Nutrition, Metabolism & Cardiovascular Diseases (2015) 25, 287e294

Available online at www.sciencedirect.com

Nutrition, Metabolism & Cardiovascular Diseases

journal homepage: www.elsevier.com/locate/nmcd

Effect of intermittent vitamin D3 on vascular function and

symptoms in chronic fatigue syndrome e A randomised controlled
trial
M.D. Witham*, F. Adams, S. McSwiggan, G. Kennedy, G. Kabir, J.J.F. Belch, F. Khan
Medical Research Institute, University of Dundee, Ninewells Hospital, Dundee DD1 9SY, Scotland, UK
Received 2 August 2014; received in revised form 1 October 2014; accepted 14 October 2014
Available online 22 October 2014

KEYWORDS
Vitamin D;
Arterial stiffness;
Chronic fatigue;
Randomised
controlled trial

Abstract Background and aims: Low 25-hydroxyvitamin D levels are common in patients with
chronic fatigue syndrome; such patients also manifest impaired vascular health. We tested
whether high-dose intermittent oral vitamin D therapy improved markers of vascular health
and fatigue in patients with chronic fatigue syndrome.
Methods and results: Parallel-group, double-blind, randomised placebo-controlled trial. Patients
with chronic fatigue syndrome according to the Fukuda (1994) and Canadian (2003) criteria were
randomised to receive 100,000 units oral vitamin D3 or matching placebo every 2 months for 6
months. The primary outcome was arterial stiffness measured using carotid-femoral pulse wave
velocity at 6 months. Secondary outcomes included ow-mediated dilatation of the brachial artery, blood pressure, cholesterol, insulin resistance, markers of inammation and oxidative
stress, and the Piper Fatigue scale.
As many as 50 participants were randomised; mean age 49 (SD 13) years, mean baseline pulse
wave velocity 7.8 m/s (SD 2.3), mean baseline ofce blood pressure 128/78 (18/12) mmHg and
mean baseline 25-hydroxyvitamin D level 46 (18) nmol/L. 25-hydroxyvitamin D levels increased
by 22 nmol/L at 6 months in the treatment group relative to placebo. There was no effect of treatment on pulse wave velocity at 6 months (adjusted treatment effect 0.0 m/s; 95% CI
0.6 to 0.6;
p Z 0.93). No improvement was seen in other vascular and metabolic outcomes, or in the Piper
Fatigue scale at 6 months (adjusted treatment effect 0.2 points; 95% CI
0.8 to 1.2; p Z 0.73).
Conclusion: High-dose oral vitamin D3 did not improve markers of vascular health or fatigue in
patients with chronic fatigue syndrome.
Trial registration: www.controlled-trials.com, ISRCTN59927814.
2014 Elsevier B.V. All rights reserved.

Introduction
Chronic fatigue syndrome/Myalgic encephalomyelitis
(CFS/ME) is a poorly-understood syndrome characterised
by extreme tiredness, and activity limitation in the

* Corresponding author. Ageing and Health, Ninewells Hospital,

Dundee DD1 9SY, UK. Tel.: 44 01382 383086; fax: 44 01382 644972.
E-mail address: m.witham@dundee.ac.uk (M.D. Witham).
http://dx.doi.org/10.1016/j.numecd.2014.10.007
0939-4753/ 2014 Elsevier B.V. All rights reserved.

absence of other diagnoses that might explain the symptom complex. Vigorous debate continues as to both the
causes of the syndrome and the most effective approach to
therapy. Recent evidence suggests that a range of markers
of vascular health are impaired in patients with CFS/ME,
including arterial stiffness and endothelial function [1,2];
ventricular function and cardiac output are also reduced
[3]. Whether this is due to the disease state itself or to
deconditioning from low activity levels is unclear. Such
impairments might provide a causal contribution to the

