Lower Vitamin D Levels Are Associated With Depression in People With Chronic Spinal Cord Injury

Accepted Manuscript Lower vitamin D levels are associated with depression among people with chronic spinal cord injury Arcangelo Barbonetti, MD, PhD, Francesca Cavallo, PhD, Settimio D’Andrea, MD, Mario Muselli, MD, Giorgio Felzani, MD, Sandro Francavilla, MD, Felice Francavilla, MD PII:

S0003-9993(16)31296-5

DOI:

10.1016/j.apmr.2016.11.006

Reference:

YAPMR 56737

To appear in:

ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION

Received Date: 2 June 2016 Revised Date:

9 November 2016

Accepted Date: 14 November 2016

Please cite this article as: Barbonetti A, Cavallo F, D’Andrea S, Muselli M, Felzani G, Francavilla S, Francavilla F, Lower vitamin D levels are associated with depression among people with chronic spinal cord injury, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2017), doi: 10.1016/ j.apmr.2016.11.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Running head: Vitamin D and depression in SCI

2

Lower vitamin D levels are associated with depression among people with chronic spinal cord

4

injury

5

Arcangelo Barbonetti, MD, PhD,a Francesca Cavallo, PhD,a Settimio D’Andrea, MD,b Mario

6

Muselli, MD,b Giorgio Felzani, MD,a Sandro Francavilla, MD,b Felice Francavilla, MDb

RI PT

3

7 a

Spinal Unit, San Raffaele Sulmona Institute, Sulmona, Italy; and bDepartment of Life, Health and

9

Environment Sciences, University of L’Aquila, L’Aquila, Italy

10 11

M AN U

SC

8

Disclosure statement: The authors declare no competing financial interests.

12

Corresponding author

14

Arcangelo Barbonetti, MD, PhD

15

Spinal Unit, San Raffaele Sulmona Institute,

16

Via dell’Agricoltura snc,

17

67039 Sulmona, Italy

18

Phone: +39 0864 25071;

19

E-mail address: [email protected]

AC C

EP

TE D

13

1

ACCEPTED MANUSCRIPT 1

Lower vitamin D levels are associated with depression among people with chronic spinal cord

2

injury

3

Abstract

5

Objective: To determine 1) whether serum concentration of 25-hydroxyvitamin D (25(OH)D) was

6

associated with depression levels in people with chronic spinal cord injury (SCI), and 2) whether

7

any observed association was independent of potential confounders.

8

Design: Cross-sectional study.

9

Participants and Setting: One hundred patients with chronic SCI, admitted to a rehabilitation

M AN U

SC

RI PT

4

program, were consecutively recruited.

11

Main Outcome Measures: Patients underwent clinical and biochemical evaluations, including

12

assessment of 25(OH)D and the presence and severity of depressive symptoms by the interviewer-

13

assisted self-report Beck Depression Inventory-II (BDI-II).

14

Results: Depression (BDI-II score ≥14) was observed in 15 out of 28 women (53.6%) and 18 out of

15

72 men (25.0%) of the study population. They exhibited significantly lower 25(OH)D levels, lower

16

function independence in daily living activities, poorer leisure time physical activity and higher

17

body mass index.

18

Lower 25(OH)D levels were associated with higher BDI-II scores, as well as with the occurrence of

19

depression. These associations persisted after adjustment for all significant predictors of the BDI-II

20

score, selected, as possible confounders, by univariate analysis. At ROC analysis, a 25(OH)D level

21

<9,99 ng/mL had the highest accuracy in discriminating patients with depression.

22

Conclusions: Among people with chronic SCI, an inverse association exists between serum

23

25(OH)D levels and depressive symptoms, widely independent of potential confounders, especially

24

those, peculiar to this population, which can mediate the effects of depression on vitamin D levels.

AC C

EP

TE D

10

25 26

Keywords: spinal cord trauma, paraplegia, tetraplegia, 25(OH)D, Beck Depression Inventory-II

2

ACCEPTED MANUSCRIPT 27

Abbreviations: ADL, activities of daily living; ASIA, American Spinal Injury Association; BDI-II,

29

Beck Depression Inventory-II; BMI, body mass index; LTPA, leisure time physical activity; NRS,

30

Numerical Rating Scale; PTH, parathyroid hormone; ROC, Receiver Operating Characteristics;

31

SCI, spinal cord injury; SCIM, spinal cord independence measure; 25(OH)D, 25-hydroxyvitamin D.

RI PT

28

32

Introduction

34

Apart from the well-known key role of the vitamin D in calcium homeostasis and bone health, there

35

is growing evidence, over the last few decades, of its widespread pleiotropic effects. Observational

36

data have been produced on the link between low serum levels of 25-hydroxyvitamin D (25(OH)D)

37

and a higher hazard of a number of chronic diseases.1-4

38

An association between lower vitamin D levels and depression has been also reported by some

39

studies5-10 but not by others.11-13 Actually, there is a biological plausibility of a causal link between

40

hypovitaminosis D and depression. In particular, receptors of vitamin D have been detected in

41

regions of the brain involved in depression,14 vitamin D response elements have been identified in

42

the promoter regions of serotonin genes,15 and in vitamin D receptor knockout mice an increased

43

anxiety-like phenotype was observed,16,17 similar to the emotional behavior seen in animal models

44

of depression. On the other hand, the link between hypovitaminosis D and depression might be due

45

to confounders, represented by risk factors shared by both conditions.

