206

JOURNAL OF WOMEN’S HEALTHVolume 16, Number 2, 2007© Mary Ann Liebert, Inc.DOI: 10.1089/jwh.2006.0089

Prevalence of Vitamin D Insufficiency and Clinical Associations among Veiled East African

Women in Washington State

SUSAN D.L. REED, M.D., M.P.H.,1,2,3 MARY B. LAYA, M.D., M.P.H.,4JENNIFER MELVILLE, M.D., M.P.H.,3,5 SIRAD Y. ISMAIL, M.S.,6

CAROLINE M. MITCHELL, M.D.,3 and DIANA R. ACKERMAN, M.D.3

ABSTRACT

Aims: Cultural and biological factors place immigrant women from equatorial Africa at in-creased risk of vitamin D insufficiency. This could in part explain the high prevalence of fa-tigue, musculoskeletal complaints, and depressive symptoms in this population.

Methods: In a cross-sectional study of East African immigrant women in Washington State,25-hydroxyvitamin D (25(OH)D) serum concentrations and multiple measures of physical andpsychological symptoms were assessed. Mean serum 25(OH)D serum concentrations and chi-square were used to assess differences between groups. Multiple logistic regression was usedto explore differences in the symptoms of subjects with varying degrees of vitamin D defi-ciency/insufficiency.

Results: Of the 75 women interviewed and who completed surveys, 25(OH)D serum sam-ples were available in 71 subjects. All were found to have low 25(OH)D; 9 (12.3%) had �8ng/mL, 29 (40.9%) had 8.1–15 ng/mL, and 33 (44.9%) had 15.1–30 ng/mL. After controlling forage, women with 25(OH)D � 15 ng/mL were 66% less likely to drink milk than women with25(OH)D � 15 ng/mL (adjusted odds ratio [aOR] � 0.33, 95% confidence interval [CI]0.11–0.99). Musculoskeletal complaints, depressive symptoms, and fatigue did not correlatewith the severity of 25(OH)D insufficiency.

Conclusions: Vitamin D insufficiency, as reflected by low 25(OH)D serum concentrations,was a universal finding in this group of women, suggesting the need for widespread educa-tion and intervention in this and other immigrant groups at northern latitudes.

1Public Health Services Division, Fred Hutchinson Cancer Research Center, Seattle, Washington.2Department of Epidemiology, University of Washington, Seattle, Washington.3Department of Obstetrics and Gynecology, University of Washington Medical Center, Harborview Medical Cen-

ter, Seattle, Washington.4Department of Internal Medicine, University of Washington Medical Center, Seattle, Washington.5Department of Psychiatry and Behavioral Sciences, University of Washington Medical Center, Seattle, Washing-

ton.6Refugee Women’s Alliance, Seattle, Washington.This work was supported by the University of Washington Center of Excellence in Women’s Health and a Com-

munity Service Gift from Johnson and Johnson.

INTRODUCTION

VITAMIN D IS OF CRITICAL IMPORT in maintain-ing calcium homeostasis. Symptomatic

vitamin D deficiency is associated with muscu-loskeletal problems, such as bone pain, muscleweakness, and osteomalacia-related bone fragility,1,2

as well as mood disturbances and impaired neu-ropsychiatric function.3,4

Sunlight on the skin (specifically solar UV-B ir-radiation) induces the formation of pre-vitaminD3 from 7-dehydrocholesterol in the epidermis, acritical step in the synthesis of metabolically ac-tive forms of the vitamin. In the absence of ade-quate sunlight, ingestion of fatty fish or vitaminD-fortified food, such as milk, is necessary to im-prove vitamin D intake. Persons at risk of devel-oping vitamin D deficiency include persons liv-ing at northern latitudes, the homebound orinstitutionalized, dark-skinned individuals, andthose who avoid direct sunlight exposure for cul-tural or health reasons.5–9

Studies of Asian and Middle Eastern immi-grants residing in Northern European countries(British Isles, Scandinavia, and Norway) have re-vealed varying degrees of vitamin D defi-ciency.5,10 These countries are located at latitudesabove 50° N, and some have diminished UV-B ir-radiation for many months of the year.11 Plot-nikoff and Quigley12 studied both immigrantsand nonimmigrants who had persistent symp-toms of nonspecific musculoskeletal complaintsin an outpatient clinic in Minneapolis, Minnesota(latitude 44° N). In this case series, 93% of patientsdemonstrated vitamin D insufficiency.

