Loa

Ma

b

a

ARRA

KVCPBOCASBDAEI

1

rraovchdhn[

B

H

0d

Journal of Steroid Biochemistry & Molecular Biology 121 (2010) 467–470

Contents lists available at ScienceDirect

Journal of Steroid Biochemistry and Molecular Biology

journa l homepage: www.e lsev ier .com/ locate / j sbmb

ow serum levels of 25-hydroxyvitamin D (25-OHD) among psychiatricut-patients in Sweden: Relations with season, age, ethnic originnd psychiatric diagnosis�

ats B. Humblea,∗, Sven Gustafssonb, Susanne Bejerota

Department of Clinical Neuroscience, Division of Psychiatry, St. Göran, Karolinska Institutet, Stockholm, SwedenDepartment of Laboratory Medicine, Section for Clinical Chemistry, Karolinska Institutet, Stockholm, Sweden

r t i c l e i n f o

rticle history:eceived 10 November 2009eceived in revised form 25 January 2010ccepted 1 March 2010

eywords:itamin Dalcidiolarathyroid hormonelood levelsut-patients

a b s t r a c t

In a chart review at a psychiatric out-patient department, latitude 59.3◦N, a sample of patients withtests of serum 25-hydroxy-vitamin D (25-OHD) and plasma intact parathyroid hormone (iPTH) wascollected, together with demographic data and psychiatric diagnoses. During 19 months, 117 patientswere included. Their median 25-OHD was 45 nmol/l; considerably lower than published reports onSwedish healthy populations. Only 14.5% had recommended levels (over 75). In 56.4%, 25-OHD was under50 nmol/l, which is related to several unfavourable health outcomes. Seasonal variation of 25-OHD wasblunted. Patients with ADHD had unexpectedly low iPTH levels. Middle East, South-East Asian or Africanethnic origin, being a young male and having a diagnosis of autism spectrum disorder or schizophreniapredicted low 25-OHD levels. Hence, the diagnoses that have been hypothetically linked to developmen-tal (prenatal) vitamin D deficiency, schizophrenia and autism, had the lowest 25-OHD levels in this adult

hart reviewutismchizophreniaipolar disorderepressive disorderDHD

sample, supporting the notion that vitamin D deficiency may not only be a predisposing developmentalfactor but also relate to the adult patients’ psychiatric state. This is further supported by the consider-able psychiatric improvement that coincided with vitamin D treatment in some of the patients whosedeficiency was treated.

© 2010 Elsevier Ltd. All rights reserved.

thnic groupsmmigrants

. Introduction

In the field of psychiatry, interest in vitamin D is of relativelyecent origin. The discovery that the brain possesses vitamin Deceptors was decisive and first inspired the suggestion that moodnd depressive disorders may be related to vitamin D deficiencyr insufficiency [1,2]. Also, hypotheses suggesting that prenatalitamin D deficiency impairs fetal neural development, therebyontributing to adult schizophrenia [3] or childhood autism [4]ave been put forward. In the case of schizophrenia, ample evi-

ence from epidemiology and preclinical research supports thisypothesis [5,6]. A small study of in-patients with schizophre-ia showed decreased levels of 25-hydroxy-vitamin D (25-OHD)7]. In an epidemiological case-control study, however, individuals

� Special issue selected article from the 14th Vitamin D Workshop held at Brugge,elgium on October 4–8, 2009.∗ Corresponding author at: Psychiatric Services for the Elderly, Uppsala Universityospital, SE-750 17, Uppsala, Sweden. Tel.: +46 18 611 23 11.

E-mail address: mats.humble@gmail.com (M.B. Humble).