46

The relationship between low levels of vitamin D and depression has never been investigated in

47

patients with chronic SCI, in spite of a well-documented high prevalence of both vitamin D

48

deficiency18-21 and depression22 in this population. A combination of physical inactivity, due to

49

muscle wasting and disability leading to poor sunlight exposure, along with unbalanced diet,

50

coexisting illnesses and obesity, the intake of drugs that induce liver microsomal enzymes

51

accelerating vitamin D catabolism, such as antiepileptic drugs for neuropathic pain treatment, can

52

result in hypovitaminosis D. An interplay between hypovitaminosis D and physical function is also

AC C

EP

TE D

M AN U

SC

33

3

ACCEPTED MANUSCRIPT particularly plausible in this population because hypovitaminosis D could in turn directly contribute

54

in worsening of skeletal muscle health and physical function: we recently demonstrated that in

55

people with chronic SCI, a low 25(OH)D level is an independent predictor of poor physical

56

function.21 Depression could play a role in this scenario, as depression can hinder outdoor activities

57

and rehabilitation favoring hypovitaminosis D. A vicious circle including depression is conceivable

58

whether hypovitaminosis D could play a causal role in depression.

59

In this study we aimed to determine whether serum concentration of 25(OH)D was associated with

60

depression levels in people with chronic SCI, and whether any observed association was

61

independent of potential confounders.

M AN U

SC

RI PT

53

62

Methods

64

Study population

65

One hundred patients consecutively admitted to a neuromotor rehabilitation program for paraplegia

66

or tetraplegia at the Spinal Unit of San Raffaele Institute of Sulmona because of traumatic SCI were

67

enrolled. All patients had a history of SCI lasting longer than 1 year. No patient received vitamin D

68

replacement therapy. Coexisting chronic illnesses were recorded, including heart diseases, chronic

69

kidney disease, diabetes, hypertension, and dyslipidemia. No patient had severe acute or chronic

70

morbidities hindering the rehabilitative project. No patient had communication or cognitive

71

disorders. The study was approved by the local ethics committee (ASL 1 Abruzzo) and all

72

participants signed a written informed consent.

EP

AC C

73

TE D

63

74

Clinical evaluations

75

Patients underwent detailed clinic and neurologic examination by the same experienced physician

76

(A.B.), according to International Standards for Neurological Examination and Functional

77

Classification of Spinal Cord Injury guidelines, and the American Spinal Injury Association (ASIA)

4

ACCEPTED MANUSCRIPT 78

protocol was used to define level and completeness of SCI. The Numerical Rating Scale (NRS),

79

scoring from 0 to 10, was used to assess the presence and intensity of pain.

80

Measures

82

The degree of functional independence attained in activities of daily living (ADL) was assessed by

83

a physiatrist (G.F.) using the spinal cord independence measure (SCIM) and leisure time physical

84

activity (LTPA) was quantified in minutes/week using the LTPA Questionnaire for people with SCI

85

(LTPAQ-SCI), as previously described.21,23,24 Briefly, the LTPAQ-SCI is an SCI-specific measure

86

of minutes of mild, moderate, and heavy intensity LTPA performed over the previous 7 days,

87

according to the physical activity guidelines for spinal cord-injured population.25 Participants used

88

an intensity classification chart to discriminate mild, moderate, and heavy intensity LTPA,

89

according to the perceived psychophysical effort. For each intensity degree, participants recalled the

90

number of days, over the last 7 days, that they performed LTPA at each intensity. Next, they

91

recalled how many minutes per day they spent performing LTPA at that intensity. The scale was

92

scored by calculating the total number of weekly minutes of activity at each intensity (number of

93

days of activity × number of minutes of activity). As total LTPAQ-SCI scores were strongly

94

correlated with mild, moderate, and heavy subscores,23 only total scores were used for analyses.

95

The presence and severity of depressive symptoms were assessed by the interviewer-assisted self-

96

report Beck Depression Inventory-II (BDI-II),26 which was administered by the same experienced

97

psychologist (F.C.) at admission. The BDI-II, a revision of the BDI, originally designed to measure

98

the degree of depression in psychiatric patients,27 is used as a 21-item community screening

99

instrument for depressive symptoms9,28 and it has been also used for people with SCI.29,30

100

Previously published cut-off points were used to identify patients with “no depression” (score <14)

101

vs “mild to severe depression” (≥14).9,26

AC C

EP

TE D

M AN U

SC

RI PT

81

102 103

Biochemistry and hematology

5

ACCEPTED MANUSCRIPT A fasting morning venous blood sample was taken from each patient. Levels of 25(OH)D and

105

parathyroid hormone (PTH) were determined using chemiluminescent immunoassays.a Intra- and

106

inter-assay variation coefficients for 25(OH)D assessment were 4.5% and 8.5%, respectively.