Somali families from Eastern Africa immi-grated to the Seattle area of Washington Stateduring the 1990s, fleeing civil war and related vi-olence that was widespread after the fall of thegovernment and withdrawal of foreign troops. Itis common for Somali women to come to ourcommunity clinics with musculoskeletal achesand pains, fatigue, and depressive symptoms.These symptoms are often considered physicalmanifestations of psychosocial stressors or psy-chiatric illness related to dislocation and pasttrauma. However, it is recognized that these wo-men are also at risk for vitamin D deficiency/in-sufficiency. Women from this predominantlyMuslim country wear long robes or dresses andhead coverings, called hijab or hejab, which alsoserve as protection from the equatorial sun. In theSeattle area (latitude 47–48°N) UV-B irradiation

is diminished for many months of the year, andin the absence of adequate dietary vitamin D, thistraditional form of dress may place Somali wo-men at increased risk for insufficient vitamin D.

We sought to determine the prevalence of vit-amin D deficiency/insufficiency among Somaliwomen living in our community and evaluate apossible association of low serum concentrationsof 25-hydroxyvitamin D (25(OH)D) with muscu-loskeletal symptoms, fatigue, pain, and depres-sion. We also explored whether 25(OH)D levelswere related to sun exposure or milk or vitaminconsumption. As part of our study protocol, weplanned to provide all deficient women with 2months of vitamin D supplementation and a let-ter with 25(OH)D results to be taken to theirhealthcare provider at completion of the study.

MATERIALS AND METHODS

The Institutional Review Board of the FredHutchinson Cancer Research Institute with reci-procity with the University of Washington ap-proved this cross-sectional study. This pilot project was a collaborative effort between theUniversity of Washington’s Center of Excellencein Women’s Health and a local community orga-nization, the Refugee Women’s Alliance (ReWA).Somali women were recruited by ReWA throughfliers and phone calls to women who had re-ceived services in the past year at this organiza-tion. ReWA provides services to approximately67% of Somali immigrant women in the Seattlearea, assisting in legal, immigration, housing, so-cial, and health services referrals. They provideon-site English literacy and nutrition classes. Ap-proximately 300 Somali women seek services peryear. When attempting to recruit women for thisproject, the original contact information given toReWA was no longer accurate in almost 50% atstudy recruitment. The majority had moved andwere difficult to trace (original phone numberwas that of a friend with whom they were livingonly temporarily). We were able to contact a con-venience sample of approximately 130 women,and 100 agreed to participate.

Women were seen at the ReWA facilities onfour weekends in March and April of 2005. Thesedates were chosen because of the known associ-ation of temporal deficiencies in vitamin D fol-lowing winter months.11 Seventy-five of the 100women (75%) who agreed to participate came for

VITAMIN D INSUFFICIENCY AMONG VEILED EAST AFRICAN WOMEN 207

the blood draw and administration of a four-pagesurvey that included questions about demo-graphics; joint, back, or bone pain; and muscleweakness, using a simple visual analog scale with anchors of “none” and “severe.” Womenwere also queried about time spent outdoors,prior fractures, and vitamin, calcium, and milkintake. The survey contained measures for de-pressive symptoms (Patient’s Health Question-naire [PHQ9])13 and posttraumatic stress disor-der (Primary Care Posttraumatic Stress Disorderscreen [PC-PTSD]).14 The survey was first pilotedamong Somali women from the community, andadjustments were made to increase readabilityand understanding. Surveys were forward andback-translated in Somali and were administeredin Somali by bilingual staff. Blood (3 mL) wasdrawn by trained phlebotomists and taken on iceto the University of Washington Research Labo-ratory. Specimens were spun immediately on ar-rival in the laboratory, and all analyses were runtogether.