960-0760/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.oi:10.1016/j.jsbmb.2010.03.013

with psychosis did not differ from controls [8]. Concerning depres-sive disorders, two population based studies support a relationbetween lower 25-OHD levels and depressed mood [9,10]. Fourrandomized trials, none of them focussing on Major depression,nevertheless support the possible causality of vitamin D deficiency[11–14]. The few previous studies on 25-OHD levels in psychiatricpatients have presented data from in-patients [7,15]. In psychiatricresearch, hyperparathyroidism has been related to depressive dis-order [16,10], however, many cases of hyperparathyroidism aresecondary to vitamin D deficiency [17], and in most psychiatricstudies, this has been disregarded. According to Jorde et al. [9]depressed mood is more related to low 25-OHD than to elevatedintact parathyroid hormone (iPTH).

In view of this background and in order to improve the qual-ity of care for psychiatric out-patients, we considered it relevantto include measurements of 25-OHD and iPTH in our standard

procedure for evaluating the physical health of our patients. Theaim of the present chart review was to describe results on thesemeasurements from a sizeable group of patients with various psy-chiatric diagnoses, and search for possible predictors of vitamin Ddeficiency requiring treatment.

468 M.B. Humble et al. / Journal of Steroid Biochemistry & Molecular Biology 121 (2010) 467–470

Table 1Demographic data, psychiatric diagnoses, serum 25-OH-vitamin D levels and plasma intact PTH levels of 117 psychiatric out-patients from Värmdö municipality, Sweden,latitude 59.3◦N.

Diagnostic group N Age years(means)

Sex(% females)

Origin(% Southern)

25-OHD nmol/lmedian (25th and75th percentiles)

iPTH pg/lmedian (25th and75th percentiles)

Autism 10 36.5 40 0 31.5 (23, 39) 56 (39, 72)Schizophrenia 20 47.4 60 20 35 (23.5, 52.5) 52 (36, 63)Bipolar 22 53.4 55 9 48 (33, 57) 53 (46, 63)Depressive 36 41.5 53 17 47.5 (33, 72) 42 (37, 66)Anxiety 13 40.3 69 0 62 (44, 81) 47 (39, 58)ADHD 8 34.2 50 0 45 (31, 67.5) 26 (23, 41)Substance 3 42.0 33 0 40 (37, 43) 43 (27, 48)Others 5 42.4 60 0 61 (45, 80) 33 (22, 53)

F of iPTu

2

2

ciaib1

2

iiwwardre

2

rfapbd[s(Dah(Dtacor

All groups 117 43.7 54.7

or 25-OHD, a level between 75 and 250 nmol/l is recommended. Reference valuesnder 43 pg/l.

. Materials and methods

.1. Setting

The first author worked as clinical psychiatrist in the psy-hiatric out-patient department of Gustavsberg, latitude 59.3◦N,n Stockholm County. This department is the main psychi-tric facility serving the catchment area of Värmdö munic-pality with 37,376 inhabitants (2008). Data were collectedetween March 2008 and September 2009, a time span of9 months.

.2. Patients

The patients were unselected consecutive cases, with a clinicalndication for blood sampling not related to the study (approx-mately 75% of all patients seen). For instance, some patients

ith regular controls related to lithium or clozapine therapyere included. In other cases, the reason for blood sampling was

ssessment of possible organic causes of depression (e.g. hypothy-oidism or cobalamin deficiency) or screening for metabolicisturbance in patients on antipsychotic treatment. Patientsegularly taking vitamin D supplements (≥10 �g/day) werexcluded.

.3. Demographic data and psychiatric diagnoses

All patients were seen and assessed by the first author in hisoutine clinical praxis. Most of the patients had been in treatmentor several years at the unit, and previous medical records werevailable in order to care for the patients. Demographic data andsychiatric diagnoses were assessed by the first author on thisasis. In cases of comorbidity, the clinically most relevant mainiagnose was used. Diagnoses were grouped according to ICD-1018] as follows: Autism spectrum disorders, including Asperger’syndrome (F94), Schizophrenia, including Schizoaffective disordersF20, F25), Bipolar disorders (F31), Depressive disorders, includingysthymia (F32, F33, F34.1), Anxiety disorders, including Person-lity disorders, if not depressed (F40–F42, F60.3), Attention deficityperactivity disorders (ADHD) (F90), Substance use disordersF10–F19), and Others, including Organic mental disorders andelusional disorders (F05–F09, F22). Ethnic origin was based on

he patients’ personal assertion and mother tongue. Ethnicity ofdopted children was based on biological parents. Ethnicities wereategorized as Northern (Scandinavian, Finnish or Slavonic origin)r Southern (Middle East, South-East Asian, African or Mediter-anean origin).

10.3 45 (31, 60) 47 (37, 62)

H are 20–65 pg/l, however, in the absence of vitamin D deficiency iPTH should be

2.4. Blood sampling

Blood samples (most of them fasting before 10 a.m.) were takenat a primary care sampling unit and sent to the laboratory as partof usual clinical routine, together with other samples, and analyzedconsecutively. Results were communicated to the psychiatric unitconsecutively.

2.5. Laboratory analyses

Serum 25-OHD was measured by radioimmunoassay, DiaSorin,Stillwater, USA. Plasma levels of iPTH were analyzed by an electro-chemiluminescent immunoassay, Roche Diagnostics, Germany.

2.6. Statistical analysis

When distributions were non-normal, values are reported asmedians with 25th and 75th percentiles in parentheses, and non-parametric tests (Mann–Whitney U-test, Kruskal–Wallis ANOVAand Spearman Rank Order Correlation) were used. A general regres-sion model with data log transformed was used for multiplecorrelations. Statistica v. 7.1 from StatSoft Inc. was used. Diagnosticgroups with 5 or less cases were not included in the comparisonsbetween diagnoses. Because of the presumed seasonal variation of25-OHD, a seasonal adjustment was done by multiplying individ-ual values with the quotient between the mean of the year and themean of the season. Probabilities < 5% were assumed as significant.

2.7. Ethics

This was a clinical quality assurance project. Neither randomiza-tion, nor placebo treatment was implemented. Blood samples weretaken on clinical grounds, and the charts were reviewed in orderto improve clinical care. Accordingly, no research ethics commit-tee was involved and written informed consent was not obtained.However, all patients were orally informed about the vitamin Dsampling and about their results, and treated with vitamin D whenappropriate.

3. Results

Serum levels of 25-OHD was tested in 121 and iPTH in 97patients. Due to holidays and other factors, the samples were

unevenly distributed over the year with few samples taken in Jan-uary, July and November. Four users of vitamin-D supplements(≥10 �g/day) with significantly higher 25-OHD (median 93.5 (79.5,100.5) compared to 45 (31, 60) nmol/l, MW-U = 27.5; p = 0.003)were excluded, leaving 117 patients for this analysis (93 with iPTH

M.B. Humble et al. / Journal of Steroid Biochemistry & Molecular Biology 121 (2010) 467–470 469

Fig. 1. Serum 25-OH-vitamin D levels among 117 psychiatric out-patients differedbHy

sa

taoTmtpelOaht(m(wMa6papci2nhO4imocg

(ma

Fig. 2. Serum levels of 25-OH-vitamin D among psychiatric out-patients in different

min D deficiency or secondary hyperparathyroidism was related to

etween sexes (MW-U = 1226.5; p = 0.010) and between age categories (KW-(2;117) = 7.30; p = 0.026), however, the age difference was entirely driven by theounger males (males only: KW-H(2;53) = 8.18; p = 0.017, females N.S.).

ample). Demographic data and median levels of 25-OHD and iPTHre summarized in Table 1.

Both 25-OHD and iPTH were non-normally distributed. The dis-ribution of 25-OHD was: ≤25, 15.4%; 25–50: 41.0%; 50–75: 29.1%;nd >75: 14.5%, with the highest value, 115 nmol/l, attained byne Scandinavian male with personality disorder in September.hree patients had 25-OHD under 12.5 nmol/l: two Scandinavianales with post-stroke depression and schizophrenia, respec-

ively, and one Mediterranean female with schizophrenia. Of 93atients with iPTH samples, 19% had pathologically elevated lev-ls (>65 pg/l according to reference values). Only 41% had optimalevels (<43 pg/l). There was a negative correlation between 25-HD and iPTH (R = −0.32; p = 0.002). iPTH correlated linearly withge (R = 0.21; p = 0.04) whereas 25-OHD did not. On the otherand, when age was categorized, a significant difference betweenhose under 34 years and the two older categories was foundKW-H(2;117) = 7.30; p = 0.026), which was entirely driven by the

ales (Fig. 1). Females had higher 25-OHD than males; 52.534.5, 63) vs. 39 (25, 52) nmol/l (MW-U(1;117) = 1226.5; p = 0.010),hile iPTH did not differ between sexes. Southern (Middle East,editerranean, South-East Asian or African) ethnic origin wasstrong predictor of low 25-OHD; 26.5 (14.5, 34) vs. 47 (32,

1) nmol/l among those with Northern origin (MW-U(1;117) = 224;= 0.00027), but not of iPTH. In a general regression, there wasn interaction between sex and ethnic origin (Wald �2(1) = 8.04,= 0.005), females of Southern origin having more reduced 25-OHDompared to the Northern group than the Southern males. Interest-ngly, among the subgroups, those of Finnish origin had the highest5-OHD levels: 71 (53, 79) nmol/l, significantly higher than Scandi-avians in pair-wise comparison; while those from the Middle Eastad the lowest levels, 25 (13, 34) nmol/l. Seasonal variation of 25-HD was small: winter 40.5 (32, 56); spring 40 (28, 53); summer6 (25, 67); and autumn 58 (42, 73). Autumn levels were higher

n pair-wise comparisons, but this failed to reach significance in aultiple comparison. With seasonal adjustment of 25-OHD values,

nly minor numerical changes appeared, and the statistical signifi-ance of all other tests and comparisons were unchanged (data notiven).

The psychiatric diagnostic groups differed in 25-OHD levelsKW-H(5;109) = 15.86; p = 0.0073). This was mainly related to the

arkedly low levels of patients with autism spectrum disordernd schizophrenia see Fig. 2. Also the iPTH levels differed between

diagnostic groups (KW-H(5;109) = 15.86; p = 0.0073). For delineation of diagnosticgroups, see Section 2.3. In post hoc comparisons, the difference between Autismand Anxiety/Personality Disorder was significant (p = 0.014). Schizophrenia vs. Anx-iety/PersD attained p = 0.053.

diagnoses (KW-H(5;87) = 11.97; p = 0.035), see Table 1. Accordingto post hoc comparisons, this was due to the relatively low levelsof patients with ADHD. Treatment with neither lithium (N = 17) norclozapine (N = 6) had any effect on 25-OHD or iPTH.

4. Discussion and conclusion

The 25-OHD levels in our sample indicate a high prevalence ofvitamin D deficiency/insufficiency among Swedish psychiatric out-patients. In contrast, previously published samples from healthySwedish populations [19–23] (mainly based on postmenopausalwomen) have found 25-OHD means between 69 and 99 nmol/l, i.e.unexpectedly high in international comparisons [20]. The consider-ably lower levels in the present study may be related to our samplebeing younger and including males, but may also be associated withour patients’ diagnoses. The usual seasonal pattern was blunted inthis sample, indicating life styles with reduced sun exposure. Also,the expected decrease with age was absent; on the contrary, sig-nificantly lower levels were seen in the younger male patients. Inline with other European studies, immigrants (particularly females)with darker skin or diminished sun exposure (e.g. covered clothing)had markedly lower 25-OHD levels.

Interestingly, the two diagnoses that have been hypothesized aspossibly related to developmental (prenatal) vitamin D deficiency,schizophrenia [3] and autism [4], had the lowest 25-OHD levelsin this adult sample. Indeed, this may be related to diet or indoorlife-style, but unknown biological factors may also contribute. Thisfinding supports the notion that vitamin D deficiency may not onlybe a predisposing developmental factor, but also bear on thesepatients’ present psychiatric state [24]. This is further supported bythe considerable improvement of psychosis or depression, whichcoincided in several patients with effective treatment of theirvitamin D deficiency, even if other treatments were unchanged.Vitamin D was given as cholecalciferol 1600–4000 IU daily or ergo-calciferol 35,000–70,000 IU once weekly. In the absence of a controlgroup, these improvements could have been due to the naturalcourse or a placebo effect, but the possibility remains that vita-

their present psychiatric symptomatology.The relatively low iPTH levels among patients with ADHD are

of interest, since they were not related to higher 25-OHD levels.We were unable to find any previous study on PTH in ADHD. How-

4 emistr

efhboP

asfislbcn

bc[pcmhsioldoaa[rst

tsgprcd

R

[

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

(2005) 1155–1163.[34] E. Cantor-Graae, J.P. Selten, Schizophrenia and migration: a meta-analysis and

70 M.B. Humble et al. / Journal of Steroid Bioch

ver, blunted PTH response to vitamin D deficiency is an establishedeature of magnesium deficiency [25], and magnesium deficiencyas repeatedly been described in ADHD [26,27]. Our finding isased on only 8 individuals with ADHD and may be spurious. Inur view, however, it indicates that further investigations of theTH-magnesium relations in ADHD are warranted.

There are several limitations of this study. There may have beenselection bias. Patients in seemingly perfect somatic health were

ometimes excluded; however, such patients were rare among therst author’s clientele. The diagnoses were not ascertained withtructured diagnostic rating scales. Arguably, diagnoses based on aengthy clinical treatment history may be at least as valid as thoseased on one cross-sectional rating. Finally, there was no healthyontrol group of similar age and sex distribution. The different diag-ostic groups may to some extent serve as substitute for this.

The clinical relevance of hypovitaminosis D in our sample maye underscored by the increased somatic morbidity, especiallyardiovascular disorders and diabetes, among psychiatric patients28]. To what extent vitamin D deficiency contributes to this, is atresent unknown. Furthermore, two common treatments in psy-hiatric clinical praxis have been reported to negatively affect boneineral density (BMD): Several antipsychotic agents may lead to

yperprolactinemia, which in the long-term may cause osteoporo-is [29,30]. Secondly, long-term use of selective serotonin reuptakenhibitors has been found to decrease BMD and increase the riskf fractures [31,32]. Since vitamin D deficiency may further chal-enge the psychiatric patients’ bone health, this remediable factoreserves increased interest. Our findings also call for focussedbservance of vitamin D status in immigrant patients. In Sweden,s in other sun deprived countries, increased rates of psychosesnd autism among dark skinned immigrants have been reported33–35,24]. While controlled studies of the effect of vitamin Deplenishment on the psychiatric symptoms are still wanting, theomatic risks of continued deficiency should alert clinicians to meethe specific needs of this population.

In conclusion, the previous optimistic view on vitamin D sta-us in Sweden, mainly derived from healthy elderly females, is notupported by this younger sample including both sexes and immi-rants. Considering the high proportion of low 25-OHD levels in theresent sample, a substantial number of psychiatric patients are atisk for unfavourable health outcomes related to vitamin D defi-iency, and may benefit from detection and treatment of vitamin Deficiency.

eferences

[1] W.E. Stumpf, T.H. Privette, Light, vitamin D and psychiatry. Role of 1,25 dihy-droxyvitamin D3 (soltriol) in etiology and therapy of seasonal affective disorderand other mental processes, Psychopharmacology (Berl.) 97 (1989) 285–294.

[2] M. Berk, K.M. Sanders, J.A. Pasco, F.N. Jacka, L.J. Williams, A.L. Hayles, S. Dodd,Vitamin D deficiency may play a role in depression, Med. Hypotheses 69 (2007)1316–1319.

[3] J. McGrath, Hypothesis: is low prenatal vitamin D a risk-modifying factor forschizophrenia? Schizophr. Res. 40 (1999) 173–177.

[4] J.J. Cannell, Autism and vitamin D, Med. Hypotheses 70 (2008) 750–759.[5] D.K. Kinney, P. Teixeira, D. Hsu, S.C. Napoleon, D.J. Crowley, A. Miller, W.

Hyman, E. Huang, Relation of schizophrenia prevalence to latitude, climate,fish consumption, infant mortality, and skin color: a role for prenatal vitaminD deficiency and infections? Schizophr. Bull. 35 (2009) 582–595.

[6] D.W. Eyles, F. Feron, X. Cui, J.P. Kesby, L.H. Harms, P. Ko, J.J. McGrath, T.H. Burne,Developmental vitamin D deficiency causes abnormal brain development, Psy-choneuroendocrinology 34 (Suppl. 1) (2009) S247–S257.

[7] B. Schneider, B. Weber, A. Frensch, J. Stein, J. Fritz, Vitamin D in schizophrenia,

major depression and alcoholism, J. Neural Transm. 107 (2000) 839–842.

[8] J.J. McGrath, M.G. Kimlin, S. Saha, D.W. Eyles, A.V. Parisi, Vitamin D insufficiencyin south-east Queensland, Med. J. Aust. 174 (2001) 150–151.

[9] R. Jorde, K. Waterloo, F. Saleh, E. Haug, J. Svartberg, Neuropsychological functionin relation to serum parathyroid hormone and serum 25-hydroxyvitamin Dlevels. The Tromsø study, J. Neurol. 253 (2006) 464–470.

[

y & Molecular Biology 121 (2010) 467–470

10] W.J. Hoogendijk, P. Lips, M.G. Dik, D.J. Deeg, A.T. Beekman, B.W. Penninx,Depression is associated with decreased 25-hydroxyvitamin D and increasedparathyroid hormone levels in older adults, Arch. Gen. Psychiatry 65 (2008)508–512.

11] A.T. Lansdowne, S.C. Provost, Vitamin D3 enhances mood in healthy subjectsduring winter, Psychopharmacology (Berl.) 135 (1998) 319–323.

12] F.M. Gloth III, W. Alam, B. Hollis, Vitamin D vs broad spectrum phototherapyin the treatment of seasonal affective disorder, J. Nutr. Health Aging 3 (1999)5–7.

13] R. Vieth, S. Kimball, A. Hu, P.G. Walfish, Randomized comparison of the effectsof the vitamin D3 adequate intake versus 100 mcg (4000 IU) per day on bio-chemical responses and the wellbeing of patients, Nutr. J. 3 (2004) 8.

14] R. Jorde, M. Sneve, Y. Figenschau, J. Svartberg, K. Waterloo, Effects of vita-min D supplementation on symptoms of depression in overweight and obesesubjects: randomized double blind trial, J. Intern. Med. 264 (2008) 599–609.

15] M. Berk, F.N. Jacka, L.J. Williams, F. Ng, S. Dodd, J.A. Pasco, Is this D vitaminto worry about? Vitamin D insufficiency in an inpatient sample, Aust. N. Z. J.Psychiatry 42 (2008) 874–878.

16] C. Joborn, J. Hetta, J. Rastad, H. Ågren, G. Åkerström, S. Ljunghall, Psychiatricsymptoms and cerebrospinal fluid monoamine metabolites in primary hyper-parathyroidism, Biol. Psychiatry 23 (1988) 149–158.

17] W.D. Fraser, Hyperparathyroidism, Lancet 374 (2009) 145–158.18] The ICD-10 Classification of Mental and Behavioural Disorders, Clinical Descrip-

tions and Diagnostic Guidelines, World Health Organization, Geneva, 1992.19] K. Landin-Wilhelmsen, L. Wilhelmsen, J. Wilske, G. Lappas, T. Rosén, G. Lind-

stedt, P.A. Lundberg, B.A. Bengtsson, Sunlight increases serum 25(OH) vitaminD concentration whereas 1,25(OH)2D3 is unaffected. Results from a generalpopulation study in Göteborg, Sweden (The WHO MONICA Project), Eur. J. Clin.Nutr. 49 (1995) 400–407.

20] P. Lips, T. Duong, A. Oleksik, D. Black, S. Cummings, D. Cox, T. Nickelsen, A globalstudy of vitamin D status and parathyroid function in postmenopausal womenwith osteoporosis: baseline data from the multiple outcomes of raloxifeneevaluation clinical trial, J. Clin. Endocrinol. Metab. 86 (2001) 1212–1221.

21] A. Burgaz, A. Åkesson, A. Öster, K. Michaëlsson, A. Wolk, Associations of diet,supplement use, and ultraviolet B radiation exposure with vitamin D status inSwedish women during winter, Am. J. Clin. Nutr. 86 (2007) 1399–1404.

22] H. Salminen, M. Sääf, H. Ringertz, L.E. Strender, The role of IGF-I and IGFBP-1status and secondary hyperparathyroidism in relation to osteoporosis in elderlySwedish women, Osteoporos. Int. 19 (2008) 201–209.

23] A. Burgaz, A. Åkesson, K. Michaëlsson, A. Wolk, 25-Hydroxyvitamin D accumu-lation during summer in elderly women at latitude 60 degrees N, J. Intern. Med.266 (2009) 476–483.

24] M.J. Dealberto, Why are immigrants at increased risk for psychosis? VitaminD insufficiency, epigenetic mechanisms, or both? Med. Hypotheses 68 (2007)259–267.

25] R.K. Rude, S.B. Oldham, C.F. Sharp Jr., F.R. Singer, Parathyroid hormone secretionin magnesium deficiency, J. Clin. Endocrinol. Metab. 47 (1978) 800–806.

26] T. Kozielec, B. Starobrat-Hermelin, Assessment of magnesium levels in childrenwith attention deficit hyperactivity disorder (ADHD), Magnes. Res. 10 (1997)143–148.

27] M. Mousain-Bosc, M. Roche, A. Polge, D. Pradal-Prat, J. Rapin, J.P. Bali,Improvement of neurobehavioral disorders in children supplemented withmagnesium-vitamin B6. I. Attention deficit hyperactivity disorders, Magnes.Res. 19 (2006) 46–52.

28] W.W. Fleischhacker, M. Cetkovich-Bakmas, M. De Hert, C.H. Hennekens, M.Lambert, S. Leucht, M. Maj, R.S. McIntyre, D. Naber, J.W. Newcomer, M. Olfson,U. Ösby, N. Sartorius, J.A. Lieberman, Comorbid somatic illnesses in patientswith severe mental disorders: clinical, policy, and research challenges, J. Clin.Psychiatry 69 (2008) 514–519.

29] A.M. Meaney, V. O’Keane, Bone mineral density changes over a year in youngfemales with schizophrenia: relationship to medication and endocrine vari-ables, Schizophr. Res. 93 (2007) 136–143.

30] T. Kishimoto, K. Watanabe, N. Shimada, K. Makita, G. Yagi, H. Kashima,Antipsychotic-induced hyperprolactinemia inhibits the hypothalamo-pituitary-gonadal axis and reduces bone mineral density in male patients withschizophrenia, J. Clin. Psychiatry 69 (2008) 385–391.

31] K. Saag, Mend the mind, but mind the bones!: balancing benefits and potentialskeletal risks of serotonin reuptake inhibitors, Arch. Intern. Med. 167 (2007)1231–1232.

32] G. Ziere, J.P. Dieleman, T.J. van der Cammen, A. Hofman, H.A. Pols, B.H. Stricker,Selective serotonin reuptake inhibiting antidepressants are associated with anincreased risk of nonvertebral fractures, J. Clin. Psychopharmacol. 28 (2008)411–417.

33] E. Cantor-Graae, K. Zolkowska, T.F. McNeil, Increased risk of psychotic disorderamong immigrants in Malmö: a 3-year first-contact study, Psychol. Med. 35

review, Am. J. Psychiatry 162 (2005) 12–24.35] M. Barnevik-Olsson, C. Gillberg, E. Fernell, Prevalence of autism in children born

to Somali parents living in Sweden: a brief report, Dev. Med. Child Neurol. 50(2008) 598–601.