107

Standard methods and commercial kitsb were used for all of the other biochemical/hematologic

108

measurements.

RI PT

104

109

Statistical analysis

111

Statistical analysis was performed using the R statistical software, version 3.0.3.c After ascertaining

112

the skewed distribution of all data with the Shapiro-Wilk test, the Wilcoxon rank-sum test was used

113

to compare continuous variables, while proportional differences were assessed by the chi-square

114

test or the Fisher exact test, as appropriate. Correlation between 25(OH)D levels and BDI-II score

115

was explored by Spearman’s rank test. A multivariate linear regression analysis of square root-

116

transformed values of all continuous variables was used to assess the independent association of

117

higher BDI-II scores with significant predictors selected by univariate analysis. The independent

118

association between 25(OH)D levels and depression (BDI-II score ≥14) was assessed by logistic

119

regression analysis with adjustment models for possible confounders. A Receiver Operating

120

Characteristics (ROC) analysis over 25(OH)D levels was used to detect a threshold value providing

121

an accurate discriminating ability in predicting depression.

122

The sample size of 100 subjects exhibited a statistical power >90% in revealing a correlation

123

coefficient (rho) and a β-coefficient = 0.5 between 25(OH)D levels and BDI-II score with an α

124

level <0.05 at 2 tails test.

AC C

EP

TE D

M AN U

SC

110

125 126

Results

127

Characteristics of the study population

6

ACCEPTED MANUSCRIPT A BDI-II score ≥14 (depression) was observed in 15 out of 28 women (53.6%) and in 18 out of 72

129

men (25.0%) of the study population. Table 1 shows the characteristics of the study population

130

categorized by BDI-II depression status.

131

Serum 25(OH)D levels were significantly lower and PTH levels significantly higher in patients with

132

BDI-II score ≥14 (depressed) with respect to those scoring <14 (non-depressed). Patients with

133

depression exhibited a higher body mass index (BMI), were engaged in a significantly poorer

134

LTPA, had a lower function independence degree in performing ADL (as evaluated by SCIM) and

135

reported a higher intake of psychotropic drugs compared with non-depressed patients. Level,

136

completeness and distance from injury did not differ significantly between the two groups.

M AN U

SC

RI PT

128

137

Univariate associations

139

As shown in table 2, at the univariate regression analysis including putative determinants of

140

depression status, a higher BDI-II score was significantly associated with lower 25(OH)D levels,

141

higher BMI, shorter distance from injury, poorer engagement in LTPA and lower functional

142

independence degree. Other significant predictors of a higher BDI-II score were the female gender,

143

the spring/summer season and the intake of psychotropic medications. At the Spearman’s rank test

144

25(OH)D levels were strongly correlated with BDI-II score (r = -0.67; p < 0.0001).

EP

AC C

145

TE D

138

146

Multivariate regressions

147

All significant predictors of a higher BDI-II score, selected by the univariate analysis, were

148

included in a multivariate linear regression analysis. Lower 25(OH)D levels were independently

149

associated with a higher BDI-II score, explaining 50% (β = -0.82; p <0.0001) of the BDI-II score

150

variability (Fig. 1). Among the other predictors, only distance from injury and LTPA exhibited a

151

significant independent inverse association with BDI-II score, explaining, however, only 4% and

152

6% of its variability, respectively (Fig. 1).

7

ACCEPTED MANUSCRIPT As shown in table 3, at the multivariate logistic regression analysis, lower 25(OH)D levels were

154

associated with depression (BDI-II score ≥14), after adjustment for gender, BMI, seasonality,

155

distance from injury and intake of psychotropic medications. This association persisted after further

156

adjustment for LTPA and functional independence degree.

157

At ROC analysis, a 25(OH)D level <9,99 ng/mL discriminated patients with depression with a

158

sensitivity of 72.7% and a specificity of 79.1% (AUC: 0.84; 95% CI: 0.76-0.91) (Fig. 2).

159

When the study population was categorized according to this cut-off value, depression was

160

exhibited by 23 of 36 patients with 25(OH)D <9.99 ng/mL (63.8%) but only by 10 of 64 patients

161

with 25(OH)D levels ≥9.99 ng/mL (15.6%) (p <0.0001).

M AN U

SC

RI PT

153

162

Discussion

164

In the present study we examined, for the first time, the relationship between the occurrence of

165

depressive symptoms, assessed by the BDI-II, and serum levels of 25(OH)D in subjects with

166

chronic SCI. Lower 25(OH)D levels were associated with a higher BDI-II score, as well as with the

167

occurrence of a depression status (BDI-II ≥14). These associations persisted after adjustment for

168

confounders. A 25(OH)D level <9,99 ng/mL had the highest accuracy in discriminating patients

169

with depression.

