method for investigating sample size sufficiency in qualitative research is a saturation point, and thus, will be used to justify the sample size of the OFGs and semi-structured questionnaires. Data collection and analysis will be conducted concurrently and when saturation has been achieved, described in thematic analysis as when no new codes are identified in participants accounts, recruitment will cease (Vasileiou et al., 2018). From previous qualitative studies, saturation commonly occurs near 30 participants, which set the target sample size for this study. The goal is to run four to five OFGs with six participants in each group (total n= 24-30).

Participant eligibility

Inclusion criteria: To be included in this study, participants must be adults ≥18 years of age with a diagnosis of CF, living in Ireland. Exclusion criteria: Participants will be excluded from the study if they are receiving enteral nutrition, following a prescription diet for another medical condition (e.g., coeliac disease, are pregnant) or are post-transplant. Adults must be on a stable medical regimen for at least four weeks prior to commencing the study with no recent pulmonary exacerbations involving the administration of oral or intravenous antibiotics or glucocorticoids.

Data collection

Both quantitative and qualitative data will be collected as part of the study. Participants will be asked to complete a demographic and self-reported health questionnaire offered in online and paper-based versions. Patient-reported clinical and demographic data will be collected to describe the interview sample, including age, sex, most recent BMI, and lung function (forced expiratory volume in 1 second (FEV₁%)); CF-related comorbidity status (e.g., presence of CF-related diabetes, liver disease) and medications including the use of genetic modulator agents.

As part of the quantitative dietary assessment, participants will be advised to complete an estimated three-day food diary, either paper-based or using the smartphone application Libro from the Dietary Analysis Software Nutritics, Version 5.7 Research Edition (Nutritics, RRID:SCR_022154) (Nutritics Ltd.: Dublin, Ireland). Participants will also complete a semi-quantitative FFQ. The FFQ is adapted from the European Prospective Investigation of Cancer (EPIC) FFQ (Harrington et al., 2008), validated for use in the Irish general population (Harrington, 1997).

The DQI-I and the DASH score will be used to derive diet quality scores for each participant (Fung et al., 2008; Harrington et al., 2011). DII® has been developed to collate all the individual food or nutrient intake components that have a known effect on markers of inflammation (Shivappa et al., 2014). Based on the intake of each of these components in the index, an individual’s diet can be assessed for its pro- or anti-inflammatory potential.

PROMs used in the study include the following: CFQ-R, a disease-specific, validated tool that encompasses general domains of HRQoL as well as domains specific to CF (FDA, 2009; Henry et al., 2003; Modi & Quittner, 2003; Quittner et al., 2005). The universal evaluation tool, EQ-5D-5L, produces a single index measurement of health status and elicits means of comparability for treatments of differing medical conditions (Eidt-Koch et al., 2009; Rabin & Charro, 2001; Solem et al., 2016). The 20-item PAGI-SYM and the 12-item PAC-SYM are validated instruments for monitoring abdominal symptoms (Rentz et al., 2004; Revicki et al., 2004; Slappendel et al., 2006).

Participants will be invited to take part in an OFG, or semi-structured interview (max. one hour) facilitated by a trained, experienced dietitian, conducted over Microsoft Teams. OFGs will be conducted in groups of four to six. However, if participants prefer to be interviewed face-to-face or by telephone this will be facilitated. OFGs and interviews will be video, and audio recorded for later transcription and analysis. The readability of OFG questions will be measured using the Flesch-Kincaid Readability Test, a validated test that produces a value based on readability level and education level (Flesch, 2007; Heydari, 2012). The questions designed for the OFGs/interviews will cover three broad topics: views on diet advised/promoted for PWCF; current consumption patterns; enablers and barriers to eating a healthier diet. Once all topics have been discussed, participants will be invited to complete the PPQ administered in the video call chatroom or via email, to facilitate engagement. The PPQ was developed on Microsoft Forms and will be used to gather quantitative information to reinforce beliefs of PWCF articulated in the OFGs and semi-structured interviews.

Data management

The primary researcher will be responsible for the storage of hard copies of signed consent forms and questionnaires, which will be kept in secure filing cabinets in line with the 2018 Data Protection Act (Data Protection Act, 2018). Electronic data will be saved on a cloud-based password-protected database. Data will be anonymised prior to data extraction. The Electronic Data Capture system (EDC) Castor, Version 1.6 (Castor – Electronic Data Capture System RRID:SCR_022150) (Ciwit B.V., The Netherlands) a cloud-based, clinical data management platform will be used to store all quantitative data collected. Personal data will only be available by key researchers involved in the study and will be anonymised before being entered into a dataset to be analysed. After the dataset has been analysed and the report/paper written, the dataset will be preserved for seven years before being destroyed. Any participants who are interested in the results of the research will be flagged and both individual and overall results will be disseminated after study completion.

