An Appraisal of the Nutritional Advice Provided to People with Spinal Cord Injuries in the UK

Table of Contents

Abstract

Acknowledgements

Table of Contents

Tables and Graphs

1.0 Introduction
1.1 Spinal Cord Injuries (SCI)
1.2 Unique nutritional needs
1.3 Nutrition for Secondary Health problems
1.3.1 Obesity and chronic diseases
1.3.2 Bowel Management problems
1.3.3 Bladder Management problems
1.3.4 Pressure Sores
1.3.5 Osteoporosis
1.4 Nutritional advice for people with SCI
1.5 Aims and objectives of the project

2.0 Methods
2.1 Study design
2.2 Comparison of Nutritional Advice
2.2.1 Identification of Nutritional Advice
2.2.2 Analysis
2.3 Cross-sectional Survey
2.3.1 Survey Design and Sampling
2.3.2 Analysis
2.4 Ethical considerations

3.0 Results
3.1 Available Nutritional advice
3.2 Consistency of Nutritional advice
3.3 Adequacy and Relevance of Nutritional advice
3.3.1 Adequacy of ‘given’ advice
3.3.2 Adequacy of ‘found’ advice
3.3.3 Perceptions of Adequacy
3.3.4 Adequacy of learnt advice
3.3.5 Relevance of Nutritional advice

4.0 Discussion
4.1 Consistency
4.2 Adequacy and Relevance
4.3 Limitations
4.4 Recommendations and Conclusions
References

5.0 Appendices
5.1 SCI Anatomy and Physiology
5.2 Participant Information Sheet 1
5.3 Questionnaire for Health care Professionals
5.4 Summary Tables of Nutritional advice
5.5 Survey for people with SCI
5.6 Organisations involved in recruitment
5.7 Participant Information Sheet 2
5.8 Verbatim survey responses

Abstract

Introduction: People with spinal cord injuries (SCI) have unique nutritional needs due to their predisposition to obesity and chronic diseases, bowel and bladder related health concerns, pressure sores and osteoporosis. Currently there are no evidence-based nutrition guidelines provided for this neglected group and with approximately 40,000 people with SCI in the UK, this is a cause for concern.

Aim and Objectives: The overall aim of this project was to: appraise the nutritional advice provided to adults with SCI, living in the UK and to identify limitations with this advice. The specific objectives of this project were: (1) To identify the different types of nutritional advice provided to adults with SCI; (2) To appraise the consistency of this nutritional advice; (3) To appraise the adequacy of this nutritional advice and its relevance to people with SCI; (4) To identify the key limitations with the nutritional advice currently provided, in terms of its consistency, adequacy and relevance; (5) To make recommendations, based on these findings.

Methods: Multiple methods were selected: (1) sources of nutritional advice (n=11) were selected by purposive sampling and compared for ‘consistency’ using directed content analysis. Sources included: two responses from a cross-sectional questionnaire completed by health care professionals; and nine sources of textual advice provided by organisations or websites in print, html or PDF format. (2) A cross-sectional, web-based survey was provided to people with SCI (n=69) via convenience sampling and enabled their views to be investigated regarding the ‘adequacy’ and ‘relevance’ of nutritional advice provided to them. Their responses were analysed using simple frequency counts and basic content analysis.

Results: Comparison of the eleven types of nutritional advice revealed many inconsistencies in the way that advice was provided and in the type of nutrients, foods or health conditions which advice was provided for. Analysis of the web- based survey results revealed: inadequacies regarding the nutritional advice that people with SCI had been provided with in terms of its delivery, timing, form, content and extent to which it was approved by people with SCI; it also showed that secondary health conditions were ‘very important’ determinants of food choice for people with SCI, thus potentially encouraging them to follow nutritional advice.

Conclusion: Nutritional advice provided to people with SCI in the UK shows great inconsistency. There is also limited evidence to suggest that it is inadequate but relevant for people with SCI to follow. Due to these findings the health of people with SCI is at great risk. An evidence-based appraisal of the literature is required; greater standardisation and co-operation between organisations providing advice; and a need for research -to help protect the health of people with SCI.

Acknowledgements

I would like to thank everyone who has generously given their time to contribute to my research project: the respondents who filled in my survey; the health professionals, who took the time to fill in my questionnaires; the organisations who were willing to provide a link for my online survey on their websites; friends and family; and the staff in the school of Integrated Health. Most importantly I would like to thank Veronica Tuffrey for her initial support and enthusiasm and Liza Draper for her patience, invaluable advice and for offering to sift through countless e-mails to help me successfully complete my project.