170

When the association between 25(OH)D levels and depression was explored in the general

171

population, the largest cross-sectional studies produced controversial results after adjustment for

172

confounders. No association was reported in a sample of 3,916 subjects aged >20 years, enrolled in

173

the National Health and Nutrition Examination Survey (NHANES).12 No association had been also

174

reported by a previous study on a sample of 3,262 community residents middle-aged and elderly

175

Chinese.11 On the contrary, significant independent associations were reported in 7,970 subjects

176

aged 15-39 years participating in the NHANES III,6 in a sample of 3,369 middle-aged and older

177

men participating in the European Male Ageing Study (EMAS)9 and in a national population survey

178

on 4,002 Jordanian subjects.10 Although the association was significant in a meta-analysis by

AC C

EP

TE D

163

8

ACCEPTED MANUSCRIPT Anglin et al.,31 according to the authors, the quality of the evidence from most of the studies was

180

low. Prospective studies on incident depression could be more informative. However, 3 cohort

181

studies included in the meta-analysis by Anglin31 reported controversial results, as no association

182

was found in older Chinese men,13 whereas an association was found in older Italian women and

183

men, participating in the InCHIANTI Study7 and in patients with cardiovascular disease.8 Although

184

an increased hazard ratio of depression for the lowest vs highest vitamin D categories was reported,

185

as stated by the authors of the meta-analysis, all the cohort studies had problems with bias

186

especially the largest one,8 they had significant heterogeneity and lacked precision. Therefore

187

uncertainty remains about the actual association between hypovitaminosis D and depression.

188

The first datum arising from this study is the strong linear inverse association between vitamin D

189

and depression levels in subjects with chronic SCI. Although the demonstration of the independence

190

of this association in this population was a challenge, due to the high prevalence of correlates

191

shared by both of these conditions, the significant linear inverse association persisted at the

192

multivariate regression analysis, including all significant predictors of a higher depression level

193

(BDI-II score) selected by the univariate analysis. Noteworthy, vitamin D level was the strongest

194

predictor of the depression levels, explaining 50% (β=-0.82; p <0.0001) of the BDI-II score

195

variability. Only distance from injury and LTPA also exhibited a significant independent inverse

196

association with BDI-II score (Fig. 1), but with a much lower predictive power.

197

A strong association was also observed between lower vitamin D levels and depression status (BDI-

198

II ≥14). It persisted at the multivariate logistic regression analysis after progressive adjustments for

199

confounding factors, also including (in the Model III, Table 3) physical function measures (LTPA

200

and SCIM), as major putative confounders. This is relevant to shed light on the nature of this

201

association, which is still uncertain. Other than a direct causal role of hypovitaminosis D in

202

depression, both conditions could represent markers of poor health without a direct interaction, or

203

depression itself may contribute to lower 25(OH)D levels by reducing the outdoor physical activity,

204

with consequent poorer sunlight exposure and higher risk of obesity. The persistence of a strong

AC C

EP

TE D

M AN U

SC

RI PT

179

9

ACCEPTED MANUSCRIPT association after a careful adjustment for confounders, especially the aforesaid ones mediating the

206

effect of depression on vitamin D levels, strengthens the likelihood of a direct causal link of

207

hypovitaminosis D with depression.

208

Although we found significantly higher levels of PTH among patients with depression with respect

209

to those without depression, this variable was not included in the multivariate regression analysis,

210

because no significant linear association between PTH levels and BDI-II score was revealed at the

211

univariate regression analysis and also because PTH was not available for all patients. However, no

212

independent association had been found in previous studies on able-bodied people.8-10 Actually, a

213

putative role of PTH in the etiology of neuropsychological disorders has been linked to primary

214

hyperparathyroidism32 and hypercalcemia.33 Although a group of participants from the Tromsø

215

health study with secondary hyperparathyroidism exhibited a higher score on the cognitive/affective

216

BDI-II subscale when compared to controls,34 this association was univariate and not adjusted for

217

potential confounding factors.

218

With regard to vitamin D deficiency, when the bone health was used as primary outcome, levels

219

below 20 ng/mL have been suggested as insufficient.1 Interestingly, in the present study, at ROC

220

analysis, a 25(OH)D value close to 10 ng/mL was the cut-off with the highest accuracy in

221

discriminating patients with depression (BDI-II ≥14), with a sensitivity of 72.7% and a specificity

222

of 79.1%.

223

Study limitations and strengths

224

Some limitations can be recognized in the present study, firstly, the limited sample size. It,

225

however, was quite large considering the peculiar type of study population, which, moreover, was

226

rather homogeneous. In fact, only patients with a neurologically stable traumatic SCI lasting longer

227

than 1 year were included in the study, thus ruling out any interference of clinical variables related

228

to sub-acute phase, and no patient had severe chronic or acute illnesses hindering the rehabilitative

229

program. Likely this could be relevant to the demonstration of significant and independent

230

associations, in spite of the limited size of our study population. Moreover, although analyses were

AC C

EP

TE D

M AN U

SC

RI PT

205

10

ACCEPTED MANUSCRIPT adjusted for several potential confounders, we cannot be certain that other unmeasured confounding

232

factors have not influenced the associations under investigation. In particular, sexual and

233

bowel/bladder dysfunctions may play a role in determining depressive symptoms.35 However, the

234

severity of the impairment of sexual and bowel/bladder functions is dependent upon the level and

