Seasonal, ethnic and gender variations in serum vitamin D₃ levels in the local population of Peterborough

Abstract

Vitamin D is a fundamentally important hormone involved in calcium absorption, bone mineralization and parathyroid hormone production. Vitamin D deficiency may result in a myriad of diseases, such as osteomalacia, rickets and has more recently been identified as a risk factor for diabetes. Accurate determination of serum vitamin D levels is, therefore, paramount when assessing an individual for related pathologies against an appropriate reference range for interpretation. The purpose of this study was to assess vitamin D status between ‘healthy’ Caucasian and ‘healthy’ Asian populations of Peterborough, taking into account seasonal serum vitamin D variation. In addition, we evaluated whether a reference range for serum vitamin D of the local population should be race and/or seasonal specific. Using a Chromsystems high-pressure liquid chromatography vitamin D₂/D₃ methodology upon Agilent 1100 hardware, serum vitamin D status was assessed in 200 subjects of varying age, gender and ethnicity using summer (n = 106) and winter (n = 94) cohorts. Serum vitamin D₃ levels were significantly lower (up to 70%; P ≤ 0.0001) in both men and women of the Asian population in comparison to the Caucasian population during both the summer and winter periods. Vitamin D₃ levels of the Caucasian cohort displayed significant variation between summer and winter (P≤0.0001), while the overall Asian population displayed no significant seasonal variation in vitamin D status. The reference range produced by the Caucasian cohorts (8.2–53.7 µg/l) complements published studies, while the Asian cohort displayed significantly lower limits (3.6–26.7 µg/l). Currently, no genetic predisposition to lower vitamin D status in Asians compared with Caucasians has been established. Therefore, the Caucasian range was implemented for all ethnicities, as this conforms to the national consensus of deficiency within the UK. This reference range indicates that 34% of the ‘healthy’ Asian community are vitamin D deficient compared with 2% of Caucasians. Further research is required to increase vitamin D deficiency awareness in Asian communities and highlight the potential role of supplementation.

Vitamin D₃, Asian, seasonal variation, gender, reference range

Introduction

Vitamin D is a fat soluble steroid hormone present in the body as either ergocalciferol (D₂) or cholecalciferol (D₃). The biochemical difference being a double bond between carbons 22 and 23 and a methyl group upon carbon 24 of D₂. A minority of vitamin D is absorbed in the gut from food products as D₂/D₃. The majority of vitamin D is produced in vivo as D₃ when skin is exposed to the ultraviolet B wavelength of sunlight and this pathway produces more than 90% of the body's vitamin D.^{1, 2}

Vitamin D is an important component in many bodily processes from molecular to systemic roles and is different from other vitamins in that it may be described as a hormone. Roles include vital functions in mineral calcium and phosphate homeostasis, cell differentiation, regulation of parathyroid hormone inhibition, T-cell suppression and macrophage deregulation, cancer pathogenesis and erythropoiesis.^2–9 Pathologies traditionally associated with a loss of vitamin D homeostasis are due to hypercalcaemia (hypervitaminosis D) or hypocalcaemia (hypovitaminosis D), which causes rickets and osteomalacia.¹⁰ However, current research is now identifying new risk factors for inadequate vitamin D levels such as diabetes mellitus (types I and II), multiple sclerosis, ischaemic heart disease, tuberculosis in Asians and effects on postnatal head and linear growth.¹¹

Currently, to investigate the vitamin D-related disorders the metabolite 25-hydroxyvitamin D₂/D₃ is routinely measured as this has the longest half-life in vivo of all the significant metabolites. As the D₃ metabolite represents the majority of vitamin D in vivo there has been much research upon the effectiveness of its production. Many factors are associated with impacting the rate of D₃ biosynthesis, such as the amount of time and skin area exposed to sunlight, the time of day exposed (early morning sun is weaker than during mid-afternoon), the latitude upon the globe, season (particularly latitudes of >50° North or South of the equator) and skin pigmentation (melanin absorbs UVB photons).^{2, 12–14} Cultural and religious dress (e.g. burkhas) are also shown to contribute to lower vitamin D₃ production in specific communities.¹⁵

In Greater Peterborough, a significant proportion of the population is formed of ethnic Asians. Previous studies have noted distinct variations in serum vitamin D concentrations of Caucasian and Asian populations living at Northern latitudes.^{16, 17} This study will aim to evaluate the difference between serum vitamin D levels of the ‘healthy’ Caucasian and ‘healthy’ Asian populations of Peterborough, taking into account seasonal variation and differences between genders. This data will then be used to assess the possible causes of any variation, and the need for a race and/or seasonal specific reference range for serum vitamin D interpretation of the local population. The requirement for a local reference range is to allow for the accurate assessment of vitamin D status, as there is currently no national serum vitamin D reference range or international calibration standard available to enable accurate comparison of different methodologies.