Data processing and analysis

All statistical analyses will be conducted in SPSS® Statistics for Windows, Version 28 (IBM SPSS Statistics, RRID:SCR_016479) (IBM Corp., Released 2019). Data will be presented as means ± standard deviations (SD), medians and interquartile ranges (IQR), and frequencies (n and %) as appropriate.

Macronutrient and micronutrient intakes will be quantified as absolute intake and energy-adjusted intake (amount per 1000kcals). Appropriate descriptive statistics will be used to describe the baseline characteristics of the cohort. Distribution of data will be tested using the Kolmogorov Smirnov test of normality, where a significance of p > 0.05 will identify data as being normally distributed. P-values will be considered significant if the p-value shows a result of p < 0.05. Independent t-tests (or Mann Whitney for non-parametric variables) will be used to compare continuous variables between participant grouping of two categories (e.g., sex differences). One-way analysis of variance (ANOVA) (or Kruskal-Wallis for non-parametric variables) tests will be applied to compare continuous variables between participant groupings of three or more categories (e.g., age categories) and Post Hoc Tukey tests will be performed to assess differences between specific categories. Pearson's Chi-square will be used to test differences between categorical variables. Total diet quality scores and scores for each component of the score will be calculated and compared between demographic groups using appropriate statistical tests as previously described. Associations between diet quality, participant demographic variables and patient reported outcome measures will be investigated. Pearson or Spearman rho correlations will be used to investigate associations between diet quality scores and DII® scores. Multiple (linear) regression models will be performed to assess what variables most strongly predict increased diet quality scores and an increased dietary inflammatory index. Dietary patterns will be identified from the FFQ data by principal components analysis.

A qualitative inductive thematic analysis method described by Braun and Clarke (2006) will be completed using NVivo^® (NVivo, RRID:SCR_014802) (QSR International Pty Ltd.), a qualitative data analysis computer software package, to analyse data collected from the OFGs and semi-structured interviews. Previous research has found this tool useful in organising large amounts of data (Castleberry, 2014; Castleberry & Nolen, 2018). The approach to thematic analysis by Braun and Clarke (2006) consists of six-phases: 1) data immersion; 2) generating codes; 3) generating themes; 4) reviewing potential themes; 5) defining and naming themes; 6) producing a report.

Transcripts from each OFG and interview will initially be read several times to immerse the researcher in the data, a preparation phase prior to further analysis to familiarise the researcher with descriptions and language of participants (Braun & Clarke, 2006). Following this, the researcher will code the data inductively in NVivo^® and a second researcher will also code the transcripts. After refining of codes is complete, subsequential themes will be actively generated supported by discussion with the wider research team.

Ethical approval and informed consent

This study has received funding by the Health Research Board (HRB) and Health Research Charities Ireland (HRCI) grant initiative with CFI. Ethical approval has been granted by University of Limerick Education and Health Sciences, Research and Ethics Committee, University Hospital Limerick (Ref: 090/2021; Date of Approval: 22^nd September 2021) and Cork University Hospital (Ref: ECM 4 (o) 11/01/2022& ECM 3 (bb) 22/02/2022; Date of approval: 26^th January 2022), and University Hospital Galway Research and Ethics Committee (Ref: C.A. 2709; Date of approval: 10^th March 2022) for all aspects of the study. Once PWCF volunteer to participate in the study a screening questionnaire will be used to determine participant’s eligibility. Once completed, all eligible participants will be provided with a participant information form and consent form (PICF) to read and sign prior to taking part in the study. Depending on participant preference, all questionnaires will be offered in either an online or paper-based version. All participants will be informed that participation is voluntary, and they can withdraw at any point during the study. Once written consent is obtained, participants are enrolled in the research study. Verbal consent will be requested for recording of the OFGs or semi-structured interviews during each introduction. This study is in compliance with the Declaration of Helsinki (WMA, 2014).

Data dissemination

A report on the findings will be prepared and presented to the funding body HRB, CFI, and to national and international project collaborators. Findings will also be presented at the CFI annual conference. Manuscripts will be prepared for publication and submitted to relevant journals. Abstracts will be submitted to conferences.

Study status

Currently the research team are undergoing recruitment of participants via CFI membership forums, CF hospital clinics and online platforms.

Underlying data

No data are associated with this article.

Extended data

Figshare: Diet quality in cystic fibrosis study. https://doi.org/10.6084/m9.figshare.c.5964891. (Greaney et al., 2022).

This project contains the following extended data:

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).