235

completeness of SCI,35 which were analysed in the present study: both level and completeness of

236

SCI were not different between depressed and non-depressed groups and they were not associated

237

to BDI-II score at the univariate regression analysis. The demonstration of the strong association

238

between 25(OH)D levels and depressive symptoms in people with chronic SCI, despite the limited

239

sample size of our series, could also lies in the high prevalence of hypovitaminosis D and

240

depression in this population, as well as in the careful selection and assessment of confounders and

241

putative determinants shared by both of these conditions. In particular, in the present study, the

242

entire spectrum of physical activities was explored. In fact, besides the functional independence in

243

performing activities of daily living (scored by the SCIM), we also assessed leisure time physical

244

activities (quantified by the LTPA Questionnaire), a major confounding factor, as related to sunlight

245

exposure and risk of obesity.

EP

246

TE D

M AN U

SC

RI PT

231

Conclusions

248

Among people with chronic SCI, an inverse association exists between serum 25(OH)D levels and

249

depression symptoms, widely independent of potential confounders, especially those, peculiar to

250

this population, which can mediate the effect of depression on vitamin D levels. This strengthens

251

the likelihood of a direct causal link of hypovitaminosis D with depression, which is still

252

controversial in the general population. Prospective/intervention studies could reinforce the

253

evidence of the independence of this association from confounding factors peculiar to people with

254

SCI, in favor of its generalizability.

AC C

247

255 256

References

11

ACCEPTED MANUSCRIPT 1. Holick MF. Vitamin D deficiency. N Engl J Med 2007;357:266-81.

258

2. Pittas AG, Lau J, Hu FB, Dawson-Hughes B. The role of vitamin D and calcium in type 2

259

diabetes. A systematic review and meta-analysis. J Clin Endocrinol Metab 2007;92:2017-29.

260

3. Pilz S, Dobnig H, Winklhofer-Roob B, Riedmüller G, Fischer JE, Seelhorst U, Wellnitz B,

261

Boehm BO, März W. Low serum levels of 25-hydroxyvitamin D predict fatal cancer in

262

patients referred to coronary angiography. Cancer Epidemiol Biomarkers Prev

263

2008;17:1228-33.

systematic review. Nat Rev Cardiol 2009;6:621-30.

SC

265

4. Pilz S, Tomaschitz A, Ritz E, Pieber TR. Vitamin D status and arterial hypertension: a

M AN U

264

RI PT

257

266

5. Hoogendijk WJ, Lips P, Dik MG, Deeg DJ, Beekman AT, Penninx BW. Depression is

267

associated with decreased 25-hydroxyvitamin D and increased parathyroid hormone levels

268

in older adults. Arch Gen Psychiatry 2008;65:508-12.

6. Ganji V, Milone C, Cody MM, McCarty F, Wang YT. Serum vitamin D concentrations are

270

related to depression in young adult US population: the Third National Health and Nutrition

271

Examination Survey. Int Arch Med 2010;3:29.

TE D

269

7. Milaneschi Y, Shardell M, Corsi AM, Vazzana R, Bandinelli S, Guralnik JM, Ferrucci L.

273

Serum 25-hydroxyvitamin D and depressive symptoms in older women and men. J Clin

274

Endocrinol Metab 2010;95:3225-33.

AC C

EP

272

275

8. May HT, Bair TL, Lappé DL, Anderson JL, Horne BD, Carlquist JF, Muhlestein JB.

276

Association of vitamin D levels with incident depression among a general cardiovascular

277

population. Am Heart J 2010;159:1037-43.

278

9. Lee DM, Tajar A, O'Neill TW, O'Connor DB, Bartfai G, Boonen S, Bouillon R, Casanueva

279

FF, Finn JD, Forti G, Giwercman A, Han TS, Huhtaniemi IT, Kula K, Lean ME, Punab M,

280

Silman AJ, Vanderschueren D, Wu FC, Pendleton N; EMAS study group.Lower vitamin D

281

levels are associated with depression among community-dwelling European men. J

282

Psychopharmacol 2011;25:1320-8.

12

ACCEPTED MANUSCRIPT 283

10. Jaddou HY, Batieha AM, Khader YS, Kanaan SH, El-Khateeb MS, Ajlouni KM. Depression

284

is associated with low levels of 25-hydroxyvitamin D among Jordanian adults: results from

285

a national population survey. Eur Arch Psychiatry Clin Neurosci 2012;262:321-7.

287

11. Pan A, Lu L, Franco OH, Yu Z, Li H, Lin X. Association between depressive symptoms and

RI PT

286

25-hydroxyvitamin D in middle-aged and elderly Chinese. J Affect Disord 2009;118:240-3. 12. Zhao G, Ford ES, Li C. Associations of serum concentrations of 25-hydroxyvitamin D and

289

parathyroid hormone with surrogate markers of insulin resistance among US adults without

290

physician-diagnosed diabetes: NHANES, 2003–2006. Diabetes Care 2010;33:344-7.

SC

288

13. Chan R, Chan D, Woo J, Ohlsson C, Mellström D, Kwok T, Leung P. Association between

292

serum 25-hydroxyvitamin D and psychological health in older Chinese men in a cohort

293

study. J Affect Disord 2011;130:251-9.