All specimens were run using the DiaSorin(Stillwater, MN) 25(OH)D equilibrium radioim-munoassay (RIA). First, a rapid extraction of25(OH)D and other hydroxylated metabolitesfrom serum or plasma with acetonitrile was per-formed. Next, the specimen, antibody, and tracerwere incubated for 90 minutes at 20–25°C. Phaseseparation was accomplished after a 20-minuteincubation at 20–25°C with a second antibody-precipitating complex. An additional buffer(phosphate gelatin buffer containing 0.1% sodiumazide) was added after this incubation and priorto centrifugation to aid in reducing nonspecificbinding.

Serum concentrations of 25(OH)D (ng/mL)were grouped as mildly (15.1–30), moderately(8.1–15), or severely insufficient/deficient (�8.0).These concentrations are equivalent to 38–75,20–37, and �20 nmol/L, respectively. Contro-versy exists as to the appropriate definitions ofvitamin D insufficiency or deficiency. In pre-dominantly Caucasian populations, parathyroidhormone (PTH) normalization has been reportedto occur at 25(OH)D concentrations of 30ng/mL,15,16 bone mineral density (BMD) toplateau at approximately 36 ng/mL,17 and opti-mal fracture prevention to be achieved at ap-proximately 40 ng/mL.17 These norms have notbeen established for Somali immigrant women,although a lower threshold may be appropriatefor African American women.18 As vitamin D in-

sufficiency has been defined as 21–30 ng/mL, wechose 30 ng/mL as the lower bound of normal.Severe 25(OH)D in our research laboratory wasdefined as �8 ng/mL. Because of the small num-bers of Somali women in the study and the current unknown appropriate designations forvitamin D insufficiency/deficiency for this pop-ulation, we chose to establish fairly equal num-bers in the moderate and mildly insufficient/deficient categories and retain our research labo-ratory’s definition of severe deficiency. We al-lowed dichotomization of 25(OH)D at 15 ng/mLfor some analyses.19

All women were invited back to receive theirresults in sealed envelopes and attend a nutritionclass within 2 months after having their blooddrawn. The class emphasized the importance ofvitamin D supplementation. All women with vi-tamin D deficiency/insufficiency were given a 2-month supply of cholecalciferol, with doses ad-justed for the 25(OH)D result (1, 2, or 3 tablets of400 IU for mild, moderate, or severe defi-ciency/insufficiency, respectively). Investigatorswere present with bilingual interpreters to givethe results. Women were asked to take the writ-ten results of vitamin D screening to their primarycare providers along with recommendations fortherapy within the next month.

Mean serum concentrations of 25(OH)D weredetermined by age, years of schooling, employ-ment, health status, and clinical symptoms.Prevalence of vitamin D deficiency/insufficiencywas calculated. Demographic, muscle and bonehealth characteristics, depressive symptoms, fa-tigue, nutrition related to vitamin D consump-tion, and surrogates of sun exposure were evalu-ated among the three groups (mild, moderate,and severely deficient/insufficient) to assess apossible association with the severity of defi-ciency. Potential confounding by the presence ofPTSD was evaluated.

Multiple logistic regression, with key vari-ables decided a priori to be of import or to showsignificance on univariate analyses, comparedmildly deficient/insufficient to the moderateand severely deficient women, and chi-squareanalyses compared differences among the threegroups (mild, moderate, and severely defi-cient/insufficient) (p � 0.05). Confidence inter-vals (CI) with 95% limits were generated. Sta-tistical analyses were performed using StataSEsoftware version 8.2 for personal computer (Col-lege Station, TX).

REED ET AL.208

RESULTS

Characteristics of the women are shown inTable 1. Of the 75 women who were interviewedand completed surveys, vitamin D serum sam-ples were available in 71 women, and all werefound to have at least vitamin D insufficiency.Nine (12.3%) had serum concentrations �8ng/mL (severe deficiency), 29 (40.9%) had serumconcentrations �8 ng/mL but �15 ng/mL (mod-erate deficiency), and 33 (44.9%) had serum con-centrations �15 ng/mL but �30 ng/mL (mild deficiency/insufficiency). Overall mean serumconcentration was 14.4 ng/mL (95% CI 13.0–15.9).