Tables and Graphs

List of Tables

Table 1: Types of Nutritional advice available for people with SCI in the UK

Table 2: A comparison of Nutritional advice provided to people with SCI

Table 3: Nutritional advice actually provided to people with SCI in the UK

List of Graphs

Graph 1: Number of respondents who were GIVEN Nutritional advice to help manage their SCI?

Graph 2: The first source of Nutritional advice, which UK respondents received, to help them manage their SCI

Graph 3: The forms of Nutritional advice that UK respondents have received 30 for their SCI

Graph 4: UK respondents’ interest in the Nutritional advice they have received to help manage their SCI

Graph 5: Who was responsible for arranging the Nutritional advice given to 31 UK respondents for their SCI?

Graph 6: How UK respondents rate the DETAIL and DEPTH of Nutritional 31 advice they have received for their SCI

Graph 7:How UK respondents rate the USEFULNESS of Nutritional advice 32 they have received for their SCI

Graph 8: How UK respondents rate the KNOWLEDGE gained from the Nutritional advice they have received for their SCI

Graph 9: How UK respondents rate the PRACTICALITY of following the Nutritional advice they have received for their SCI

Graph 10: How UK respondents rate the SCI Relevance of Nutritional advice they have received for their SCI

Graph 11: How many UK respondents would have preferred to have received more Nutritional advice for their SCI

Graph 12: Have UK respondents FOUND other sources of Nutritional advice to help them manage their SCI

Graph 13: Other types of Nutritional advice UK respondents have found to 35 help manage their SCI

Graph 14: How UK respondents rate the DEPTH and DETAIL of Nutritional 35 Advice they have found, for their SCI

Graph 15: How UK respondents rate the USEFULNESS of Nutritional advice they have found, for their SCI

Graph 16: How UK respondents rate the KNOWLEDGE gained from the 36 Nutritional advice they have found, for their SCI

Graph 17: How UK respondents, rate the PRACTICALITY of following the 36 Nutritional advice they have found, for their SCI

Graph 18: How UK respondents, rate the SCI RELEVANCE of Nutritional 37 advice they have found, for their SCI

Graph 19: The importance of nutrition and health factors in determining 42 what people with SCI eat

Graph 20: The importance of Nutritional advice relative to other factors in 43 determining what people with SCI eat

I Introduction

1.1 Spinal Cord Injuries (SCI)

The spinal cord is a bundle of nerves, which carry messages to and from the brain in order to control body function, sensation and movement. A ‘spinal cord injury’ (SCI) occurs when the spinal cord is damaged, usually due to a fracture of the vertebral column, which encases and protects the spinal cord (WHO, 1996).

SCI injuries are classified according to the level of the corresponding vertebra at which the injury to the spinal cord occurs. For instance, a C5 vertebral break is called a C5 level injury (See: Appendix 5.1). An SCI paralyses the muscles, causing a loss of ‘sensory function’ (sensation) and ‘motor function’ (movement) according to the level of the injury (WHO, 1996). Furthermore, it disrupts the ‘autonomic nervous system’ which controls metabolism and involuntary body function (Claus-Walker and Halstead, 1982a; Claus-Walker and Halstead, 1982b; Lynch and Frizelle, 2006).

According to the ‘International standards for Neurological and Functional classification of Spinal Cord Injury’, a ‘paraplegic’ is someone who has injured the thoracic, lumbar or sacral region of their spinal cord and retains complete arm function, with varying degrees of trunk and lower body function; and a ‘tetraplegic’ (or ‘quadriplegic’) is someone who has injured the cervical region of their spinal cord, with impaired arm, trunk and lower body function. They also classify SCI by the degree of function lost (See: Appendix 5.1). Someone with a ‘complete’ injury has lost entire movement and sensation below the level of the injury, whereas someone with an ‘incomplete’ injury has retained some degree of motor and/or sensory function below the level of the injury (Maynard et al., 1997).

1.2 Unique nutritional needs

Despite the type of injury, people with ‘chronic SCI’¹ have unique nutritional needs relative to the general population, because they are predisposed to many ‘secondary health conditions’² (Tomey et al., 2005; Phillips and Gater, 2007). These conditions result from their reduced capacity for movement and sensation and/or from the metabolic and hormonal changes caused by disruption to their autonomic nervous system (Claus-Walker and Halstead, 1982b; Lehman, 1995; Kocina, 1997; Bauman et al., 1999; Chen et al., 2000; Garstang and Miller-Smith, 2007).