294 295

M AN U

291

14. Eyles DW, Smith S, Kinobe R, Hewison M, McGrath JJ. Distribution of the vitamin D receptor and 1 alpha-hydroxylase in human brain. J Chem Neuroanat 2005;29:21-30. 15. Wang TT, Tavera-Mendoza LE, Laperriere D, Libby E, MacLeod NB, Nagai Y, Bourdeau

297

V, Konstorum A, Lallemant B, Zhang R, Mader S, White JH. Large-scale in silico and

298

microarray-based identification of direct 1,25-dihydroxyvitamin D3 target genes. Mol

299

Endocrinol 2005;19:2685-95.

301 302 303

EP

16. Burne TH, Johnston AN, McGrath JJ, Mackay-Sim A. Swimming behaviour and post-

AC C

300

TE D

296

swimming activity in Vitamin D receptor knockout mice. Brain Res Bull 2006;69:74-8. 17. Kalueff AV, Lou YR, Laaksi I, Tuohimaa P. Increased anxiety in mice lacking vitamin D receptor gene. Neuroreport 2004;15:1271-4.

304

18. Oleson CV, Patel PH, Wuermser LA. Influence of season, ethnicity, and chronicity on

305

vitamin D deficiency in traumatic spinal cord injury. J Spinal Cord Med 2010;33:202-13.

306

19. Javidan AN, Sabour H, Latifi S, Vafa M, Shidfar F, Khazaeipour Z, Shahbazi F, Rahimi A,

307

Razavi SH. Calcium and vitamin D plasma concentration and nutritional intake status in

13

ACCEPTED MANUSCRIPT 308

patients with chronic spinal cord injury: a referral center report. J Res Med Sci 2014;19:881-

309

4. 20. Barbonetti A, Vassallo MR, Felzani G, Francavilla S, Francavilla F. Association between

311

25(OH)-vitamin D and testosterone levels: evidence from men with chronic spinal cord

312

injury. J Spinal Cord Med 2016;39:246-52.

RI PT

310

21. Barbonetti A, Sperandio A, Micillo A, D’Andrea S, Pacca F, Felzani G, Francavilla S,

314

Francavilla F. Independent association of vitamin D with physical function in people with

315

chronic spinal cord injury. Arch Phys Med Rehab 2016;97:726-32.

317

22. Williams R, Murray A. Prevalence of depression after spinal cord injury: a meta-analysis.

M AN U

316

SC

313

Arch Phys Med Rehab 2015;96:133-40.

318

23. Barbonetti A, Vassallo MR, Pacca F, Cavallo F, Costanzo M, Felzani G, Francavilla S,

319

Francavilla F. Correlates of low testosterone in men with chronic spinal cord injury.

320

Andrology 2014;2:721-8.

24. Barbonetti A, Caterina Vassallo MR, Cotugno M, Felzani G, Francavilla S, Francavilla F.

322

Low testosterone and non-alcoholic fatty liver disease: Evidence for their independent

323

association in men with chronic spinal cord injury. J Spinal Cord Med 2016 Feb 25:1-7.

324

25. Ginis KA, Hicks AL, Latimer AE, Warburton DE, Bourne C, Ditor DS, Goodwin DL,

325

Hayes KC, McCartney N, McIlraith A, Pomerleau P, Smith K, Stone JA, Wolfe DL. The

326

development of evidence-informed physical activity guidelines for adults with spinal cord

327

injury. Spinal Cord 2011;49:1088-96.

329 330 331

EP

AC C

328

TE D

321

26. Beck AT, Steer RA, Brown GK. Manual for the Beck Depression Inventory-II. San Antonio, TX: Psychological Corporation; 1996. 27. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561-71.

14

ACCEPTED MANUSCRIPT 332

28. Segal DL, Coolidge FL, Cahill BS, O’Riley AA. Psychometric properties of the Beck

333

Depression Inventory II (BDI-II) among community-dwelling older adults. Behav Modif

334

2008;32:3-20. 29. Ataoğlu E, Tiftik T, Kara M, Tunç H, Ersöz M, Akkuş S. Effects of chronic pain on quality

RI PT

335

of life and depression in patients with spinal cord injury. Spinal Cord 2013;51:23-6.

336

30. Xue S, Arya S, Embuldeniya A, Narammalage H, da Silva T, Williams S, Ravindran A.

338

Perceived functional impairment and spirituality/religiosity as predictors of depression in a

339

Sri Lankan spinal cord injury patient population. Spinal Cord 2016, May 3.

SC

337

31. Anglin RE, Samaan Z, Walter SD, McDonald SD. Vitamin D deficiency and depression in

M AN U

340

adults: systematic review and meta-analysis. Br J Psychiatry 2013;202:100-7.

341

32. Silverberg SJ, Bilezikian JP, Bone HG, Talpos GB, Horwitz MJ, Stewart AF. Therapeutic

343

controversies in primary hyperparathyroidism. J Clin Endocrinol Metab 1999;84:2275-85.