The mean age of participants was 44.1, with arange of 18–75 years. Most women who were se-verely deficient were at the extremes of the agerange, 18–21 or �51. The majority of women hadnot completed high school (85.5%), and this wasparticularly true of women with severe 25(OH)Ddeficiency (100%). Less than 10% of the womenwere employed outside of the home. Most, 94%,considered themselves to be in good, very good,or excellent health, and self-proclaimed healthstatus did not correlate with serum 25(OH)D con-centration. None of the women were smokers,

and none reported thyroid dysfunction, or renalor hepatic disease. Four women had diabetes, butnone of the 4 had severe 25(OH)D deficiency.Mean serum concentrations did not vary by age, schooling, employment, or self-proclaimedhealth status.

Table 2 summarizes findings from an analysisof the association of clinical symptoms withseverity of 25(OH)D deficiency. The presence ofjoint, back, and bone pain, muscle weakness,falling, fatigue, or depression was not associatedwith lower mean 25(OH)D serum concentrations.Over 40% of women had at least mild joint, back,or bone pain. Muscle weakness in the prior 3months was reported by 24.9% of women. Wo-men with severe 25(OH)D deficiency were morelikely to have fallen in the prior 3 months, butthis was not statistically significant (22.2%, p �0.18). At least moderate fatigue in the previousweek was reported in 26.8% of women. Minor tomoderate depressive symptoms were noted in 14(19.7%). These findings did not change when con-trolled for the presence of PTSD (15.5% of wo-men).

Prior fractures, vitamin use, and calcium usewere uncommon in all women, 5.5%, 12.7% and

VITAMIN D INSUFFICIENCY AMONG VEILED EAST AFRICAN WOMEN 209

TABLE 1. CHARACTERISTICS OF WOMEN WITH MILD, MODERATE, AND SEVERE VITAMIN D DEFICIENCY AND MEAN 25(OH)D SERUM CONCENTRATIONS BY GROUP

Vitamin D deficiency

Mild Moderate Severe Mean 25(OH)D(15.1–30 ng/mL) (8.1–15 ng/mL) (�8 ng/mL) Total concentration

(n � 33) (n � 29) (n � 9) (n � 71) (95% CI)

Age, years�21 4 (12.1) 4 (13.8) 4 (44.5) 12 (16.9) 13.5 (8.5–18.5)22–40 10 (30.3) 4 (13.8) 1 (11.1) 15 (21.1) 16.7 (13.1–20.3)41–50 6 (18.2) 8 (27.6) 0 14 (19.8) 13.8 (10.8–16.8)51–60 6 (18.2) 6 (20.7) 3 (33.3) 15 (21.1) 14.2 (10.5–17.9)�61 7 (21.2) 7 (24.1) 1 (11.1) 15 (21.1) 13.7 (11.7–15.6)

School, years0 22 (68.7) 13 (46.4) 5 (55.6) 40 (58.0) 15.3 (13.2–17.3)1–5 3 (9.4) 5 (17.7) 1 (11.1) 9 (13.0) 13.3 (10.1–16.5)6–11 4 (12.5) 3 (10.7) 3 (33.3) 10 (14.5) 12.2 (8.7–15.7)�12 3 (9.4) 7 (25.2) 0 10 (14.5) 12.7 (9.2–16.1)

Currently enrolledYes 17 (51.5) 21 (72.4) 8 (88.9) 46 (64.8) 13.1 (11.4–14.8)No 16.8 (14.4–19.3)

EmployedYes 3 (10.7) 3 (10.3) 0 6 (9.1) 14.5 (8.2–20.7)No 13.8 (12.3–15.2)

HealthExcellent 14 (42.4) 14 (48.3) 7 (77.8) 35 (49.3) 13.4 (11.2–15.7)Very good 4 (12.1) 8 (27.6) 0 12 (16.9) 15.3 (11.9–18.7)Good 12 (36.4) 6 (20.7) 2 (22.2) 20 (28.2) 15.4 (12.8–17.9)Fair 3 (9.1) 1 (3.4) 0 4 (5.6) 16.3 (7.5–25.1)

10.0%, respectively. None of these factors was as-sociated with the serum 25(OH)D concentration(data not shown). Most women (56.3%) walkedoutside at least 10 minutes per day every day. The9 women who were never or rarely outside for atleast 10 minutes per day had mean serum con-centrations of 25(OH)D of 11.2 ng/mL (95% CI7.0–15.4) compared with all other women whowere outside for 10 minutes at least one day perweek, with 25(OH)D of 14.9 ng/mL (95% CI13.4–16.4).