The key nutritionally-modifiable, secondary health conditions for people with SCI include:

1. Obesity and other chronic diseases
2. Bowel management problems.
3. Bladder management problems:
4. Pressure Sores
5. Osteoporosis

The ways in which these secondary health conditions may affect nutritional needs will now be discussed.

1.3 Nutrition for Secondary Health problems

1.3.1 Obesity and other chronic diseases

People with SCI are predisposed to becoming overweight and obese, because paralysis is thought to decrease their energy requirements, by affecting all three components of energy expenditure (Liou et al., 2005; Phillips and Gater, 2007). Whether paralysis reduces Thermic Effect of Feeding’ (TEF)³ in the SCI population, is disputed (Monroe et al., 1998; Bucholz and Pencharz, 2004). It has been shown however, that people with SCI have a reduced ‘Basal Metabolic Rate’ (BMR)⁴ (Mollinger et al., 1985; Bucholz et al., 2003a) - with their Resting Metabolic Rate £T (RMR)⁵ estimated to be 14-27% lower, compared to the general population (Bucholz and Pencharz, 2004). This is probably due to the atrophy of muscle tissue, caused by paralysis (Nuhlicek et al., 1988; Liou et al., 2005). Paralysis also leads to a reduced capacity for physical activity⁶ the likely reason why people with SCI show reduced Physical Activity Level’s⁷ (PAL) below the 1.70, recommended for optimal health (FAO/WHO/UNU, 2001; Bucholz et al., 2003b).

Analysis of the 1994-1995 National Health Interview Survey (NHIS) in the USA (n=25,626) found that after adjusting for socio-economic factors, people with lower extremity mobility difficulties were 2.5 times more likely to be obese than able­bodied people (Odds Ratio [OR] 2.5; 95% CI 2.3 - 2.7, p>0.05) (Weil et al., 2002). Furthermore, it has recently been suggested that the Body Mass Index (BMI) Indicator, may be underestimating the obesity problem for people with SCI (Jones et al., 2003). This is problematic, since obesity already causes 30,000 deaths every year in the UK (Haslam et al., 2006) and cost the National Health Service (NHS) in England, £480 million in 1998 (Avenell, 2004, p.1).

The increased fat mass of people with SCI, also puts them at risk of chronic diseases such as diabetes and cardio-vascular disease (CVD), of which coronary heart disease (CHD) is said to be their number one killer (Zlotolow et al., 1992; Kocina, 1997; Liou et al., 2005; Myers et al., 2007). Their predisposition to CVD is also exacerbated by abnormal lipid metabolism, as they show lower levels of the protective High Density Lipoprotein Cholesterol (HDL) compared to the general population (Liang et al., 2007; Liang et al., 2008).

A seven year, longitudinal study on men with SCI (n= 361), reported their increased risk of dying from diabetes (Relative Risk [RR]: 2.62; 95% CI: 1.19 - 5.77) and heart disease (Relative Risk [RR]: 3.66 95% CI: 1.77 - 7.78) compared to men from the general population (Garshick et al., 2005); while another study (n=545) showed that people with complete injuries had a 44% excess risk of developing CVD (95% CI: 1.16 - 1.77) (Groah et al., 2001).

People with SCI are often unable to exercise (Haisma et al., 2006), so the role of nutrition is imperative for preventing obesity and chronic diseases (Liou et al., 2005; Phillips and Gater, 2007). Yet, ‘limited educational resources are available on nutritional issues and weight control for this high-risk population’ (Chen et al., 2006, p.83). Furthermore, the literature appears to show inconsistencies.

For instance, a diet low in saturated fats and high in omega-3 polyunsaturated fats has been associated with a reduced incidence of CHD in the general population (WHO, 1990). Although, in the SCI population one study shows no association between dietary intake and serum lipid levels[8] in the SCI population (Zlotolow et al., 1992) and another that supplementation with omega-3 fatty acids has no effect on cholesterol or triglyceride levels (Javierre et al., 2005). Similarly, Cox et al., (1985) suggest that the energy requirements for quadriplegics are 22.7 kcal/kg body-weight/day and paraplegics are 27.9 kcal/kg body-weight/day. Yet Liu et al., (1996) states that energy requirements may be higher, for tetraplegics with pressure sores.

1.3.2 Bowel management problems

Bowel dysfunction, results from the autonomic nervous system disruption caused by an SCI and does vary by type and level of injury (Lynch and Frizelle, 2006). Constipation, irregular defecation and faecal incontinence - which are collectively referred to as ‘neurogenic bowel’ - account for the majority of all gastrointestinal problems reported by all people with SCI (Chung and Emmanuel, 2006).