Methods

Patient selection

Serum specimens were selected from the sample archive stored at Peterborough District Hospital Clinical Biochemistry laboratory. The inclusion criteria used to select patients suitable for the study were subjects between the ages of 18 and 65 and Caucasian (white British) and Asian (Indian, Pakistani or Bangladeshi). The exclusion criteria included pregnant women, subjects with abnormal serum calcium, phosphate or alkaline phosphatase levels (using local reference ranges) and subjects with history or evidence of the following conditions: rickets or osteoporosis, renal failure, diabetes mellitus, autoimmune disease, gastrointestinal malabsorption, any form of cancer. The suitable subjects were also screened for their locality to Greater Peterborough (by their home postcode) to enable creation of a representative population for the study.

Specimens were anonymized and segregated into sex and ethnic background by using the demographical information provided on the laboratory database. Specimens were collected between the period of 1 December 2007 to 31 January 2008 (winter cohort) and 1 June to 31 July 2008 (summer cohort). The number of subjects for the summer and winter periods, respectively, were as follows: Caucasian male, 29 and 27; Caucasian female, 28 and 25; Asian male, 29 and 27; Asian female, 20 and 15. Suitable samples for the Asian cohort were significantly more difficult to obtain than those for the Caucasian population. This was due to only 7.3% of the total Greater Peterborough population being of Asian origin compared with the Caucasian contingent of 87.2%.¹⁸

Sample requirements and storage

The analysis of serum vitamin D₃ required peripheral blood collected into a serum gel Z/4.7 ml (Sarstedt monovette^®) bottle. The samples were centrifuged at 5127 g for 12 min within 8 h of venopuncture. All samples that were collected had been stored at 4°C for a maximum of 7 days post centrifugation. Aliquots taken for the study were stored at −20°C in sterile 7 ml pots in light protected conditions until analysis (maximum of 11 months). All storage methods complied with requirements of the methodology stated by the manufacturer Chromsystems (Munich, Germany).

Vitamin D₂/D₃ analysis

Specimens were analysed using an Agilent 1100 high pressure liquid chromatography (HPLC) system with Chromsystems vitamin D₂/D₃ methodology (calculating 25-hydroxyvitamin D_2/3). The methodology allows determination of vitamin D through an isocratic HPLC system. 500 µl of sample underwent protein precipitation and selective solid phase extraction to remove interfering contaminants and concentrate the analytes. To aid accurate quantification an internal standard (stable vitamin D derivative) was added to monitor the extraction process. The HPLC system was used to calculate the vitamin D₂/D₃ concentrations at 265 nm using a multi-wavelength detector with retention times of 4.75 min for D₃ and 5.2 min for D₂. The lower limit of detection assessed locally was 5 µg/l and the methodology performed with CVs of 2.65% for D₂ and 2% for D₃. Chromatographs were automatically integrated and vitamin D₂/D₃ concentrations calculated by the Agilent ChemStation B.03.01 software through peak height. All HPLC runs were calibrated and quality controlled using certified products produced by Chromsystems. All procedures were carried out in line with the manufacturer's guidelines.

Data analysis

All statistical analysis was performed using the Analyse-It: method evaluation statistical package. The population distribution was identified using the Pearson's coefficient (Pc) indicating a predominantly non-parametric distribution. Box and whisker plots located potential outliers within the data set. Two-tailed Mann–Whitney U (non-parametric) tests were used to statistically analyse for significant difference between subgroups medians, as the results were not predicted and the data were of unmatched nature and unequal size. Results were interpreted by analysing the critical value (<0.05 indicates significant difference). The ‘normal’ reference ranges (95% confidence level) were produced non-parametrically by excluding the highest 2.5% and lowest 2.5% of the cohort in question.