The factor that tended to correlate best withseverity of vitamin D deficiency was milk in-take, although this was not significant in uni-variate analysis (p � 0.07) (data not shown). Themean 25(OH)D serum concentration of womenwho did not drink milk was 12.9 ng/mL (95%CI 9.4–14.7) compared with women who drankone to two glasses of milk per day (15.5 ng/mL,95% CI 13.8–17.1). Not one woman drank morethan two glasses of milk per day. Because milkintake was a key factor identified a priori as apotential predictor for vitamin D deficiency/in-sufficiency, multivariate analysis was per-formed. After controlling for age, women with25(OH)D concentrations �15 ng/mL were 66%less likely to drink milk than women with25(OH)D concentrations of �15 ng/mL or more(adjusted odds ratio [aOR] � 0.33, 95% CI0.11–0.99). The proportions of women drinkingzero, one, or two cups of milk per day were di-rectly associated with increasing 25(OH)Dserum concentrations (Fig. 1).

DISCUSSION

Recent studies suggest consumption of at least700–800 IU of vitamin D per day is needed foroptimal prevention of both fractures and falls,presumably from musculoskeletal weakness andpain.17 A target serum concentration over 30 ng/mL of 25(OH)D has been recommended.15,17,20

Not one Somali immigrant woman evaluated inour study met these recommended standards.Mean 25(OH)D serum concentration for the en-tire population was 14.4 ng/mL, lower than the17.7 ng/mL observed in African American wo-men and the 33.1 ng/mL observed in white wo-

REED ET AL.210

TABLE 2. MEAN SERUM 25(OH)D CONCENTRATIONS IN WOMEN WITH SYMPTOMS OF PAIN, MUSCLE WEAKNESS, HISTORY OF FALLING, FATIGUE, AND DEPRESSION COMPARED WITH WOMEN WITHOUT

Symptoms present Symptoms absent

Mean serum Mean serumconcentration (ng/mL) (95% CI) concentration (ng/mL) (95% CI)

Paina

Joint 14.8 13.2–16.4 14.2 11.9–16.4Back 14.7 12.6–16.8 14.2 12.2–16.2Bone 14.4 12.6–16.3 14.0 12.3–16.5

Muscle weaknessa 16.4 14.1–18.5 13.8 12.1–15.5Fallinga 14.5 13.0–16.0 15.5 9.6–21.4Fatigueb 14.5 12.0–17.0 14.4 12.7–16.2Depressionc 15.5 11.8–19.1 14.2 12.6–15.8

aBothered by in the past 3 months.bRated � 5/10 in past week.cMild to moderate by PHQ9 survey.13

FIG. 1. The association of milk consumption with meanserum 25(OH)D serum concentrations. Low milk con-sumption was associated with low 25(OH)D.

men participating in the National Health and Nu-trition Examination Survey (NHANES III).19

In our study, the lowest serum concentrationsof 25(OH)D were observed in women who didnot drink milk. Women with serum concentra-tions �15 ng/mL were 66% less likely to drinkmilk than women with concentrations of at least15 ng/mL (aOR 0.33, 95% CI 0.11–0.99). A simi-lar association was observed in NHANES IIIamong adults19 and in a study of adolescents inBoston,21 but not in a Canadian study.22 Our find-ings were similar to those of the NHANES analy-sis by Nesby-O’Dell et al.19 African American wo-men who consumed milk at least three times perweek were half as likely to have serum vitaminD concentrations �15 ng/mL, as were those wo-men who drank no milk. Similar findings wereobserved among Caucasian women.

Although we had anticipated that serum25(OH)D concentrations would correlate withmusculoskeletal pain, weakness, fatigue, and de-pressive symptoms, this small pilot study did notcorroborate that assumption. We had presumedthat some women in the study would have nor-mal 25(OH)D concentrations. However, of 71 wo-men, there were none who met the criteria for vitamin D sufficiency, and, therefore, the com-parison of vitamin D deficient/insufficient wo-men with women found to be vitamin D suffi-cient could not be made. The nonspecificmusculoskeletal complaints reported in other im-migrant populations attributed to severe vitaminD deficiency12,23–30 may well have been appreci-ated in our population if we had had normal con-trols and a greater number of women with severevitamin D deficiency. All the studies cited, how-ever, lacked controls or were case reports, withthe exception of the study by Erkal et al.30 Theyfound that Turkish immigrants in Germany withlow 25(OH)D were more apt to complain of pro-longed bone and muscle aches and pains thanGerman men and women living in Germany orTurkish men and women living in Turkey (sta-stistical analysis not shown). The theory thatmusculoskeletal pain is associated with vitaminD deficiency is also corroborated by a report ofpatients with chronic low back pain in Saudi Ara-bia. Al Faraj et al.31 showed that oral therapy with5,000–10,000 IU/day of vitamin D amelioratedback pain in all 299 (100%) individuals who werefound to have 25(OH)D concentrations of �9ng/mL as compared with only 42(68.9%) indi-

viduals with 25(OH)D concentrations �9 ng/mL(p � 0.001).

Others have investigated whether traditionaldress or nutrition was more important in deter-mining vitamin D deficiency. In Tunisian women,nutrition and multiparity were more importantthan wearing a veil.32 In a group of veiled and un-veiled Turkish7 and Kuwaiti women8 with pre-sumed similar dietary habits, women who wereveiled were more likely to be vitamin D deficient,a finding similar to that of Erkal et al.30 amongTurkish women living in Turkey as well as Turk-ish women who immigrated to Germany. Anotherstudy from Turkey evaluated severity of vitaminD deficiency by the degree of veiling and foundthat women with minimal sun exposure due toclothing were more deficient than others with non-traditional Muslim dress.9 Holvik et al.33 evaluatedthe prevalence and predictors of vitamin D defi-ciency among Pakistani, Vietnamese, Turkish, SriLankan, and Iranians living in Norway and foundthat a diet rich in vitamin D (fatty fish and codliver oil supplements) was strongly associatedwith greater concentrations of 25(OH)D. NHANESIII analyses showed profound differences in vita-min D serum concentration between U.S. AfricanAmerican women and U.S. Caucasian women.Forty-two percent of African American women,compared with 4.2% of white women, had25(OH)D concentrations �15 ng/mL.19 Thus, itappears likely that the combination of traditionaldress, diet low in vitamin D, and dark skin mayplay an important role in affecting serum concen-trations of 25(OH)D.

Our study had several limitations. The analy-sis was based on a convenience sample of immi-grant women who visited a community organi-zation and had stable contact information over 1year. It is possible that immigrants with poorerhealth or fewer resources are more likely to seekservices, but one could also argue that amongthose who initially sought services who could notbe reached because they had no home or phoneat follow-up had even fewer resources. Thus, ourresults may be biased. Our sample was small,which limited the study’s power to detect differ-ences between groups. We were unable to evalu-ate parathyroid, thyroid-stimulating hormone (torule out fatigue due to hypothyroidism), sedi-mentation rate (to assess other etiologies of mus-culoskeletal pain), or urinary calcium in this pi-lot study, but it is doubtful that having these data

VITAMIN D INSUFFICIENCY AMONG VEILED EAST AFRICAN WOMEN 211

would have changed our results overall. Lastly,we had limited dietary and medical informationand did no physical assessments.

The strength of this study was apparent in thecollaborative efforts of a community organizationand an academic team. This study is one of onlytwo12 to report on vitamin D deficiency/insuffi-ciency among immigrants to North America andthe only one to show universally low 25(OH)Dserum concentrations. Quality data collectionwith culturally sensitive tools was essential. Theability to reach this immigrant populationthrough grassroots recruitment efforts resulted inexcellent participation rates that are rarelyachieved by academic groups working in isola-tion, without community partners.

CONCLUSIONS

The profound finding of 100% hypovita-minosis D in this population of Somali immigrantwomen deserves additional study in order to as-sess the associated important health ramificationsfor immigrant Muslim women who live in north-ern climates, have dark skin, and wear traditionaldress. Further team efforts that can assess im-portant health disparities for underserved mi-nority populations and improve health outcomesare needed. Interventions that assess supplemen-tation through the Women, Infants and Children(WIC) nutrition programs and enhance our un-derstanding of immigrant dietary habits will beimportant areas to incorporate into future trans-lational projects.34

ACKNOWLEDGMENTS

We thank Irina Vivyanochka, Rother Rashid,Sofia Haji, Someireh Amairfaiz, Tricia Sills,Megan Rehder, and Drs. Jessie Marrs, BarbaraNorquist, Dionne Piggott, and Prashanti Aryalfor volunteering their time to make this commu-nity project possible.

REFERENCES

1. Utiger RD. The need for more vitamin D. N Engl JMed 1998;338:828.

2. Mowe M, Haug E, Bohmer T. Low serum calcidiolconcentration in older adults with reduced muscularfunction. J Am Geriatr Soc 1999;47:220.

3. Jorde R, Waterloo K, Saleh F, Haug E, Svartberg J.Neuropsychological function in relation to serumparathyroid hormone and serum 25-hydroxyvita-min D levels The Tromso study. J Neurol 2006;253:464.

4. Gloth FM, Alam W, Hollis B. Vitamin D versus broadspectrum phototherapy in the treatment of seasonalaffective disorder. J Nutr Health Aging 1999;3:5.

5. Glerup H, Mikkelsen K, Poulsen L, et al. Commonlyrecommended daily intake of vitamin D is not suffi-cient if sunlight exposure is limited. J Intern Med2000;247:260.

6. Dawodu A, Absood G, Patel M, et al. Biosocial fac-tors affecting vitamin D status of women of child-bearing age in the United Arab Emirates. J Biosoc Sci1998;30:431.

7. Guzel R, Kozanoglu E, Guler-Uysal F, Soyupak S,Sarpel T. Vitamin D status and bone mineral densityof veiled and unveiled Turkish women. J WomensHealth Gend Based Med 2001;10:765.

8. el-Sonbaty MR, Abdul-Ghaffar NU. Vitamin D defi-ciency in veiled Kuwaiti women. Eur J Clin Nutr1996;50:315.

9. Alagol F, Shihadeh Y, Boztepe H, et al. Sunlight ex-posure and vitamin D deficiency in Turkish women.J Endocrinol Invest 2000;23:173.

10. Feleke Y, Abdulkadir J, Mshana R, et al. Low levelsof serum calcidiol in an African population comparedto a North European population. Eur J Endocrinol1999;141:358.

11. Webb AR, Kline L, Holick MF. Influence of seasonand latitude on the cutaneous synthesis of vitamin D3:Exposure to winter sunlight in Boston and Edmontonwill not promote vitamin D3 synthesis in human skin.J Clin Endocrinol Metab 1988;67:373.

12. Plotnikoff GA, Quigley JM. Prevalence of severe hy-povitaminosis D in patients with persistent, nonspe-cific musculoskeletal pain. Mayo Clin Proc 2003;78:1463.

13. Spitzer RL, Kroenke K, Williams JB. Validation andutility of a self-report version of PRIME-MD: ThePHQ primary care study. Primary Care Evaluation ofMental Disorders. Patient Health Questionnaire.JAMA 1999;282:1737.

14. Prins A, Ouimette P, Kimerling R, et al. The primarycare PTSD screen (PC–PTSD): development and op-erating characteristics. Prim Care Psych 2004;9:9.

15. Holick MF, Siris ES, Binkley N, et al. Prevalence of vi-tamin D inadequacy among postmenopausal NorthAmerican women receiving osteoporosis therapy. JClin Endocrinol Metab 2005;90:3215.

16. Chapuy MC, Preziosi P, Maamer M, et al. Prevalenceof vitamin D insufficiency in an adult normal popu-lation. Osteoporos Int 1997;7:439.

17. Bischoff-Ferrari HA, Giovannucci E, Willett WC, Di-etrich T, Dawson-Hughes B. Estimation of optimalserum concentrations of 25-hydroxyvitamin D formultiple health outcomes. Am J Clin Nutr 2006;84:18.

18. Harris SS. Vitamin D and African Americans. J Nutr2006;136:1126.

REED ET AL.212

19. Nesby-O’Dell S, Scanlon KS, Cogswell ME, et al. Hy-povitaminosis D prevalence and determinants amongAfrican American and white women of reproductiveage: Third National Health and Nutrition Examina-tion Survey, 1988–1994. Am J Clin Nutr 2002;76:187.

20. Glowacki J, LeBoff MS. Vitamin D insufficiency andfracture risk. Curr Opin Endocrinol Diab 2004;11:353.

21. Gordon CM, DePeter KC, Feldman HA, Grace E,Emans SJ. Prevalence of vitamin D deficiency amonghealthy adolescents. Arch Pediatr Adolesc Med 2004;158:531.

22. Vieth R, Cole DE, Hawker GA, Trang HM, Rubin LA.Wintertime vitamin D insufficiency is common inyoung Canadian women, and their vitamin D intakedoes not prevent it. Eur J Clin Nutr 2001;55:1091.

23. Gallino A, Radvila A. Immigrant osteomalacia.Schwiez Med Wochenschr 1982;112:163.

24. Petrides PE, Bartl R, Buchele W, Moser E. Chief symp-toms: Bone pain and muscle weakness. Internist (Berl)1991;32:58.

25. Nellen JF, Smulders YM, Jos Frissen PH, Slaats EH,Silberbusch J. Hypovataminosis D in immigrant wo-men: Slow to be diagnosed. BMJ 1996;312:570.

26. Jensen JE, Hitz MF. Osteomalacia—A frequently over-looked condition among refugees and immigrants.Ugeskr Laeger 2000;162:6250.

27. Coster A, Ringe JD. Osteomalacia in immigrants:Therapeutic management of two cases. Med Klin(Munich) 2000;95:451.

28. Wolf G, Aigner RM, Wurzer H, Schwarz T. Osteo-malacia in vitamin D deficiency in a 32-year-old im-migrant: Case report of a rare disease. Nuklearmedi-zin 2001;40:N34.

29. Grootjans-Geerts I, Wielders JP. A pilot study of hy-povitaminsis D in apparently healthy, veiled Turkishwomen: Severe vitamin D deficiency in 82%. Ned Ti-jdschr Geneeskd 2002;146:1100.

30. Erkal MZ, Wilde J, Bilgin Y, et al. High prevalence ofvitamin D deficiency, secondary hyperparathy-roidism and generalized bone pain in Turkish immi-grants in Germany: Identification of risk factors. Os-teoporos Int 2006;17:1133.

31. Al Faraj S, Al Mutairi K. Vitamin D deficiency andchronic low back pain in Saudi Arabia. Spine 2003;28:177.

32. Meddeb N, Sahli H, Chahed M, et al. Vitamin D de-ficiency in Tunisia. Osteoporos Int 2005;16:180.

33. Holvik K, Meyer HE, Haug E, Brunvand L. Preva-lence and predictors of vitamin D deficiency in fiveimmigrant groups living in Oslo, Norway: The OsloImmigrant Health Study. Eur J Clin Nutr 2005;59:57.

34. Davila F. Food and fasting in Somali culture. Eth-nomed June 2001. Available at ethnomed.org/ethnomed/cultures/Somali/som_food.html

Address reprint requests to:Susan D. Reed, M.D., M.P.H.

Obstetrics and GynecologyHarborview Medical Center

325 9th AvenueBox 359865

Seattle, WA 98104

E-mail: reeds@u.washington.edu

VITAMIN D INSUFFICIENCY AMONG VEILED EAST AFRICAN WOMEN 213