Due to this loss of voluntary control, people with SCI have to develop specific bowel management programs using stimulating agents such as a ‘suppository, mini-enema, manual digital stimulation, or oral medication’ (Matthews-Kirk et al., 1997, p.59). This takes considerable time and is not full-proof.

A study on SCI subjects (n=161) found that 41% were spending more than one hour, on each episode of bowel evacuation and 33% reported having faecal incontinence at least once a month (Harari et al., 1997). Another study (n=115) reported that bowel management was a cause of significant distress for 54% of SCI subjects (p<0.005) (Glickman and Kamm, 1996). Furthermore, bowel management problems were responsible for 11% of SCI hospital readmissions over a 10 year period in an Australian hospital (Middelton et al., 2004).

Diet manipulation is very important for bowel management due to the effects it can have on the colonic transit time, thus reducing their likelihood of being constipated or susceptible to faecal incontinence (Chung and Emmanuel, 2006; Lynch and Frizelle, 2006). Diet could therefore indirectly promote their health and well-being (Matthews-Kirk et al., 1997). Yet discrepancies exist about ideal nutrition for bowel management. For instance, the benefits of a high fibre diet are generally advised for people with SCI (Banwell et al., 1993; Matthews-Kirk et al., 1997). Yet a study by Cameron et al., (1996) found that increasing fibre intake through the addition of bran, significantly increased the colonic transit time for a group of SCI subjects from 28.2hours to 42.2hours (p<0.05). Requirements may also vary depending on the level of injury, as a study reported that people with lower level injuries required more dietary modification than people with upper level injuries, due to experiencing greater difficulties with bowel management (Yim et al., 2001).

1.3.3 Bladder management problems

Disruption to the autonomic nervous system as a result of an SCI, also leads to ‘neurogenic bladder’, in which control over urination is lost (Chen et al., 2000; Matsumoto et al., 2001). This can result in a bladder that constantly spasms and contracts urine intermittently or a ‘flaccid bladder’ in which the bladder fails to empty (WHO, 1996). Due to such reasons some people have ‘indwelling catheters⁹ ’ (Maynard and Diokno, 1982), but having a catheter or a flaccid bladder can predispose people to ‘Urinary tract infections’ (UTI’s) (Matsumoto et al., 2001). People with SCI are also prone to urinary stones due to their neurogenic bladder and the risk is higher with higher levels of injury (Chen et al., 2000).

In a 50 year follow up of 20 people with SCI, 17 reported UTI’s (Zelig et al., 2000). It is estimated that 7% of people with SCI will develop kidney stones within the first 10 years after their injury, although a higher level of injury is associated with a higher risk (Chen et al., 2000). Furthermore, 36% of people with indwelling catheters develop bladder stones within the first 8 years after their injury (DeVivo et al., 1985). Diseases of the genitourinary tract are said to be the fourth cause of mortality amongst people with SCI (Imai, 2000), such that prevention of these complications, through nutrition is key (Houda, 1993). However, there appears to be conflicting evidence regarding the type of nutritional practice which people with SCI should follow to prevent UTI’s and stones, particularly regarding juice intake.

People with SCI are typically advised to avoid orange juice as it lowers the pH of their urine (Chen et al., 2002). Another study, however, recommends that orange juice may be protective against urinary stones as it is rich in citrate (Wabner, 1993). Cranberry juice is suggested for protection against bladder stones (Chen et al., 2002) and urinary tract infections due to its effect at lowering urine pH (Reid et al., 2001). Although, a Cochrane review suggests that it is not possible to determine the benefits of Cranberry Juice based on the evidence, except for women from the able-bodied population for whom it shows some benefits (Jepson and Craig, 2008).

1.3.4 Pressure Sores

People with SCI are also predisposed to pressure sores - areas of damaged skin, which usually develop on the bony prominences of the body. These are caused by people with SCI ‘sitting or lying in the same position for too long without moving’ (WHO, 1996, p.6) and due to their impaired sensation, they are often unable to feel a pressure sore developing (WHO, 1996).

The point prevalence of pressure sores in a UK community sample of people with SCI (n=472), was 99 (23%) (Raghavan et al., 2003) and a nine year, multi-centre study in the USA (n= 3361) found that pressure sores were more common for subjects with complete injuries (Chen et al., 2005).