288

disease syndrome, but are important in their own right as

markers of increased risk of cardiovascular events.
A large body of observational data links low vitamin D
status, measured as circulating 25-hydroxyvitamin D
(25OHD) levels, with vascular disease, including hypertension and the metabolic syndrome, as well as an
increased risk of future cardiovascular events [4]. Very low
25OHD levels are seen in osteomalacia, a disease primarily
of bone, but also characterised by fatigue, muscle pain and
evidence of a skeletal myopathy. Vitamin D supplementation successfully reverses the myopathy of osteomalacia,
although the effect of vitamin D on vascular disease has
been mixed [5,6] and is the topic of several large, ongoing
trials.
Activity limitation is a hallmark of CFS/ME, and the
consequent lack of sunlight exposure limits opportunities
for ultraviolet light driven vitamin D synthesis in the skin.
Unsurprisingly, circulating 25-hydroxyvitamin D (25OHD)
levels are considerably lower in patients with CFS than in
healthy controls [7]. Furthermore, correlations have been
observed between lower 25OHD levels and arterial stiffness, impaired endothelial function, markers of inammation and oxidative stress [8] which may in themselves
mediate adverse effects on vascular function.
Observational data cannot provide robust evidence of
causality, particularly in view of the fact that any illness
reducing activity levels or leading to an inammatory state
[9] will tend to display low 25OHD levels; prospective
randomised trials are therefore required to test whether
vitamin D supplementation can reverse the impairments
in vascular health and symptoms seen in CFS. We therefore
conducted a double blind, placebo controlled trial of highdose intermittent vitamin D3 supplementation to test
whether this intervention could improve markers of
vascular function, markers of inammation and oxidative
stress, as well as symptoms, in patients with CFS/ME.
Methods
Study design and population
We conducted a parallel group, placebo controlled, double
blinded, randomised trial of oral vitamin D supplementation on cardiovascular disease risk in patients with diagnosed CFS. Participants were recruited from the
Connective Tissue Disease Clinic at Ninewells Hospital,
Dundee, Scotland, via advertising in the ME Research UK
magazine and through local ME patient support groups.
Inclusion criteria for the trial were: age 18 or over, diagnosed CFS which fullled the Fukuda (1994) [10] and Canadian (2003) criteria [11], and serum 25OHD level
<75 nmol/L. Exclusion criteria were: history of osteoporosis, sarcoidosis, renal stones, or metastatic malignancy,
already taking pharmacological vitamin D preparations
(sh oils were permitted), liver function tests (bilirubin,
alanine, aminotransferase or alkaline phosphatase)
>3  upper limit of normal, corrected serum calcium
>2.60 mmol/L or <2.15 mmol/L, and estimated glomerular
ltration rate <40 ml/min. Patients who were unable to

M.D. Witham et al.

give informed written consent, who were of child bearing

age and were not taking reliable contraception, who were
pregnant, who were diagnosed with psychiatric disorders
within the past 5 years, substance abuse/dependence, or
who were diagnosed with an eating disorder at any time
were also excluded from the study. Following the
screening visit, eligible patients were invited for baseline
assessments of arterial stiffness, endothelial function, and
metabolic and inammatory markers. The study was cosponsored by the University of Dundee/NHS Tayside and
was carried out according to the Principles of Good Clinical
Practice and the Declaration of Helsinki. Research ethics
approval was obtained from the East of Scotland Research
Ethics Committee (reference 10/S0501/61) and written
informed consent was obtained from all participants. The
UK Medicines and Healthcare Regulatory Agency (MHRA)
approved the trial (EuDRACT number 2010-019096-29).
The trial was registered on the International Standard
Randomised
Controlled Trials
Network register:
ISRCTN59927814. The protocol is available from the
authors.
Intervention and randomisation
Once the baseline measurements had been completed, a
single dose of 100,000 units of oral vitamin D3 (Vigantol
oil, 20,000 units vitamin D3 per ml, Merck KGaA, Darmstadt, Germany) or matching placebo (Mygliol oil, Merck
KGaA, Darmstadt, Germany) was administered to patients
randomised in a 1:1 ratio. Tayside Pharmaceuticals, Dundee, Scotland prepared the study medication bottles with
sequential medication study numbers according to a
computer-generated number sequence used for randomisation, stored in sealed envelopes until the end of the
study. Each patient received the next numbered study
medication bottle, ensuring allocation concealment. The
study dose was repeated at 2 months and 4 months and
the medication was ingested in the presence of the
research team to ensure 100% adherence. Participants and
healthcare providers remained blinded to the treatment
allocation until the trial data had been analysed.
Outcome measures
Outcome measures were performed at baseline, 2 months
and 6 months. All outcomes were measured and analysed
blind to treatment allocation. The primary outcome was
change in arterial stiffness, measured by carotid-femoral
pulse wave velocity, from baseline to 6 months, compared
between the two groups. Arterial stiffness has important
effects on ventricular output as well as being a marker of
future cardiovascular events [12]; low 25OHD levels are
associated with increased arterial stiffness [8,13] which is
itself prevalent in patients with CFS/ME [1].
Arterial stiffness
Peripheral pressure waveforms were recorded at the
femoral, carotid and radial artery by applanation