Results

Two hundred subject samples were analysed during the study (94 winter samples; 106 summer samples), with seasonal groups subdivided into sex and ethnic origins. The mean subject age was 43 years (range 18–65 years). The number of male subjects was 56% (112) compared with 44% (88) females. The serum vitamin D concentration was obtained from the chromatographs produced by the Agilent ChemStation B.03.01 software, using an internal vitamin D standard. Only vitamin D₃ was analysed due to the unquantifiable concentrations of D₂ in the subjects studied (data not shown for brevity).

Serum vitamin D₃ concentrations of the Caucasian and Asian populations between summer and winter periods are shown in Fig. 1. The statistical ‘outliers’ highlighted [results >1.5 of inter-quartile range (IQR)] have not been excluded from the study as there is no evidence suggesting that they were produced through analytical error. The result >3 IQR was repeated to confirm the analytical validity. Therefore, the results must be regarded as genuine and represent the respective populations studied, as the variation is more likely environmental in nature (i.e. sunlight exposure and diet).

Figure 1.

Serum vitamin D3 levels between ethnicities and season. Serum samples were taken from different population groups to determine vitamin D₃ levels. Data are presented as box and whisker plots (‘+’ represent >1.5 and <3 inter-quartile range or ‘×’ >3 inter-quartile range are depicted on the plots. ^#P < 0.001, total Caucasian summer compared with total Caucasian winter, total Asian winter or total Asian summer cohorts. Sample sizes were as follows: Caucasian Winter, n = 52; Caucasian Summer, n = 57; Asian Winter, n = 42; Asian Summer, n = 49).

The data clearly shows seasonal fluctuations in the Caucasian subjects where the median value decreased from 32.5 to 21.1 µg/l (−35%; P < 0.0001), whereas the Asian cohort only decreased from 9.8 to 8.6 µg/l (−12%; P = 0.1169) in the winter season. These results also indicated significant differences between the Caucasian cohort when compared with the Asian cohort for both winter and summer periods, respectively (P < 0.0001; Fig. 1). This indicates that the majority of the Caucasian population consistently have higher serum vitamin D₃ concentrations than the Asian cohort. In fact the median for the Caucasian population in the winter is double that of the summer Asian cohort.

When the data were further sub-grouped in terms of gender more discrete differences were displayed (Fig. 2). A similar pattern is observed for Caucasian men between summer and winter seasons where the values decrease from 32.9 to 22.1 µg/l (−49%; P = 0.0037) and females from 30.5 to 18.8 µg/l (−62%; P = 0.0117), respectively. However, a significant difference is now shown for Asian males where the values decrease from 12.5 to 8.3 µg/l (−51%; P = 0.0022), but no change was shown for Asian females of 8.1 vs. 9.6 µg/l (+16%; P > 0.05) (Fig. 2).

Figure 2.

Serum vitamin D3 levels between gender, ethnicity and season. Data are presented as box and whisker plots (Caucasian male winter, n = 27; Caucasian male summer, n = 29; Caucasian female winter, n = 25; Caucasian female summer, n = 28; Asian male winter, n = 27; Asian male summer, n = 29; Asian female winter, n = 15; Asian female summer, n = 20).

Further statistical comparisons between the different population groups are summarized in Table 1. There were significant differences between the Caucasian males and Asian males in the winter and summer populations (P≤0.0001) and this finding was mirrored by the statistical analysis between the Caucasian females and Asian females in the winter and summer periods (P≤0.0001). No significant difference was observed between the Caucasian males and females either in the summer or winter periods as shown in Fig. 2. However, Asian males appear to have slightly higher vitamin D₃ concentrations than the females, most prominently in the summer (P≤0.0001).

Table 1.

Comparisons of different population groups over the same season

Population group comparison	Season	Calculated Mann–Whitney U-value	Significance (*P-value)
Caucasian males vs. Asian males	Winter	67	<0.0001
Caucasian males vs. Asian males	Summer	49	<0.0001
Caucasian females vs. Asian females	Winter	27	<0.0001
Caucasian females vs. Asian females	Summer	10	<0.0001
Caucasian males vs. Caucasian females	Winter	317	0.7005
Caucasian males vs. Caucasian females	Summer	371	0.5709
Asian males vs. Asian females	Winter	228	0.5116
Asian males vs. Asian females	Summer	103	<0.0001

*P < 0.05, statistically significant.