If adequate pressure is not relieved, pressure sores can very quickly turn into open wounds taking months to heal (WHO, 1996). They can also cause ‘psychosocial disruptions, with loss of time from work, overwhelming expense, damage to self esteem, and disruptive changes in personal relationships’ (Lehman, 1992, p.110).

It is estimated that treatment of pressure sores costs the USA $1.2 billion a year (Byrne and Salzburg, 1996). In some cases they can also become infected and cause death amongst people with SCI (WHO, 1996).

Pressure sores are associated with nutrition (Meyers, 2000). One study on SCI subjects (n=826) found that eating a healthy diet was associated with a reduced likelihood of pressure sores (Odds Ratio [OR]: 0.77; 95% CI: 0.62 - 0.97) (Krause and Broderick, 2004). Although, there are still discrepancies about what that healthy diet should be. Phillips and Gater (2007) caution against protein nutritional supplements for people with SCI, - as their high-caloric content may encourage obesity - although, they do suggest that their protein requirements may be the same or higher than pre-injury to help prevent pressure sores. A small randomised control trial (n=16), however, found that high-protein/energy supplements did not cause weight gain, nor did they show any effect on pressure sore healing. Although Vitamin C and Zinc supplemented group were the only group whose pressure sores improved over the three week period (Desneves et al., 2005).

1.3.5 Osteoporosis

Finally, people with SCI are predisposed to ‘disuse osteoporosis¹⁰, due to their reduced capacity for weight-bearing and movement, which is brought about by paralysis of the legs (Elias and Gwinup, 2006). It has been reported that all people with SCI have some form of disuse osteoporosis, below the level of their injury (Claus-Walker and Halstead, 1982b).

This type of osteoporosis is serious as it can lead to increased lower extremity fractures which have been reported to occur for people with SCI at an incidence rate of 1 - 34% (Jiang et al., 2006). Increased leaching of calcium from the bones also leads to the increased likelihood of kidney stones, (Elias and Gwimup, 2006), which were discussed previously (see 1.3.3 Bladder Management). This raises debate about the type of nutrition required to prevent leaching from the bones, without increasing the likelihood of kidney stones. Therefore, some sources advise a high calcium diet for maintaining bone density (Chen et al., 2002; Segatore, 1995). While others caution against calcium excess, due to the chance of increasing kidney stones (Houda, 1993).

1.4 Nutritional advice for people with SCI

Due to the secondary conditions just discussed, it seems likely that people with SCI have unique nutritional needs relative to the general population, which if addressed could help improve their health, well-being and chances of survival, as well as saving on health-expenditure costs. What those nutritional needs are, is not so clear, based on the discrepancies briefly outlined in the literature. This wouldn’t normally present a problem, except ‘there are currently no evidence-based guidelines for nutritional support in SCI’ (Phillips and Gater, 2007, p.65).

There are approximately 40,000 people with SCI in the UK (Nichols, 2005). This is a sufficient number to warrant sound nutritional advice especially since Liou et al., (2005) suggest that the prevention of secondary health conditions ‘should be a major component of health care for people with disabilities’ (p.321). However, since there are, no evidence-based nutritional guidelines, to help prevent these secondary health conditions, this raises the question: what type of nutritional advice are people with SCI currently provided with in the UK and how sound is it?

To the best of the researcher’s knowledge, this hasn’t been assessed yet. Thus, an appraisal of the nutritional advice provided to people with SCI, in the UK is necessary on several grounds. Firstly it needs to be appraised for its ‘consistency’, considering that clear discrepancies exist in the literature. Secondly it needs to be appraised for its ‘adequacy’, because at the minimum, people with SCI require access to nutritional advice of a satisfactory standard. Finally, it needs to be appraised for its ‘relevance’, so that pragmatic factors can be identified, which may encourage and prevent people with SCI, from following nutritional advice.

1.5 Aims and objectives of the project

The overall aim of the project is:

To appraise the nutritional advice provided to adults with spinal cord injuries, living in the UK and to identify limitations with this advice.

The specific objectives are:

1. To identify the different types of nutritional advice provided to adults with spinal cord injuries.
2. To appraise the consistency of this nutritional advice.
3. To appraise the adequacy of this nutritional advice and its relevance to people with SCI.
4. To identify the key limitations with the nutritional advice currently provided, in terms of its consistency, adequacy and relevance.
5. To make recommendations, based on these findings.

A few stipulations should be noted regarding the aim and objectives of the project: -The advice ‘provided’ - will include all advice which is able to be accessed by people with SCI from within the UK, including both formal sources ‘provided’ by health professionals as well as nutritional advice ‘provided’ by books, internet etc.