Effect of intermittent vitamin D3 in chronic fatigue syndrome

tonometry using the validated SphygmoCor pulse waveform analysis system (AtCor Medical) [14]. Augmentation
index e a ratio of the pressure increase to pulse pressure
normalised for a heart rate of 75 beats per minute
(AIx@75) e was calculated from the central aortic pulse
wave by the internal Sphygmocor algorithm. Pulse wave
velocity in metres/second was derived by recording the
pressure waveform at the femoral & carotid artery and
comparing time of arrival with time of the ECG R-wave. If a
measurement could not be obtained at the femoral artery
the radial artery was used as an alternative site to capture
the carotid to radial pulse wave velocity. The coefcient of
variation (within-person test variability) for PWV in our
laboratory is 6.3%
Blood pressure
Blood pressure was measured in the supine position using
an automatic sphygmomanometer (Omron M6, OMRON
Healthcare Europe, Hoofddorp, Netherlands). Three readings were taken and the mean blood pressure was calculated from the second and third readings.
Endothelial function
Flow mediated dilation (FMD) of the brachial artery was
measured according to standard guidelines [15]. In brief,
the diameter of the brachial artery was measured in
longitudinal section using a 5e8 MHz linear array
transducer (Acuson Sequoia 512, Siemens Medical Solutions, Berkshire, UK). Images were recorded for 1 min at
baseline before inating a sphygmomanometer cuff,
placed around the forearm, to 50 mm above systolic
pressure or to a minimum of 200 mmHg for 5 min. After
cuff release brachial artery images were recorded for a
further 2 min. Images were stored in the DICOM (Digital
Imaging and Communications in Medicine) le format
and analysed using the dedicated edge detection software Vascular Tools Brachial Analyser (Medical Imaging
Applications LLC, Coralville, USA). Once a stable baseline
had been re-established the change in brachial artery
diameter was measured in response to sublingual
administration of 0.4 mg of glyceryl trinitrate (GTN, Lipha
Pharmaceuticals Ltd, West Drayton, UK), an endotheliumindependent vasodilator. The coefcient of variation
(within-person test variability) for FMD in our laboratory
is 10.9%.
Metabolic and inammatory markers
Blood was sampled from the antecubital fossa after
minimal venous occlusion using size 21 gauge buttery
cannula, and blood was allowed to ow freely. The blood
was collected into BD Vacutainer SST Tubes (BD Diagnostics, USA) for serum preparation. The serum tubes
were centrifuged for 15 min at 1300g. The serum was
removed and placed into various aliquots before being
stored at
20  C for vitamin D analysis and
70  C for all
other blood tests. High sensitivity enzyme linked

289

immunosorbent assays (ELISA) were used to measure

Interleukin- 6 (IL-6), tumour necrosis factor-a (TNF-a)
(R&D Systems, UK) and C-reactive protein (CRP) (Kalon
Biological Ltd, UK). Assays from Immunodiagnostic Systems (IDS) were used to measure 25OHD (ELISA) and
1,25-dihydroxy vitamin D (1,25OHD) [radioimmunoassay
(RIA)]. Markers of global oxidative stress d-ROMs (derivatives of reactive oxygen metabolites) and BAP (biological antioxidant potential) tests were measured on a
FRAS 4 (H&D srl, Italy) integrated analytical system which
consists of a dedicated photometer and an incorporated
centrifuge. Each of the samples was tested in duplicate
and all of the samples from one subject - i.e. all 3 visits were measured on one assay, to account for inter-assay
variation. Homeostatic model of assessment for insulin
resistance (HOMA-IR) was calculated as [fasting glucose
(mmol/L)  fasting insulin (mU/L)]/22.5. Serum calcium,
glucose, creatinine and intact parathyroid hormone were
measured as part of routine clinical workow by the
department of Biochemical Medicine, Ninewells Hospital,
Dundee, using a multichannel Roche analyser system.
Coefcients of variation for 25OHD were 2% (inter-assay)
and 2.5% (intra-assay); for 1,25OHD coefcients of variation were 13% (inter-assay) and 4.4% (intra-assay). For
PTH, the normal range was 1.0e6.9 pmol/L; the coefcient of variation was 6% (at 1.3 pmol/L) to 12% (at
12 pmol/L).
Fatigue
The Piper fatigue scale [16] was selected as a multidimensional measure of fatigue, encompassing cognitive,
behavioural, sensory and affective domains of fatigue. It
has been previously recommended for use in patients with
CFS/ME [17], and has shown response to interventions in
other fatigue conditions [18].
Statistical analyses
The trial was analysed by intention to treat. Two sided p
values of <0.05 were taken as signicant for all outcomes.
For all normally distributed outcomes, ANOVA was used to
compare results between groups at each follow up timepoint, adjusting for baseline values of the outcome. For
non-normally distributed variables, change scores between baseline and follow up were derived, which were
normally distributed. These values were then compared
using ANOVA in order to derive an estimate of treatment
effect size.
The sample size for the trial was calculated based on
the primary outcome. Previous studies have suggested that
a 1 m/s difference in pulse wave velocity is of clinical
importance; based on our previous work in this population
[1,2,8] we calculated that to detect a 0.9 m/s difference
with a standard deviation of 1.0 m/s would require 20
patients per group. Allowing for likely dropout over the 6
month study period, we planned to recruit a total of 25
patients per group (50 total).

290

M.D. Witham et al.

Results
64 potential participants were given information on the
trial, of whom 50 were randomised into the trial between
August 2011 and March 2012. Participant ow through the
trial is shown in Fig. 1. Baseline details of participants are
shown in Table 1. The group allocated to placebo differed
in a number of important ways from the intervention
group; the placebo group had higher baseline blood
pressure, worse endothelial function, stiffer arteries, worse
symptoms and higher levels of inammation and oxidative
stress than the intervention group. One participant in each
group had malabsorption syndrome (from celiac disease in
the treatment group, and from an enteroenterostomy in
the placebo group).
Tables 2e5 show the main trial results. Further
adjustment for baseline systolic blood pressure, baseline
25OHD level, baseline total cholesterol and baseline CRP
was performed for endothelial function and arterial stiffness measures; no signicant treatment effects were seen
after these further adjustments (FMD adjusted treatment
effect at 6 months
0.1% [95% CI
2.0
1.9] p Z 0.93; PWV

Figure 1

adjusted treatment effect at 6 months 0.0 m/s [95% CI

0.6
0.6] p Z 0.90; AIx adjusted treatment effect at 6
months 1% [95% CI
4
6] p Z 0.73).
Table 6 shows all adverse events reported during the
trial. One serious adverse event (a diagnosis of bowel
cancer requiring hospitalisation) was noted in the vitamin
D group; a further diagnosis of ductal breast carcinoma in
situ was noted in the vitamin D group. Other adverse
events were minor, self-limiting illnesses, including exacerbations of pre-existing CFS symptoms, and were equally
distributed across the two groups.
Discussion
This is the rst randomised controlled trial to our knowledge that has investigated the role of vitamin D supplementation in patients with CFS/ME. Although the focus of
our trial was on the vascular effects of vitamin D, we also
measured a range of metabolic, inammatory and symptom parameters. Despite a statistically and clinically signicant increase in 25OHD levels from vitamin D

CONSORT diagram of participant ow through the trial.

Effect of intermittent vitamin D3 in chronic fatigue syndrome

291

Table 1 Baseline details.

Mean age (SD)

Male sex (%)
Current smoker (%)
Body mass index
(Kg/m2) (SD)
Previous cancer (%)
Previous diagnosis
of vascular disease (%)
Previous diagnosis of
diabetes mellitus (%)
Statin use (%)
Antiplatelet use (%)
Antihypertensive use (%)
Median number of
medications [IQR]
Serum creatinine
(mmol/L) (SD)
Mean adjusted serum
calcium (mmol/L) (SD)
Mean 25OHD
(nmol/L) (SD)
Mean 1,25OHD
(pmol/L) (SD)
Mean PTH (pmol/L) (SD)
Mean systolic blood
pressure (mmHg) (SD)
Mean diastolic blood
pressure (mmHg) (SD)
Mean pulse rate
(beats/min) (SD)
Mean total cholesterol
(mmol/L) (SD)
Mean LDL cholesterol
(mmol/L) (SD)
Mean HDL cholesterol
(mmol/L) (SD)
Mean triglycerides
(mmol/L) (SD)
Median HOMA-IR (IQR)
Mean brachial artery
FMD (%) (SD)
Mean pulse wave
velocity (m/s) (SD)
Mean augmentation
index (%) (SD)
Median C-reactive
protein (mg/l) (IQR)
Mean IL-6 (pg/ml) (SD)
Mean TNF alpha
(pg/ml) (SD)
Mean d-ROMS
(U.Carr) (SD)
Mean BAP (mmol/L) (SD)
Piper Behavioural (SD)
Piper Affective (SD)
Piper Sensory (SD)
Piper Cognitive (SD)
Piper total (SD)

Vitamin D
group (n Z 25)

Placebo
group
(n Z 25)

48.1 (12.0)
7 (28)
4 (16)
28.8 (7.9)

50.7 (13.1)
5 (20)
6 (24)
29.8 (5.4)

0.47
0.51
0.73
0.60

1 (4)
1 (4)

1 (4)
3 (12)

1.0
0.61

3 (12)

1 (4)

0.61

2
0
4
2

4
3
5
4

(8)
(6)
(20)
[4]

0.67
0.12
0.71
0.04

68 (12)

71 (10)

0.30

2.27 (0.06)

2.31 (0.08)

0.05

44 (15)

48 (20)

0.44

48 (17)

43 (19)

0.36

6.0 (2.6)
125 (17)

5.0 (2.0)
132 (18)

0.15
0.16

76 (10)

81 (12)

0.14

70 (11)

70 (9)

0.85

5.2 (1.2)

5.7 (1.3)

0.17

3.2 (0.9)

3.6 (1.2)

0.19

1.5 (0.5)

1.2 (0.3)

0.07

1.2 (0.7)

2.0 (0.8)

0.002

2.6 (3.0)
7.7 (3.1)

2.2 (3.7)
6.5 (2.9)

0.78
0.17

7.3 (2.6)

8.3 (1.9)

0.13

16 (13)

27 (10)

0.001

0.6 (3.2)

2.7 (4.6)

0.04

1.6 (0.8)
1.5 (0.7)

2.5 (1.8)
1.9 (0.8)

0.05
0.09

389 (99)

473 (143)

0.04

2128 (407)
6.6 (1.8)
7.4 (1.6)
6.3 (1.9)
5.1 (1.5)
6.3 (1.6)

1815 (480)
8.0 (1.3)
7.8 (1.6)
7.5 (1.5)
6.1 (1.4)
7.3 (1.2)

0.03
0.006
0.45
0.02
0.04
0.02

(4)
(0)
(16)
[4]

25OHD, 25-hydroxyvitamin D; 1,25OHD, 1,25-hydroxyvitamin D;

PTH, parathyroid hormone; FMD, ow-mediated dilatation of the
brachial artery; LDL, low density lipoprotein; HDL, high density lipoprotein; HOMA-IR, homeostasis model assessment e insulin
resistance; IL-6, interleukin 6; TNF, tumour necrosis factor; d-ROMS,
derivatives of Reactive Oxygen Metabolites; BAP, biological antioxidant potential.

supplementation, we found no evidence of improvement

in vascular health, cholesterol, insulin resistance, inammation, oxidative stress or symptoms of fatigue or
depression.
Recent studies have demonstrated that patients with
CFS/ME have impaired vascular function when compared
to patients without these syndromes; CFS/ME patients
have worse endothelial function and arterial stiffness, and
the degree of vascular dysfunction is related to markers of
inammation and oxidative stress, as well as to circulating
25OHD levels [8]. The aetiology of vascular dysfunction in
CFS/ME is however likely to be complex. Inammation and
oxidative stress, thought to be part of the core aetiopathology of CFS/ME [19,20], are known to impair endothelial function via a range of mechanisms. However the
clinical hallmarks of CFS/ME are fatigue and reduced activity, which in turn are also known to impair endothelial
function, increase blood pressure and exacerbate inammation [21]. Low activity levels are associated with lower
exposure to sunlight and hence lower 25OHD levels. Thus
the association between low 25OHD levels, inammation,
oxidative stress and vascular dysfunction in CFS/ME may
not necessarily be causal.
There are a number of possible reasons why we did not
nd an effect of vitamin D supplementation on vascular or
inammatory markers in this trial. Firstly, as noted above,
low 25OHD levels may be a result of CFS/ME, rather than
being causal for either CFS/ME or the associated vascular
dysfunction. Secondly, a larger dose of vitamin D, or daily
rather than intermittent dosing may be required. It has
been argued that 25OHD levels above 75 nmol/L are
required for optimum health [22], although this gure was
extrapolated from observational studies. Although the
mean 25OHD level in the current trial did not reach this
level (in part because levels were measured eight weeks
after each dose, beyond the one week peak expected with
bolus vitamin D dosing), we have noted improvements in
endothelial function, blood pressure and B-type natriuretic
peptide with vitamin D supplementation in other selected
patient groups [23e25]. Improvements in these previous
trials were seen after 8 weeks of treatment, suggesting
that the 6 month treatment period in the current study
should be adequate to see effects. Nevertheless, we cannot
exclude the possibility that larger, more frequent dosing
could still produce a clinically relevant effect. A third
possibility is that vitamin D supplementation is effective
only for those patients with very low baseline 25OHD
levels. Only 6/50 (12%) of participants had 25OHD levels
below 25 nmol/L, thus we are unable to test this hypothesis using the existing trial data. Lastly, CFS is a syndrome,
not a disease; it is quite possible that the syndrome encompasses a variety of different aetiopathological entities,
some of which might respond to vitamin D supplementation. Until the means exist to better characterise CFS
subtypes, we are unable to test whether vitamin D supplementation may benet some, but not all, patient groups
with CFS.
Our study had a number of strengths. The randomised,
placebo controlled design reduced the risk of bias, and the

292

M.D. Witham et al.

Table 2 Vascular outcome measures.

Pulse wave
velocity (m/s)
Augmentation
index (%)
Systolic blood
pressure (mmHg)
Diastolic blood
pressure (mmHg)
Pulse (beats/min)
FMD (%)

2
6
2
6
2
6
2
6
2
6
2
6

mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths

Vitamin D (SD)

Placebo (SD)

Adjusted treatment
effect (95% CI)a

7.2
6.9
15
15
123
124
76
75
74
74
7.4
7.4

8.5
8.1
25
25
130
133
83
82
76
73
6.1
6.9

0.2
0.0
2
1

2

2

5

2

3
1
0.9
0.1

0.44
0.93
0.16
0.60
0.59
0.55
0.04
0.34
0.40
0.78
0.17
0.93

(2.5)
(2.4)
(12)
(12)
(17)
(17)
(8)
(9)
(12)
(13)
(2.6)
(2.5)

(1.7)
(1.4)
(9)
(10)
(16)
(15)
(9)
(10)
(14)
(11)
(2.8)
(3.4)

(
0.7e0.3)
(
0.6e0.6)
(
1e6)
(
4e6)
(
10e6)
(
10e5)
(
9e0)
(
7e2)
(
9e4)
(
5e6)
(
0.4e2.2)
(
1.6e1.7)

FMD, ow-mediated dilatation of the brachial artery.

a
Adjusted for baseline outcome measure values.

Table 3 Vitamin D, calcium and PTH.

25OHD (nmol/L)
1,25OHD (pmol/L)
Adjusted serum
calcium (mmol/L)
PTH (pmol/L)

2
6
2
6
2
6
2
6

mths
mths
mths
mths
mths
mths
mths
mths

Vitamin D (SD)

Placebo (SD)

Adjusted treatment
effect (95% CI)a

55
64
53
56
2.30
2.31
7.0
5.8

44
44
36
35
2.29
2.30
6.4
7.8

15
22
14
18
0.03
0.04
0.4

2.4

0.001
0.001
0.003
0.008
0.23
0.10
0.72
0.006

(18)
(21)
(19)
(22)
(0.09)
(0.07)
(5.9)
(3.2)

(21)
(22)
(17)
(24)
(0.08)
(0.09)
(3.3)
(4.6)

(7e22)
(10e34)
(5e23)
(5e31)
(
0.02e0.08)
(
0.01e0.09)
(
1.9e2.8)
(
4.1 to
0.7)

25OHD, 25-hydroxyvitamin D; 1,25OHD, 1,25-hydroxyvitamin D; PTH, parathyroid hormone.

a
Adjusted for baseline outcome measure values.

group of patients studied in this trial were carefully phenotyped to ensure that they met both the Fukuda and the
Canadian criteria for diagnosis of CFS/ME. 25OHD levels
were raised by the intervention. The study duration was

longer than that required to show changes in vascular

markers in our previous studies of vitamin D interventions. Unfortunately, there was signicant imbalance in baseline characteristics between groups, despite

Table 4 Metabolic, inammatory and oxidative stress outcome measures.

Total cholesterol (mmol/L)

LDL cholesterol (mmol/L)
HDL cholesterol (mmol/L)
Triglycerides (mmol/L)
Median HOMA-IR
Median C-reactive protein (mg/L)
IL-6 (pg/ml)
TNF-alpha (mg/ml)
d-ROMS (U.Carr)
BAP (mmol/L)

2
6
2
6
2
6
2
6
2
6
2
6
2
6
2
6
6
6

mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths
mths

Vitamin D (SD) or [IQR]

Placebo (SD) or [IQR]

Adjusted treatment effect (95% CI)a

5.1 (1.3)
4.8 (1.1)
2.9 (0.9)
2.8 (0.8)
1.4 (0.5)
1.3 (0.4)
1.4 (0.9)
1.4 (0.8)
1.1 [3.2]
1.9 [2.9]
0.7 [8.3]
0.9 [5.8]
2.4 (1.8)
2.4 (1.9)
1.7 (0.8)
1.5 (0.6)
377 (134)
2046 (349)

5.5 (1.1)
5.7 (1.2)
3.4 (1.1)
3.6 (1.2)
1.2 (0.3)
1.3 (0.3)
2.1 (1.0)
1.9 (0.9)
3.9 [5.4]
2.2 [4.2]
2.1 [9.4]
1.8 [2.8]
2.7 (1.9)
2.1 (1.1)
1.9 (0.8)
2.1 (1.2)
448 (72)
2113 (579)

0.0

0.3

0.1

0.2
0.0

0.1
0.0
0.2

2.7

0.3
0.0
0.4
0.2
0.7
0.2

0.3

32

146

0.82
0.10
0.55
0.31
0.72
0.14
0.90
0.31
0.09
0.75
0.98
0.60
0.66
0.11
0.06
0.28
0.30
0.37

(
0.3e0.4)
(
0.8e0.1)
(
0.4e0.2)
(
0.6ee0.2)
(
0.1e0.1)
(
0.2e0.0)
(
0.4e0.5)
(
0.2e0.5)
(
5.8e0.4)
(
2.4e1.7)
(
4.0e4.0)
(
1.1e1.9)
(
0.7e1.1)
(
0.2e1.6)
(
0.4e0.0)
(
0.8e0.3)
(
93e29)
(
471e179)

LDL, low density lipoprotein; HDL, high density lipoprotein; HOMA-IR, homeostasis model assessment e insulin resistance; IL-6, interleukin 6;
TNF, tumour necrosis factor; d-ROMS, derivatives of Reactive Oxygen Metabolites; BAP, biological antioxidant potential.
a
Adjusted for baseline outcome measure values.

Effect of intermittent vitamin D3 in chronic fatigue syndrome

293

Table 5 Symptom scores.

Vitamin D
(SD)

Placebo
(SD)

Piper fatigue scale

Behavioural
6.7 (1.6)
7.5 (2.2)
Affective
7.6 (1.4)
7.5 (2.2)
Sensory
6.5 (1.5)
7.1 (2.2)
Cognitive
5.5 (1.1)
5.9 (1.7)
Piper total
6.6 (1.1)
7.0 (1.9)
Hospital anxiety and depression score
Anxiety
5.7 (4.0)
5.0 (4.4)
Depression
5.7 (3.1)
7.6 (4.6)
a

Adjusted
treatment
effect (95% CI)a

0.1 (
1.1e1.3)
0.3 (
0.6e1.3)

0.1 (
1.2e1.0)
0.1 (
0.8e1.0)
0.2 (
0.8e1.2)

0.88
0.48
0.85
0.85
0.73

0.4 (
0.9e1.8)

1.0 (
2.6e0.5)

0.53
0.18

In conclusion, we found no evidence that intermittent,

high-dose vitamin D supplementation improves vascular
health, inammation, oxidative stress or symptoms in
patients with CFS/ME, although the intervention appeared
not to worsen any outcome measures.
Conict of interest
MDW and FK have received previous funding for vitamin D
research from Diabetes UK; Chief Scientist Ofce, Scottish
Government; Chest Heart and Stroke Scotland; Heart
Research UK; Tenovus Scotland.

Adjusted for baseline outcome measure values.

Acknowledgements
randomisation; this, together with the small size of the
trial, makes interpretation of the results more difcult. Our
trial was conducted in a single centre, on an overwhelmingly white, adult cohort; thus the results cannot
necessarily be generalised to patients from other ethnic
groups, children or other geographic areas. Alternative
scales for measuring fatigue (e.g. the Krupp Fatigue
Severity Scale) may provide better sensitivity and face
validity in CFS [26,27].
It is reassuring that no excess of adverse events was
seen in the treatment group. An aetiological hypothesis
that has been proposed previously suggests that low
vitamin D levels are a host response to infection by unknown microbes [28], and further suggests that vitamin D
supplementation could worsen the chronic infection by
suppressing the host response to such an infection. We
found no evidence of worsening symptoms over the six
month study follow up to support this hypothesis,
although our study was not designed to examine longterm outcomes or any rebound effects after cessation of
therapy. Although we did not nd evidence of benet on
the selected vascular markers studied in this trial, it is of
interest that PTH levels were signicantly lower in the
vitamin D group by six months. PTH may have vasculotoxic
effects independent of low 25OHD levels [29], and such
effects may be manifest in ways that we did not measure
in this trial, for example by promoting left ventricular
hypertrophy [30].

Table 6 Adverse events.

Infection
Musculoskeletal
Gastrointestinal
Dermatological
Cardiovascular
New cancer
Hypercalcemia
Ureteric colic
Other
Total:
Deaths
Hospitalisations

Vitamin D

Placebo

14
2
6
0
1
2
0
0
8
33
0
1

9
2
12
1
2
0
0
0
13
39
0
0

Funded by ME Research UK, grant number MERUK/2009/2.

We acknowledge the nancial support of NHS Research
Scotland through NHS Tayside in facilitating this research.
We thank Merck KGaA for donation of Vigantol oil and
placebo oil. The funder had no role in the management,
analysis or interpretation of the trial.
References
[1] Spence VA, Kennedy G, Belch JJ, Hill A, Khan F. Low-grade
inammation and arterial wave reection in patients with chronic
fatigue syndrome. Clin Sci (Lond) 2008 Apr;114(8):561e6.
[2] Newton DJ, Kennedy G, Chan KK, Lang CC, Belch JJ, Khan F. Large
and small artery endothelial dysfunction in chronic fatigue syndrome. Int J Cardiol 2012;154(3):335e6.
[3] Hollingsworth KG, Hodgson T, Macgowan GA, Blamire AM,
Newton JL. Impaired cardiac function in chronic fatigue syndrome
measured using magnetic resonance cardiac tagging. J Intern Med
2012 Mar;271(3):264e70.
[4] Beveridge LA, Witham MD. Vitamin D and the cardiovascular
system. Osteoporos Int 2013;24(8):2167e80.
[5] Elamin MB, Abu Elnour NO, Elamin KB, Fatourechi MM, Alkatib AA,
Almandoz JP, et al. Vitamin d and cardiovascular outcomes: a
systematic review and meta-analysis. J Clin Endocrinol Metab
2011;96(7):1931e42.
[6] Pittas AG, Chung M, Trikalinos T, Mitri J, Brendel M, Patel K, et al.
Systematic review: vitamin D and cardiometabolic outcomes. Ann
Intern Med 2010;152(5):307e14.
[7] Berkovitz S, Ambler G, Jenkins M, Thurgood S. Serum 25-hydroxy
vitamin D levels in chronic fatigue syndrome: a retrospective
survey. Int J Vitam Nutr Res 2009;79(4):250e4.
[8] Witham MD, Kennedy GY, Belch JJF, Hill A, Khan F. Association
between vitamin D status and markers of vascular health in patients with chronic fatigue syndrome/myalgic encephalomyelitis
(CFS/ME). Int J Cardiol 2014;174(1):139e40.
[9] Reid D, Toole BJ, Knox S, Talwar D, Harten J, OReilly DS, et al. The
relation between acute changes in the systemic inammatory
response and plasma 25-hydroxyvitamin D concentrations after
elective knee arthroplasty. Am J Clin Nutr 2011;93(5):1006e11.
[10] Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A.
The chronic fatigue syndrome: a comprehensive approach to its
denition and study. International Chronic Fatigue Syndrome
Study Group. Ann Intern Med 1994;121(12):953e9.
[11] Carruthers BM, Jain AK, De Meirleir KL. Myalgic encephalomyelitis/chronic fatigue syndrome: clinical working case denition,
diagnostic and treatment protocols. J Chronic Fatigue Sy 2003;
11(1):117.
[12] Ben-Shlomo Y, Spears M, Boustred C, May M, Anderson SG,
Benjamin EJ, et al. Aortic pulse wave velocity improves cardiovascular event prediction: an individual participant meta-analysis
of prospective observational data from 17,635 subjects. J Am Coll
Cardiol 2014;63(7):636e46.

294
[13] Al Mheid I, Patel R, Murrow J, Morris A, Rahman A, Fike L, et al.
Vitamin D status is associated with arterial stiffness and vascular
dysfunction in healthy humans. J Am Coll Cardiol 2011;58(2):
186e92.
[14] Laurent S, Cockcroft J, Van Bortel L, Boutouyrie P, Giannattasio C,
Hayoz D, et al. Expert consensus document on arterial stiffness:
methodological issues and clinical applications. Eur Heart J 2006;
27(21):2588e605.
[15] Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D,
Charbonneau F, Creager MA, et al. Guidelines for the ultrasound
assessment of endothelial-dependent ow-mediated vasodilation of the brachial artery: a report of the International Brachial
Artery Reactivity Task Force. J Am Coll Cardiol 2002;39(2):
257e65.
[16] Piper B. Measuring fatigue. In: Stromborg M, Olsen J, editors. Instruments for clinical health care research. London: Jones and
Barlett; 1997. p. 482e96.
[17] Rodriguez T. The challenge of evaluating fatigue. J Am Acad Nurse
Pract 2000;12(8):329e38.
[18] Favrat B, Balck K, Breymann C, Hedenus M, Keller T, Mezzacasa A,
et al. Evaluation of a single dose of ferric carboxymaltose in
fatigued, iron-decient women e PREFER a randomized, placebocontrolled study. PLoS One 2014;9(4):e94217.
[19] Mihrshahi R, Beirman R. Aetiology and pathogenesis of chronic
fatigue syndrome: a review. N Z Med J 2005;118(1227):U1780.
[20] Morris G, Berk M, Galecki P, Maes M. The emerging role of autoimmunity in myalgic encephalomyelitis/chronic fatigue syndrome
(ME/CFS). Mol Neurobiol 2014;49(2):741e56.
[21] Pedersen BK. The anti-inammatory effect of exercise: its role in
diabetes and cardiovascular disease control. Essays Biochem 2006;
42:105e17.

M.D. Witham et al.

[22] Bischoff-Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson-Hughes B. Estimation of optimal serum concentrations of 25hydroxyvitamin D for multiple health outcomes. Am J Clin Nutr
2006;84(1):18e28.
[23] Sugden JA, Davies JI, Witham MD, Morris AD, Struthers AD.
Vitamin D improves endothelial function in patients with type 2
diabetes mellitus and low vitamin D levels. Diabet Med 2008;25:
320e5.
[24] Witham MD, Dove FJ, Dryburgh M, Sugden JA, Morris AD,
Struthers AD. The effect of different doses of vitamin D(3) on
markers of vascular health in patients with type 2 diabetes: a
randomised controlled trial. Diabetologia 2010;53(10):2112e9.
[25] Witham MD, Dove FJ, Sugden JA, Doney AS, Struthers AD. The
effect of vitamin D replacement on markers of vascular health in
stroke patients - a randomised controlled trial. Nutr Metab Cardiovasc Dis 2012;22(10):864e70.
[26] Taylor RR, Jason LA, Torres A. Fatigue rating scales: an empirical
comparison. Psychol Med 2000;30(4):849e56.
[27] Jason LA, Evans M, Brown M, Porter N, Brown A, Hunnell J, et al.
Fatigue scales and chronic fatigue syndrome: issues of sensitivity
and specicity. Disabil Stud Q 2011;31(1).
[28] Waterhouse JC, Perez TH, Albert PJ. Reversing bacteria-induced
vitamin D receptor dysfunction is key to autoimmune disease.
Ann N Y Acad Sci 2009;1173:757e65.
[29] Kestenbaum B, Katz R, de Boer I, Hoofnagle A, Sarnak MJ,
Shlipak MG, et al. Vitamin D, parathyroid hormone and cardiovascular events among older adults. J Am Coll Cardiol 2011;58(14):
1433e41.
[30] Saleh FN, Schirmer H, Sundsfjord J, Jorde R. Parathyroid hormone
and left ventricular hypertrophy. Eur Heart J 2003;24(22):
2054e60.