344

33. Schram MT, Trompet S, Kamper AM, de Craen AJ, Hofman A, Euser SM, Breteler MM,

345

Westendorp RG. Serum calcium and cognitive function in old age. J Am Geriatr Soc

346

2007;55:1786-92.

TE D

342

34. Jorde R, Waterloo K, Saleh F, Haug E, Svartberg J. Neuropsychological function in relation

348

to serum parathyroid hormone and serum 25-hydroxyvitamin D levels. The Tromsø study. J

349

Neurol 2006;253:464-70.

AC C

EP

347

350

35. Barbonetti A, Cavallo F, Felzani G, Francavilla S, Francavilla F. Erectile dysfunction is the

351

main determinant of psychological distress in men with spinal cord injury. J Sex Med

352

2012;9:830-6.

353 354

Suppliers

355

a

356

b

357

c

LIAISON® - DiaSorin (25(OH)D) and Medical Systems (PTH). Instrumentation Laboratory Co.

The R Foundation for Statistical Computing.

15

ACCEPTED MANUSCRIPT 358

Figure legends

360

Fig 1 Multivariate regression analysis of the relationship between putative predictors of depression

361

status (selected by univariate analysis) and Beck Depression Inventory-II (BDI-II) score.

362

Regression was performed on square root-transformed values. β-coefficient (95% CI) indicates the

363

standardized coefficient of regression.

RI PT

359

364

Fig 2 Receiver operating characteristics (ROC) analyses over 25(OH)D levels, showing the

366

threshold level which provides accurate discriminating ability in predicting a Beck Depression

367

Inventory-II (BDI-II) score ≥14.

AC C

EP

TE D

M AN U

SC

365

1

ACCEPTED MANUSCRIPT Table 1. Characteristics of the study population categorized by depression status BDI-II depression category Depressed

(BDI-II score <14)

(BDI-II score ≥14) n = 33

Demographic and lifestyle variables 49 [22-89]

Gender – no. (%) 54 (81.0)

Females

13 (19.0)

M AN U

Males

Educational level – no. (%) Primary School High School University

18 (55.0)

0.3

0.01

15 (45.0)

18 (29.6)

10 (30.3)

34 (50.7)

17 (51.5)

15 (22.4)

6 (18.2)

TE D

Employment – no. (%) Student

58 [20-79]

SC

Age (years)

p value

RI PT

No depression

n = 67

0.8

3 (4.5)

1 (3.0)

13 (19.4)

8 (24.0)

14 (20.9)

3 (9.0)

37 (55.2)

21 (64.0)

29 (43.3)

12 (36.4)

30 (44.7)

17 (51.5)

Divorced/Separated

5 (7.5)

3 (9.1)

Widowed

3 (4.5)

1 (3.0)

Current smokers – no. (%)

24 (35.8)

11 (33.3)

0.9

Alcohol intake ≥ 1 drink/day – no. (%)

35 (52.2)

11 (33.3)

0.11

525 [0-1225]

350 [0-1155]

0.02

16.1 [4.12-34.4]

9.51 [4-15.1]

<0.0001

9.4 [8.3-10.6]

9.3 [8.1-10.3]

0.07

Unemployed

Retired

EP

Employed

0.5

Marital Status – no. (%) Single

AC C

1

Married/de facto

LTPA (min/week)

0.9

Blood biometric measures 25(OH)D (ng/mL)* Calcium (mg/dL)

2

ACCEPTED MANUSCRIPT PTH (ng/L)†

26.45 [6.4-100.6]

42.4 [15-124]

0.02

0.8 [0.5-1.9]

0.7 [0.4-1.8]

0.1

25.4 [15.21-37.2]

27.7 [12.3-33]

0.03

Systolic Blood Pressure (mmHg)

116 [82-150]

115 [93.3-143.3]

0.7

Diastolic Blood Pressure (mmHg)

73.3 [55-86]

70 [55-85]

0.07

23 (69.7)

0.44

17 [14-38]

<0.0001

21 (63.6)

0.002

5.2 [1.25-27]

0.3

11 (33.3)

0.5

Creatinine (mg/dL)

≥ 1 coexisting illness – no. (%)‡

45 (67.2)

BDI-II score

7 [1-13]

Psychotropic drugs – no. (%)

20 (29.8)

Distance from injury (years)

6.3 [1.1-48]

Level of lesion – no. (%) 28 (41.8)

Thoracic-lumbar spine Lesion completeness Complete motor lesion (ASIA A-B) Incomplete motor lesion (ASIA C-D)

Functional independence (SCIM score) Pain intensity (NRS score) Season of evaluation

EP

Autumn-Winter – no. (%)

39 (58.2)

22 (66.7)

37 (55.2)

23 (69.7)

30 (44.8)

10 (30.3)

15 (22.4)

3 (9.1)

0.1

57 [21-98]

39 [13-89]

0.001

5 [1-11]

6 [1-10]

0.2

42 (62.7)

11 (33.3)

0.01

25 (37.3)

22 (66.7)

TE D

Able to walk – no. (%)

M AN U

Cervical spine

0.2

AC C

Spring-Summer – no. (%)

SC

BMI (kg/m2)

RI PT

Clinical and injury-related variables

2

Data were expressed as median [min-max] for continuous variables and as percentages when

3

categorical. BDI-II, Beck Depression Inventory-II; LTPA, Leisure Time Physical Activity;

4

25(OH)D, 25-hydroxyvitamin D; PTH, Parathyroid hormone; BMI, Body Mass Index; ASIA,

5

American Spinal Injury Association; SCIM, Spinal Cord Independence Measure; NRS, Numeric

6

Rating Scale.