- The project will only appraise the advice provided to adults with SCI who are ‘aged 18-60’, because factors such as growth and ageing may well affect the nutritional advice provided (FAO/WHO/INIT, 1990).
- The project will only focus on the nutritional advice provided to people for management of their ‘chronic SCI’ (which for simplicity, will hereafter be referred to as ‘SCI’), as nutritional advice provided to people for ‘acute SCI’, is likely to be unique (Phillips and Gater, 2007).

2.0 Methods

2.1 Study Design

It was possible to address the objectives of this project ‘from a number of different perspectives’ (Spicer, 2005, p.294). A predominantly qualitative approach was undertaken, so that nutritional advice could be appraised from the point of view of those being studied’ (Hudelson, 1994, p.62). However, the design and analysis were approached in a systematic, quantitative manner enhancing ‘scientific rigour’ (Mays and Pope, 1995, p.109).

Multiple methods were chosen, to ‘appraise’ the suggested nutritional advice. This enhanced validity as a crude form of ‘methodological triangulation¹¹ ’ (Hudelson, 1994, p.54). This also introduced time constraints, which along with other ‘pragmatic factors’ (Spicer, 2005, p.302) involving logistics and resources, restricted the overall project study design and ‘generalisability¹² ’ of findings (Green and Thorogood, 2004).

The overall study design comprised of two main parts:

Firstly, the types of nutritional advice provided to people with SCI were identified by performing a web-based search, speaking to general informants and from administering questionnaires to UK health professionals. The key messages from these sources of nutritional advice were extracted by directed content analysis and compared, which enabled the ‘consistency’ of nutritional advice to be appraised (see: 2.2 Comparison of Nutritional Advice).

Secondly, a cross-sectional survey was designed and implemented for adults with SCI to complete, regarding the nutritional advice they had been provided with.

Their responses were analysed by simple content analysis and frequency counts, which enabled the ‘adequacy’ and ‘relevance’ of nutritional advice, to be appraised from their perspective (see: 2.3 Survey for people with SCI).

The findings from these two parts revealed limitations with the current nutritional advice provided, on which recommendations were then based.

2.2 Comparison of Nutritional advice

2.2.1 Identification of Nutritional advice

To enable the ‘consistency’ of nutritional advice to be appraised for people with SCI in the UK, sources of nutritional advice were identified and selected, only if they could be accessed from within the UK. Due to time restrictions only eleven sources were selected for analysis, which limited the generalisability of findings.

UK Spinal Units

Health professionals in spinal units were a likely source of nutritional advice, as they were a first contact point for people with SCI, immediately following their injury (Houda, 1993). Two dieticians and a head nurse were contacted from two independent UK spinal units, by ‘purposive sampling’ (Bowling, 2002, p.187). Face to face interviews were not possible due to limited consultation time. Instead, a participant information sheet (see: Appendix 5.2) and link to an open-ended, web- based questionnaire (see: Appendix 5.3) were sent to them via e-mail. The questionnaire was designed on SurveyMonkey.com® (Portland, Oregon USA) - ‘an inexpensive online tool that is easy to use, with great flexibility and has the ability to collect information in a timely fashion’ (Collier et al., 2005, p, AB5). It was favoured for its convenience and because it was programmed, such that the questions could not accidentally be skipped (Rhodes et al., 2003).

Two of the health professionals filled in the questionnaire, which asked them to describe the type of verbatim nutritional advice they would provide to people with chronic SCI - according to the nominated nutrient, food group or health condition listed. These questionnaire categories served as prompts and were chosen after scanning other textual sources to see how nutritional advice was usually given to people with SCI. The third participant, was unable to fill in the questionnaire, but volunteered an SCI nutrition advice booklet - which was usually provided to people with SCI, in UK spinal units. The two questionnaires and booklet were thus included for analysis.

Respected Organisations

Respected health organisations and key SCI support networks were identified as probable sources of nutritional advice for people with SCI outside of a spinal unit. Three organisations were recommended by SCI general informants and their print- based sources of nutritional advice were identified and also included for analysis.

Web-based Sources

Due to the increasing range of ‘electronic assistive devices,’¹³ for people with disabilities, the internet was recognised as a potentially accessible source of health advice for people with SCI (Houlihan et al., 2003). Internet-based nutritional advice was sourced using a search strategy, akin to what a lay person might use. The terms - ‘nutrition AND "spinal cord injuries”’ were searched in Google.co.uk® and the first five relevant sources selected. They were selected irrespective of the country of origin, as people with SCI would not necessarily discriminate nutritional advice by country of source.

Some of these sources were not restricted to a single web-page or had additional links to other sources of nutritional advice within their text. This additional advice was thus identified and grouped with the original source, as a person with SCI would probably treat these as one source of nutritional advice. These five sources of nutritional advice were also included for analysis.

2.2.2 Analysis

The eleven sources of nutritional advice identified above (see: Table 1), were compiled and systematically reduced into ‘data display’¹⁴ tables (Miles and Hubeman, 1994) using ‘directed content-analysis¹⁵ ’ (Hsieh and Shannon, 2005).

This method was appropriate because the categories were pre-defined from the sources of nutritional advice (Hsieh and Shannon, 2005). Displaying the data in tables facilitated comparison so that the ‘consistency’ of nutritional advice was ‘appraised’ with rigour and ease (Mays and Pope, 1995)

The pre-defined categories for the coding process were identified as the nutrients, food groups or health conditions, on which nutritional advice for people with SCI were based. Each category was assigned a data display table and excerpts of un­edited nutritional advice relating to that category were then inserted into the row corresponding to the source of that advice. Qualitative analysis requires ‘imagination’ (Green and Thorogood, 2004, p.174), which could have introduced subjectivity in the coding process - but a second coder was not commissioned to enhance reliability, as this was meant to be a piece of independent research.

Tables were grouped together if nutritional advice for similar categories was limited. Repetitive advice was condensed and overly specific details omitted. Data reduction produced twenty-one summary tables, with only the key nutritional messages and rationale retained for each piece of advice. These summary tables (see: Appendix 5.4) were then compared to see if nutritional advice was consistent according to nutrient, food group or health condition. A purposively sampled selection of key inconsistencies which had emerged from this analysis, were presented in a meta-table (see: Table 2), so that the ‘consistency’ of nutritional advice could be easily visualised and appraised. The meta-table was displayed, using a similar layout to those presented by Cannon (1992) in his comparison of nutritional reports from around the world.

2.3 Cross-sectional Survey

2.3.1 Survey Design and Sampling

A cross-sectional, web-based survey was designed and implemented for participants with SCI, so that the ‘adequacy’ and ‘relevance’ of the nutritional advice they had been provided with, could be appraised from their perspective. To assess the ‘adequacy’ of this advice, participants were asked to relay the types of nutritional advice they had been given or found themselves; their satisfaction and approval of this advice; and the key nutritional messages they learnt from this nutritional advice. To investigate the ‘relevance’ of this advice, they were asked to relay the degree to which this nutritional advice influenced their food choice in comparison to specific factors. These factors were identified from papers by Tomey et al., (2005) and Phillips and Gater (2007) for example money, accessibility and often being reliant on a Personal Assistant (PA).

Survey Design

The web-based survey was created using SurveyMonkey.com® (see: Appendix 5.5) and favoured for several reasons. It enabled the ‘hidden’ SCI population of 40,000 to be reached (Schleyer and Forrest, 2000). It generated more responses within the time constraints of the project, than postal surveys or semi-structured interviews would have allowed (Schleyer and Forrest, 2000; Rhodes et al., 2003). It was also less intrusive (Wyatt, 2000) and since respondents could respond anonymously and voluntarily this limited ‘contextual bias’ ¹⁶ (Hudelson, 1994, p.5).

Finally, it was certified to be ‘fully accessible’ for people with disabilities (Survey Monkey, 2008), thus encouraging people to participate, regardless of their dexterity, due to assistive-device technology (Houlihan et al., 2003). However, because SurveyMonkey.com® restricted survey design to a maximum of ten questions, this limited the style of questions asked and depth of data generated.

Quality control was incorporated into study design: only one response was allowed from a single IP address, which minimised the possibility of multiple entries (Rhodes et al., 2003); the survey questions comprised of mostly close-ended questions and Likert scales[17] to encourage participation (Edwards et al., 2007) despite the possibility of ‘response bias'[8]’ (Bowling, 2005, p.155); and the survey was pilot-tested by two personal contacts with SCI, to ensure their understanding of key concepts (Gillham, 2000, p.35). Their advice reduced the likelihood of ‘cultural reinterpretation’^{18 19 20} errors (Hudelson, 1994) which could have been introduced by the researcher through a lack of ‘reflexivity’²¹ (Seale, 1999).

Sampling Frame

SCI affiliated organisations who appealed to the interests of people with SCI were identified (see: Appendix 5.6). Permission was gained to post a link to the survey on their websites, along with a description of the study, equivalent to a ‘participant information sheet’ (see: Appendix 5.7). Voluntary participants were thus recruited

by ‘convenience sampling’ (Bowling, 2002, p187) if they came across the link. This limited the generalisability of the findings. It also may have introduced bias, because cross-checking by ‘respondent validation²² ’ was not possible due to the anonymity of respondents (Schleyer and Forrest, 2000). It was thus not certain they had been met the relevant ‘inclusion criteria’ as requested on the web-page:

1) had chronic SCI; 2) were a UK resident; 3) had their SCI when they were aged 18-60.

2.3.2 Analysis

Respondents were filtered from the cross-sectional, web-based survey according to the following exclusion criteria: 1) If they had ‘only’ visited spinal cord units outside of the UK; 2) if they had their SCI in 2007 (as they could have acute SCI).

Analysis of close-ended responses

The close-ended responses from the remaining survey results (n=69) were analysed using simple frequency counts and presented as graphs (see: Graphs 1­20), to enhance their visual acuity (Gillham, 2000). It was decided not to stratify the results according to sex, level of injury and time of SCI, due to the small sample size. Nor were confidence intervals included as this may have lead to the false impression that the results could be generalised to the wider population.

[...]

---

¹ Chronic Spinal Cord Injuries - Life-long SCI, ‘defined as 1 or more years post-SCI’ (Grandas, 2005, p.2)

² Secondary Health Conditions - are defined as ‘preventable medical, emotional, or social problems resulting directly from an initial disabling condition’ (Liou et al., 2005, p.321)

³ Thermic Effect Feeding (TEF) is the energy required for food ingestion and digestion, which makes up approximately (FAO/WHO/UNU, 2001)

⁴ Basal Metabolic Rate (BMR) is the minimum energy needed to run the body and stay alive (FAO/WHO/UNU, 2001)

⁵ Resting Metabolic Rate (RMR) is the alternative to measuring BMR as the latter is difficult to measure clinically. RMR is ~3% higher than BMR (Goran and Astrup, 2002).

⁶ Physical Activity is the energy expended through obligatory and optional physical activities, (FAO/WHO/UNU, 2001)

⁷ Physical Activity Level (PAL) - is the Total energy expenditure ‘for 24 hours expressed as a multiple of BMR’ (FAO/WHO/UNU, 2001, p.9)

⁹ Indwelling catheterisation is when a standard catheter tube is inserted to the bladder through a hole in the abdomen, or through the genitals and left there to help drain urine regularly from the bladder into a plastic leg bag which can be emptied through a valve. A catheter is usually changed once a month (WHO, 1996)

¹⁰ Disuse osteoporosis - is defined ‘as a decrease in the amount of normal bone tissue that occurs as a result of immobilization’ (Elias and Gwinup, p.163)

¹¹ Methodological triangulation - is when ‘multiple data collection methods are used to study a single problem or research questions’ (Hudelson, 1994, p.54)

¹² Generalisability - describes ‘the extent to which findings from a study apply to a wider population’ (Green and Thorogood, 2004, p.197)

¹³ Electronic assistive devices - enable people with mobility impairments to access computer hardware and software e.g. voice-activation technology, which enables people to type without use of a key-board (Houlihan et al., 2003).

¹⁴ Data Display- is a method of reducing unedited text into ‘a visual format that presents information systematically’ (Miles and Huberman, 1994, p.91)

¹⁵ Directed Content Analysis -uses pre-defined codes, revealed from research findings, to form the necessary categories by which qualitative data is reduced (Hsieh and Shannon, 2005).

¹⁶ Contextual bias - is introduced by contextual factors such as a taboo topic or the presence of an interviewer which affects the way a participant might respond (Hudelson, 1994, p.5)

¹⁷ Likert scale - people have to choose ‘from a series of opinion statements about an issue’ (Bowling, 2002, p.289)

¹⁸ Response style bias - is introduced by close ended questions such as rating scales, in which respondents favour one category or style of response (Bowling, 2002, p.155)

¹⁹ Cultural reinterpretation - occurs ‘when questions are not meaningful to respondents in the way intended by the researcher’ (Hudelson, 1994, p.5)

²⁰ Reflexivity - is a validity check in which the researcher is mindful as to the ways they may have biased methodology, due to ‘prior assumptions and experience’ (Mays and Pope, 2000, p.51)

²¹ Respondent validation - ‘includes techniques in which the investigator’s account is compared with those of the research subjects to establish the level of correspondence between the two sets’ (Mays and Pope, 2000, p.51).