Ethnicity-specific reference ranges were produced as there were significant differences between the Caucasian and Asian cohorts all year round (Figs. 1 and 2; Table 1). This was confirmed by the calculated reference ranges, particularly the upper limit (Table 2). As there was no significant difference (Table 1) between the Caucasian male and female cohorts for the same season there is no requirement to produce a sex-specific reference range. Since the season-specific reference range showed very little variation between seasons (lower limit: 4.1%; upper limit: 1.7%) (Table 2), an annual Caucasian reference range was produced, which has increased accuracy due to the significantly larger population (n =109) so extreme results will have less impact upon the result.

Table 2.

‘Normal’ serum vitamin D₃ reference ranges (95% confidence interval) derived from the study data

Population	Lower limit (μg/l)	Upper limit (μg/l)
Caucasian Winter	7.6	50.9
Caucasian summer	7.3	51.8
Caucasian all year	8.2	53.7
Asian winter	3.6	22.4
Asian summer	3.3	29.3
Asian all year	3.6	26.7

Like the Caucasian population, there was no statistical difference between the Asian male and female groups in the winter, indicating no requirement for a sex-specific reference range. The summer Asian cohorts did show significant difference (Table 1), thus it would be desirable to produce a sex-specific reference range. However, as each cohort is <40 observations (1/α = 1/0.025), it is not statistically possible to produce an accurate range estimation. The summer and winter Asian reference ranges (Table 2), like the Caucasians, have very closely matching lower limits (0.3 µg/l variation). While the upper limits are visibly dissimilar showing 24% difference between the summer and winter. However, this was not confirmed statistically by the Mann–Whitney U analysis (Table 1), which indicated no significant difference between the two seasons, therefore an annual-specific reference range was produced (Table 2).

Discussion

Due to Peterborough's diverse population it was important to establish if there was a difference in vitamin D₃ status of the Caucasian and Asian populations. This is of particular interest and importance in diagnostic medicine for the formation of appropriate reference ranges to assess pathologies and to advise upon treatment strategies, particularly as there have been large increases in the number of cases of rickets and osteomalacia within the UK.¹⁷

This study has conclusively proven that there is a significant difference in serum vitamin D₃ concentrations between the Caucasian and Asian subject cohorts during the summer (70%) and winter (59%) seasons (Fig. 1). The data showed a strong seasonal link between serum vitamin D₃ concentrations in the Caucasian populations with a divergence in medians of 35%, whereas only a slight increase in median serum concentration occurred in the Asian population. In fact the female Asian population showed a slight decrease in the median serum vitamin D₃ concentration during the summer when compared with the winter. This is an unexpected result, which suggests that the Asian population of Peterborough has reduced levels of vitamin D₃ acquired through sunlight exposure compared with Caucasian counterparts. This hypothesis is suggested as there is no significant rise in vitamin D₃ concentration during the summer period from the winter ‘baseline’, indicating their vitamin D₃ intake is relatively constant all year round. This poses the question as to whether Asian females have any significant exposure to sunlight, otherwise a slight rise in median serum vitamin D₃ levels would be expected during the summer period as seen in the male population. The most likely factors causing reduced serum vitamin D₃ concentrations in the Asian cohort are, skin pigmentation, cultural dress and diet.

Darker skin pigmentation plays a pivotal role in reducing vitamin D₃ biosynthesis, as increased levels of melanin enhance UVB photon absorption, thus decreasing in vivo vitamin D₃ production.¹⁴ Previously, Clemens et al.¹⁹ determined that severely increased melanin pigmentation may reduce vitamin D synthesis by up to 99%, while Holick²⁰ quantified that individuals with darker skin required 10 times more exposure to sunlight to produce the same amount of vitamin D₃ as individuals with fairer skin. Furthermore, a study by Lo et al.²¹ stated that Asians have the same capacity as Caucasians to produce vitamin D₃, although longer exposure is required for equivalent erythemal response. Therefore, Peterborough with a latitude of 52°N undergoes large variations of UVB levels during the seasons due to the earth's axis (i.e. reduced during winter).^{2, 22} This would further decrease the Asian cohorts ability to produce vitamin D₃in vivo during the winter when compared with the Caucasian population.

Cultural dress (e.g. burkhas) is potentially a major contributor to reduced vitamin D₃ concentrations as the majority of the Asian population studied are of Pakistani origin. Since female Asians cover the majority of their bodies (e.g. purdah) this in turn will reduce the exposure to sunlight all year round and thus vitamin D₃ production.²³ Vitamin D₃ levels will therefore be depleted as an estimated 90% of the vitamin is produced in vivo unless replaced through diet or supplementation.² Whereas, the male Asian cohorts median vitamin D₃ concentration did increase slightly in the summer period (8.3–12.5 µg/l), probably due to the fact that they are able to attain increased sun exposure since the majority of their skin is not covered. Finally, dietary influences may also contribute to reduced vitamin D₃ concentrations. For example, high levels of fibre and phytate intake or a low calcium diet.^{24, 25}

Other factors, which were not controlled in this study, may have caused falsely raised or lowered results such as, self-supplementing with vitamin D₃, regular sun-bed use, a recent time period in ‘sun-rich’ environments, night shift workers, frequent sunscreen use (absorbing UVB) or religious orientation (diet and cultural dress). Body mass index has also been described as a contributing factor to reduced vitamin D₃ levels as Wortsman et al.²⁶ states that vitamin D deficiency is commonly associated with obesity. Therefore, a probability exists that some of the subjects may introduce bias to the study such as the results >1.5 IQR from the median. These individuals are unable to be distinguished from the ‘normal’ population through analytical means.

The Caucasian winter and summer groups were statistically different (P≤0.0001), displaying the traditional ‘peak and trough’ appearance in the seasonal cycle (Fig. 1). This finding mirrors those of other studies, which show similar patterns within Caucasian populations and confirms the notion of vitamin D₃ status being directly related to intensity of, and exposure to UVB.³ Therefore, during the winter months when the sun's rays reach the Earth at a more oblique angle, the ozone absorbs increased levels of UVB rays diminishing the body's ability to produce vitamin D₃in vivo.¹³ The maximum average atmospheric temperatures of 7.4°C and sunshine hour average of 50 in the Peterborough region for the winter months samples were significantly reduced from those of the summer months (20.8°C and 186 h).¹⁸ This means individuals are more likely to be covering a larger skin area during the winter months and reducing levels of outdoor leisure time, thus reducing exposure time. This will compound the issue of a weaker concentration of UVB radiation, therefore, causing a reduction of serum vitamin D₃ concentrations.

When observing the reference ranges produced by this study it is clear to see that the reference intervals are wide (Caucasian 8.2–53.7 µg/l; Asian 3.6–26.7 µg/l). The Caucasian lower interval concurs with the 8 µg/l suggested by Serhan et al.,¹⁶ with the Asian lower limit significantly lower than any recommended. The limit of deficiency currently accepted in the UK and most commonly quoted in the literature is 10 µg/l and is applied to all ethnic groups.^{11, 16, 27–30} This definition determines 64% of the Asian population studied over the winter and 53% during the summer to be vitamin D deficient (Table 3). Furthermore, using the lower limit of 32 µg/l suggested by Lips,³¹ 100% of the combined Asian population studied would be classed as vitamin D deficient all year round. More strikingly 74% of all the Asian females studied (both winter and summer) are deemed vitamin D deficient if the limit is 10 µg/l. This is well documented having been determined at 43% by Ford et al.¹⁷ upon outpatient clinics in Birmingham and 50% by Datta et al.³² who studied pregnant ethnic minorities attending antenatal clinics. Although, a study by Awumey et al.³³ suggested that in Asians 25(OH)D₃-24-hydroxylase is upregulated causing low serum 25-hydroxyvitamin D₃ levels. This suggests there may be a genetic predisposition to reduced vitamin D levels in the Asian population. If this were to be conclusively proven then an Asian-specific reference range would potentially be more appropriate.

Table 3.

Published serum 25-hydroxyvitamin D concentrations defining deficiency and prevalence of vitamin D deficiency in the study population

Definition of vitamin D deficiency (μg/l)	Percent of population vitamin D deficient
	Caucasian total	Caucasian winter	Caucasian summer	Asian total	Asian winter	Asian summer
<816	1.83	1.92	1.75	28.57	35.7	22.5
<1011, 28, 29	5.50	7.69	3.51	58.2	64.3	53.1
<2030	28.4	48.1	11.0	77.4	95.2	91.8
<3231	62.4	77.0	49.1	100	100	100
<4036	88.1	90.4	86.0	100	100	100

If the lower limit of the Caucasian reference range established by this study were applied to the Asian population then 38% are deficient during the winter and 31% during the summer. The Caucasian limit of deficiency is the most accurate for Peterborough as it is locally derived from populations exposed to similar climatic conditions, the same latitude, and is similar to the UK consensus figure of 10 µg/l.

The Caucasian population fare better using these increased lower limits with only 7.7% being classed as deficient during the winter and 3.5% in the summer at 10 µg/l. However, a substantial proportion of the Caucasian cohort (48%) are deemed deficient during the winter if the lower limit were set as 20 µg/l, although this drops to 11% during the summer period. Therefore, from the Caucasian population studied the lower limit suggested of 8.2 µg/l is a realistic figure, which corresponds to other studies published (Serhan et al.¹⁶), but is slightly lower than the consensus figure of 10 µg/l. Using government population statistics and the lower local reference interval (<8.2 µg/l) 1.8% (2606 people) of the Caucasian population are deemed vitamin D deficient, while 34% (4065) of the Asian population have hypovitaminosis D.³⁴

In contrast to the vast quantity of work studying the lower limit, the scientific data quantifying an upper limit is very unclear. It has been demonstrated by Dawson-Hughes et al.³⁵ that serum vitamin D levels >80 µg/l are attained by individuals not taking supplementation, but living in ‘sun-rich’ environments. This would indicate that these high levels are still within physiological parameters as vitamin D₃ is produced in vivo under tightly controlled regulatory processes. Furthermore, Vieth³⁶ states that absolute rises in serum vitamin D₃ are directly related to basal concentration (i.e. <10 µg/l showed double the increase of those with >20 µg/l) when exposed to equivalent UVB. Hence hypervitaminosis D is only likely to occur due to over supplementation. This was displayed in this study as three high results (up to 76.3 µg/l) showed no loss of bone marker homeostasis. Therefore, the upper reference range limit proposed from the Caucasian cohort of 53.7 µg/l is within levels suggested in the literature and should be considered genuine for the population studied. The Asian upper limit of 26.7 µg/l is unlikely to be a realistic upper value of vitamin D tolerance when compared with the highest Caucasian result (76.3 µg/l) with no obvious effects of hypervitaminosis D. These results suggest that the true upper interval of serum vitamin D is likely to be higher than this study suggests and is potentially greater than those studied by Dawson-Hughes et al.³⁵

However, the above comparisons of interval limits reported by other groups with the present study need to be interpreted with some caution since currently there is no international calibration standard for all vitamin D methodologies. This can cause variation in the vitamin D₃ levels recovered from different assays, which means comparison of results from different methodologies is debatable.

In summary, significant differences in serum vitamin D₃ levels exist between the Caucasian and Asian populations of Peterborough. This study has also produced a reference range, which is specific for the local population and corresponds to published reports. However, a separate reference range for the Asian population of Peterborough requires further research to indicate whether a lower range should be considered ‘optimal’ for health. There is also no requirement for a season-specific range due to the insignificant variation between seasonal ranges and the derived value of deficiency correlates with other published values.^{11, 16, 29, 30} Furthermore, this study has raised serious concerns regarding the accuracy of the traditional ‘bone markers’ for vitamin D deficient pathologies, since the markers (calcium, phosphate or alkaline phosphatase), which were used to screen for ‘healthy’ subjects were normal. This indicates the requirement for an improved programme of screening for vitamin D disorders of the Asian community since large numbers of individuals with hypovitaminosis D are currently being undiagnosed and raises the possible requirement of vitamin D supplementation in the diet.

Funding

This project was fully funded by the Clinical Biochemistry Department, Peterborough District Hospital.

Author biography

Matthew has recently graduated from the University of Westminster with first class BSc Honours in Biomedical Science. The course was undertaken on a part-time basis over 4 years while working full time at the department of Clinical Biochemistry, Peterborough District Hospital as a trainee Biomedical Scientist. Matthew is starting an MSc in Clinical Biochemistry at the University of Westminster in October 2009 and intends to continue in his career as a Biomedical Scientist.

Acknowledgements

I would like to express my sincere thanks to my workplace tutor, Robin Lucas-Evans (Peterborough District Hospital) and University tutor Dr Vinood Patel (University of Westminster) for their continued advice and support through the whole project process. I would also like to thank Dr Stephen Reed (University of Westminster) for his statistical advice.

Author notes

Supervisor: Vinood B. Patel. Department of Biomedical Science, School of Life Sciences, University of Westminster, 115 New Cavendish Street, London W1W 6UW, UK. Tel: [phone] (xtn. 3543). Email: [email protected]