7

*To convert the values for 25(OH)D to nmol/L, multiply by 2.5.

8

†

Available only for 68 patients.

3

ACCEPTED MANUSCRIPT Coexisting illnesses included: heart diseases, chronic kidney disease, diabetes, hypertension, and

TE D

M AN U

SC

RI PT

dyslipidemia.

EP

10

‡

AC C

9

1

ACCEPTED MANUSCRIPT 1

Table 2. Univariate associations of putative predictors of depression status with the Beck

2

Depression Inventory-II (BDI-II) score BDI-II score

β (95% CI)

p value

RI PT

Putative predictors

-2.67 (-3.9,-1.44)

Age (years)

0.1 (-0.08, 0.30)

Education level

-0.4 (-1.36, 0.55)

0.40

Employment status

0.27 (-0.60, 1.14)

0.53

M AN U

SC

Gender

Marital status

<0.0001 0.28

0.11 (-0.77, 1.01)

0.80

-0.05 (-0.08, -0.03)

<0.0001

-0.98 (-1.14, -0.82)

<0.0001

-0.11 (-0.05, 0.28)

0.19

0.74 (0.19, 1.29)

0.008

0.7 (-0.57, 1.98)

0.28

-0.21 (-0.38, -0.03)

0.02

-0.39 (-1.62, 0.83)

0.52

-0.01 (-1.24, 1.2)

0.97

Walking ability

-0.45 (-2.01, 1.10)

0.56

SCIM score

-0.30 (-0.46, -0.15)

0.0002

Pain intensity (NRS score)

0.047 (-0.33, 0.43)

0.8

Seasonality

1.27 (0.10, 2.44)

0.03

Intake of psychotropic drugs

1.50 (0.32, 2.70)

0.013

LTPA (min/week) 25(OH)D (ng/mL)

BMI (kg/m2) Comorbidity status

Level of the lesion

EP

Distance from injury (years)

TE D

PTH (ng/L)

AC C

Completeness of the lesion

3

Univariate regressions were performed on square root-transformed values. Abbreviations: β-

4

coefficient, the standardized coefficient of regression; CI, confidence interval; LTPA, Leisure Time

2

ACCEPTED MANUSCRIPT Physical Activity; 25(OH)D, 25-hydroxyvitamin D; PTH, Parathyroid hormone; BMI, Body Mass

6

Index; SCIM, Spinal Cord Independence Measure; NRS, Numeric Rating Scale.

7

Categorical variables were expressed as Gender: 1 = female gender and 2 = male gender; Education

8

level: 1 = primary school, 2 = high school and 3 = university; Employment status: 1 = student, 2 =

9

unemployed, 3 = employed and 4 = retired; Marital status 1 = single, 2 = married/de facto, 3 =

10

divorced/separated and 4 = widowed; Comorbidity status: 1 = no coexisting illnesses and 2 = at

11

least one coexisting illness (heart diseases, chronic kidney disease, diabetes, hypertension, and

12

dyslipidemia); Level of the lesion: 1 = thoracic-lumbar spine lesion (paraplegia) and 2 cervical

13

spine lesion (tetraplegia); Completeness of the lesion: 1 = incomplete motor lesion (ASIA scale C-

14

D) and 2 = complete motor lesion (ASIA scale A-B); Walking ability: 1 = unable to walk and 2 =

15

able to walk; Seasonality: 1 = autumn/winter and 2 = spring/summer; intake of psychotropic drugs:

16

1 = no and 2 = yes.

AC C

EP

TE D

M AN U

SC

RI PT

5

1

ACCEPTED MANUSCRIPT 1

Table 3. Odds ratios (95% confidence intervals) for the association of 25-hydroxyvitamin D levels

2

with depression (Beck Depression Inventory-II score ≥14)

Odds ratio (95% CI) Model 1*

Model 2†

Model 3‡

0.76 (0.66, 0.85)

0.78 (0.67, 0.88)

0.76 (0.66, 0.88)

0.78 (0.66, 0.90)

p< 0.0001

p = 0.0002

p = 0.0005

p = 0.001

RI PT

Unadjusted

25(OH)D (ng/mL)

*Adjusted for gender, body mass index (Kg/m2) and seasonality.

5

†

6

medications.

7

‡

8

Cord Independence Measure.

SC

3 4

M AN U

Adjusted for all variables in Model 1 plus distance from injury (years) and intake of psychotropic

AC C

EP

TE D

Adjusted for all variables in Model 2 plus Leisure Time Physical Activity (min/week) and Spinal

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT