J Endocrinol Invest (2014) 37:541–546 DOI 10.1007/s40618-014-0061-1
ORIGINAL ARTICLE
Serum 25 hydroxyvitamin D in employees of a Middle Eastern university hospital M.-H. Gannage´-Yared • E. Helou • V. Zaraket • S. Abi Akl • L. Antonios M.-L. Moussalli • S. Wakim
•
Received: 6 December 2013 / Accepted: 6 February 2014 / Published online: 22 March 2014 Ó Italian Society of Endocrinology (SIE) 2014
Abstract Background The Middle East registers the highest rate of vitamin D deficiency worldwide. In Lebanon, previous studies looked at this deficiency in schoolchildren, university students, young adults and postmenopausal women. However, no previous study was performed in hospital workers. The objective of our study was to evaluate vitamin D status in a Beirut hospital center and to look at the potential factors influencing these measurements. Methods This cross-sectional study was performed on hospital employees who came for a regular checkup at the primary health-care department. 25(OH)D measurements were performed using the Dia-Sorin chemiluminescent assay. Results 392 subjects (318 women and 74 men) were included in the study. The mean age of the participants was 41.02 ± 11.3 years. The mean 25(OH)D level was 15.61 ± 7.91 ng/ml, with no significant difference according to gender. There were no significant correlations between 25(OH)D and both BMI and age, but 25(OH)D was significantly associated with educational level (p = 0.03). There was a significant difference in 25(OH)D levels according to season (p \ 0.001) and a significant association between 25(OH)D and the reported weekly hours of sun exposure (r = 0.1, p = 0.032), but not with the reported sunscreen use. Fish consumption was positively M.-H. Gannage´-Yared (&) E. Helou V. Zaraket S. Abi Akl L. Antonios M.-L. Moussalli Department of Endocrinology, Hotel-Dieu de France Hospital, Saint-Joseph University, Beirut, Lebanon e-mail: [email protected] S. Wakim Department of Employees Primary Health Care, Hoˆtel Dieu de France Hospital, Beirut, Lebanon
associated with 25(OH)D levels (p = 0.018), while milk, dairy product or egg consumption did not achieve any significant relationship. In a stepwise linear regression analysis, fish consumption and season were the only independent predictors of 25(OH)D levels (p = 0.007 and p = 0.0001 respectively). Conclusion Vitamin D deficiency is common among hospital workers. This finding reinforces the need for vitamin D supplementation in these high-risk populations. Keywords
Vitamin D Hospital workers Lebanese
Introduction Vitamin D deficiency is very common worldwide and continues to be an unrecognized epidemic in many populations [1–3]. With the implementation of food-fortifying programs, it was thought that the deficiency had been eradicated, but its prevalence appears to be increasing [4]. It has been reported in healthy children, young and middleaged adults and the elderly. Despite ample sunshine, the Middle East registers the highest rate of rickets and vitamin D deficiency worldwide [5–8]. In large part, this is explained by limited sun exposure secondary to hot weather, dress codes, culinary habits and prolonged breastfeeding [5, 6]. In Lebanon, the deficiency affects all the subgroups of the population, more particularly schoolchildren [9], university students [10], young adults [11] and postmenopausal women [12]. Because few foods contain vitamin D precursors, exposure to sunlight is the primary determinant of vitamin D status [1]. As a result, in the United States (US) and elsewhere, most of the indoor workers are at risk for low vitamin D levels [13]. At northern latitudes, this risk is further
123
542
J Endocrinol Invest (2014) 37:541–546
compounded by the lack of sufficient UVB rays in the winter to produce vitamin D in the skin [14, 15]. Few studies have looked at the prevalence of vitamin D deficiency in medical communities. These studies were mainly performed in medical students. In the US, 25 % of medical trainees were found to have vitamin D deficiency [25 hydroxyvitamin D (25(OH)D) \20 ng/ml] [16], while this prevalence was higher among medical students from Spain [17] and India [18] (32.6 and 87.5 % respectively). Moreover, in Saudi Arabia, 96 % of medical students have 25(OH)D levels below 30 ng/ml [19]. To our knowledge, no previous study looked at vitamin D status in hospital workers of different socioeconomic levels. The objective of our study was to evaluate vitamin D status in a single hospital center of the Beirut area and to look at the potential environmental and dietetic factors influencing this status.
of dairy products (cheese, yoghurt or other dairy products) (a serving approximates 40 g) or of fish (a serving approximates 150 g) and weekly consumption of eggs. Medical history (such as diabetes, hypertension and dyslipidemia) was recorded. The body mass index (BMI) was calculated as weight divided by the square of height. None of the subjects included in the study was taking vitamin D supplementation.
Materials and methods
Statistical analysis
Subjects
The statistical analysis was performed using the software SPSS (version 13). Data were expressed as mean and standard deviation or as percentage. The Pearson coefficient of correlation was used for linear correlation between variables. ANOVA test was used to compare qualitative variables with three or more categories and Student’s t test to compare means. A stepwise multiple linear regression test was performed to identify the independent variables that affect 25(OH)D levels. A p \ 0.05 was considered to be statistically significant.
The study population consisted of 392 subjects recruited among the 900 hospital employees of both sexes working at the Hoˆtel-Dieu de France Hospital, one of biggest university hospitals of the Beirut area. The study was approved by the Ethics Committee of our hospital (File number CEHDF 527). All the subjects signed their agreement when they filled the questionnaire before participating in the study. A 25(OH)D measurement was performed for all hospital employees who came for a regular checkup that was performed approximately every 2 years with the in-house primary care physician. The first 397 subjects who presented between 2009 and 2010 during an approximate 1 year period were recruited. Five osteoporotic women treated with biphosphonates were excluded from the study. None of the subjects were taking vitamin D supplements. Hospital employees included all kind of professions (nurses, medical assistants, technicians, secretaries, medical engineers, administrative employees, etc.). Doctors and medical students were excluded from the study because they are not hospital employees. A questionnaire was filled by all the subjects. The following data were collected: age (in years), sex, weight (in kilograms), height (in centimeters), level of education (primary, secondary or university), number of weekly hours of sun exposure and use of facial sunscreen (never, rarely, most of the time or always). The period of sampling was categorized according to the four seasons (winter: January, February, March; spring: April, May, June; summer: July, August, September and fall: October, November, December). The following vitamin D-rich food were recorded: weekly cups of consumed milk, weekly servings
123
Biological parameters 25(OH)D measurements were performed using the DiaSorin chemiluminescent assay which detects both vitamin D2 and vitamin D3. The sensitivity of the assay is around 1.5 ng/ml, and both the intra- and intercoefficient of variation of the assay are \12 %. Calcium, phosphorus and creatinine were measured using a Kodak automate.
Results Baseline anthropometric, sociocultural and nutritional characteristics of the population are shown in Table 1. The study population consisted of 392 subjects (318 women and 74 men). Men were statistically older than women; they also had higher BMI and lower educational level than their female counterparts (p = 0.014, p \ 0.001 and p \ 0.001 respectively). They also used less sunscreen and reported a higher exposure to sun (p \ 0.001 and p = 0.006 respectively). No differences in dietary habits were observed according to gender. Men were more often hypertensive, dyslipidemic and diabetic than women (p \ 0.001, p = 0.001 and p = 0.001, respectively. 25(OH)D levels and their relation to gender, anthropometric parameters and educational level The mean 25(OH)D level in the whole population was of 15.61 ± 7.91 ng/ml (median 14.75, 25th percentile 4–9.4, 75th percentile 9.6–42.1); in women it was
J Endocrinol Invest (2014) 37:541–546
543
Table 1 Demographic characteristics of the study population (data are expressed as mean ± standard deviation or as percentage)
Table 2 Biological charactersitics of the study population (data are expressed as mean ± standard deviation or as percentage)
Variable
Variable
Total (n = 392)
Men (n = 74)
Women (n = 318) 15.76 ± 8.19
Total (n = 392)
Men (n = 74)
Women (n = 318)
Age (years)*
41.02 ± 11.3
43.97 ± 12.11
40.36 ± 11.04
25(OH)D (ng/ml)***
15.61 ± 7.91
14.98 ± 6.99
Weight (kg)
67.88 ± 13.8
82.35 ± 11.85
64.61 ± 11.99
Calcium (mmol/l)
2.38 ± 0.09
2.38 ± 0.09
2.39 ± 0.09
BMI (kg/m2)*** Waist circumference (cm)***
25.23 ± 4.49 84.92 ± 10.6
27.12 ± 3.5 100.7 ± 6.6
24.8 ± 4.6 81.83 ± 8.28
Phosphorus (mmol/l)***
1.21 ± 0.17
1.11 ± 0.18
1.23 ± 0.16
Creatinine (lmol/l)
62.6 ± 12.5
75.53 ± 13.6
58.52 ± 8.87
The difference of means between men and women is significant at the 0.001 level (***)
Educational level*** Primary (%)
21.5
35.6
18.3
Secondary (%)
22.8
31.5
20.8
University (%)
55.7
32.9
60.9
Milk (number of glasses/week)
1.15 ± 1.42
0.94 ± 1.19
1.2 ± 1.47
Dairy products (weekly servings)
5.32 ± 4.06
5.51 ± 4.8
5.34 ± 3.88
Fish (servings per week)
0.91 ± 0.973
0.89 ± 0.7
0.92 ± 1.02
Eggs (yellow egg/wk.)
1.07 ± 1.26
1.33 ± 1.7
1.02 ± 1.22
Sun exposure (number of hours/week)**
4.71 ± 4.3
6.2 ± 7.2
4.2 ± 5.18
Dietary habits
Use of sunscreen*** Never (%)
58.7
91.8
51.2
Rarely (%)
23.3
5.5
27.3
Most of the time (%)
9.9
2.7
11.5
Daily (%)
8.1
0
9.9
Hypertension (%)***
10.6
23.3
7.8
Dyslipidemia (%)**
24.1
38.4
20.8
Diabetes (%)**
3.3
9.6
1.9
Fig. 1 Repartition of 25(OH)D levels in quartiles according to gender
Comorbidities
The difference of means between men and women is significant at the 0.05 level (*), 0.01 level (**) or at the 0.001 level (***)
15.76 ± 8.19 ng/ml (median 14.8, 25th percentile 4–9.4, 75th percentile 21.8–42.1) and in men 14.98 ± 6.99 ng/ml (median 14.4, 25th percentile 4–8.67, 75th percentile 20.7–31.9) (Table 2, Fig. 1). There was no significant difference according to gender, even after adjustment for age, BMI and educational level. 25(OH)D levels were not significantly correlated with BMI or age (p = 0.38 and p = 0.57, respectively) even after adjustment for sun exposure and nutritional factors. However, 25(OH)D levels were found to be significantly associated with educational level (p = 0.03).
25(OH)D levels and their relation to season, sun exposure and nutritional factors 25.3 % of the 25(OH)D measurements were performed during the winter season, 32.1 % during the spring, 24.5 % during the summer and 18.1 % during the fall. The respective average values of 25(OH)D during winter, spring, summer and fall were 13.14, 13.53, 20.4 and 16.35 ng/ml (Fig. 2). 26.78 % of the population had 25(OH)D levels below 10 ng/ ml, 44.64 % had levels C10 and \20 ng/ml, 23.47 % C20 and \30 ng/ml and 5.1 % C30 ng/ml. There was a statistically significant difference in 25(OH)D levels according to season (p \ 0.001). Accordingly, there was a significant association between 25(OH)D and the reported weekly hours of sun exposure (r = 0.11, p = 0.032). The reported use of sunscreen was not significantly associated with 25(OH)D levels (p = 0.13). Finally, among the nutritional factors, only fish consumption was positively correlated with 25(OH)D levels (r = 0.12, p = 0.018), while milk, dairy
123
544
J Endocrinol Invest (2014) 37:541–546
Discussion
Fig. 2 25(OH)D repartition according to season
products or eggs were not (p = 0.12, p = 0.109 and p = 0.97, respectively). 25(OH)D levels in relation to other biological parameters No significant correlations were found between 25(OH)D levels and calcium, phosphorus or creatinine levels (p = 0.94, p = 0.31 and p = 0.62, respectively). 25(OH)D levels in relation to concomitant medical disorders There were no significant differences in 25(OH)D levels in hypertensive, dyslipidemic or diabetic subjects compared to their counterpart unaffected subjects (p = 0.33, p = 0.35, and p = 0.31 respectively). These results are still nonsignificant after adjustment for age, sun exposure, season and nutritional factors. Stepwise linear regression analysis A stepwise linear regression analysis was carried out to find the factors that may independently affect 25(OH)D levels. Fish consumption and season were the only independent predictors of 25(OH)D levels (p = 0.007 and p = 0.0001 respectively) (Table 3).
Table 3 Stepwise multiple regression analysis with 25(OH)D as a dependent variable Beta
Standard error
Significance
Constant
12.02
0.75
\0.0001
Season
1.90
0.37
\0.0001
Fish consumption
1.09
0.40
0.007
123
This study is the first to be performed in a sample of the Lebanese population including subjects with a wide age range (20–63 years) and from different socioeconomic and educational levels, thus representing several subgroups of our population. The mean 25(OH)D level observed was 15.61 ng/ml. Only 5.1 % of the population had levels higher than the recommended threshold of 30 ng/ml. In addition, 26.78 % of the population had severe vitamin D deficiency [25(OH)D \ 10 ng/ml]. It has been previously reported that worldwide, the Lebanese population is one of the most severely affected by vitamin D deficiency since 85 % of postmenopausal osteoporotic women [7] and 51.7 % of university students [10] have 25(OH)D levels below 30 ng/ml. In addition, 52 % of schoolchildren [9] and 72.8 % of young adults aged 30–50 years [11] have levels below 20 and 12 ng/ml, respectively. The mean reported 25(OH)D levels in the present study is between the one observed in the university students’ study, which is 31 ng/ml [10], and in the young adults’ study, which is 9.71 ng/ml [11]. The difference we observed with the university students can be explained by the higher students’ educational levels or by the different ages of the two populations. Conversely, the study performed in young adults [11] recruited subjects from lower socioeconomic level, with different culinary habits and during the winter season explaining their lower 25(OH)D levels. We did not observe an association between gender, age or BMI and 25(OH)D levels. A gender difference was observed in some subgroups of the Lebanese population [9, 11], but not in the university students’ study [10]. The lack of gender difference observed in our study can be explained by the fact that schedules are equally distributed in both sexes. It is also possible that the relatively small number of recruited men precludes any observed gender difference. In addition, it should be noted that ours is the first Lebanese study recruiting subjects with a wider age range compared to other studies, thus allowing studying the effect of age on vitamin D status. We did not observe an association between age and 25(OH)D levels. This could be related to the exclusion of elderly subjects older than 65 years, since aging is associated with a decline in the capacity of human skin to produce vitamin D3 [20]. Similarly to our results, studies from the US [21] or Canada [22] did not find such an association, while in contrast age was a predictor of 25(OH)D in other populations such as the Brazilian one [23]. Surprisingly, BMI was not found to be significantly correlated with 25(OH)D levels, unlike most of the other Lebanese studies [9, 12]. The lack of such relationship could be different clinical characteristics of our population, more particularly its relatively lower BMI.
J Endocrinol Invest (2014) 37:541–546
Similarly, in the students’ Lebanese study [10], the inverse relationship between BMI and 25(OH)D was nonsignificant in females. The association between vitamin D and obesity is attributed to the reduced 25(OH)D bioavailability when stored in the adipose tissue [24]. A recent metaanalysis shows significant heterogeneity between studies; the weak negative association between both parameters remained mainly significant in Western countries, but not in developing nations [25]. On the other hand, similarly to previous Lebanese studies [9, 12], our study reinforces the importance of low socioeconomic level as a risk factor for 25(OH)D deficiency. An important significant seasonal variation in 25(OH)D levels was observed in our study. This finding was also observed in the studies conducted among both Lebanese university students [10] and schoolchildren [9]. However, this influence was not reported in postmenopausal osteoporotic women probably due to the fact that most of these subjects were taking vitamin supplements. The importance of season as a risk factor for hypovitaminosis D in the Middle East was also recently highlighted [26]. Similarly, the number of weekly reported hours of sun exposure was correlated with 25(OH)D levels. Even if self-report sun exposure using questionnaires poorly correlates with objective measures, such as dosimetry [27], our finding reinforces the importance of sun exposure as a source of vitamin D. On the other side, and contrarily to what would be anticipated, sunscreen use was not found in our study to be associated with 25(OH)D levels. Even if the application of sunscreen was shown to interfere with the cutaneous synthesis of vitamin D [28], ‘real-life’ trials have failed to document such a relationship, perhaps due to inappropriate application of sunscreen [29, 30]. Vitamin D-rich food (milk, fish, eggs) consumption is limited in our population. Milk consumption (the only vitamin D-fortified food in our country) is very low, probably due to our high lactose intolerance prevalence. Dairy products, even if much more consumed, are not vitamin D fortified, explaining the absence of relationship between their consumption and 25(OH)D levels. Fish consumption was independently associated with 25(OH)D values, probably due to the fact that it is the only naturally rich vitamin D food. Similarly to our study, a recent Korean study [31] showed that serum 25(OH)D was associated with fish consumption. Finally, no relationship has been demonstrated between 25(OH)D levels and the presence in our patients of diabetes, hypertension or dyslipidemia, in contrast with other studies that found an association between vitamin D deficiency and the components of metabolic syndrome, including insulin resistance [32]. This could be explained by the younger age of our population or by the small number of subjects with these comorbidities.
545
Our study may present some limitations. It is a single center study, thus it does not reflect vitamin D statues among all hospital workers. In addition, it does not include subjects older than 65 years, for the simple reason that they are retired. In conclusion, our study shows that the prevalence of vitamin D deficiency in hospital workers is very high, since 94.9 % of the subjects had 25(OH)D values below 30 ng/ ml. In addition, 25(OH)D levels were not significantly associated with age, BMI, gender or dairy products consumption. Conversely, education level, season sampling, degree of sun exposure and fish consumption were associated with vitamin D status, while season and fish consumption were the only independent predictors. Our study is the first one to be performed in a medical community population, regrouping different professions. It reinforces the need for regular screening for vitamin D deficiency in people who work indoors and for vitamin D supplementation of this high-risk population [33]. Conflict of interest The authors declare that they have no financial competing interests.
References 1. Holick MF (2007) Vitamin D deficiency. New Eng J Med 357:266–281 2. Holick MF (2006) High prevalence of vitamin D inadequacy and implications for health. Mayo Clin Proc 81:353–373 3. Holick MF, Chen TC (2008) Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr 87:1080S– 1086S 4. Lowdon J (2008) Low vitamin D status: on the increase? J Fam Health Care 18:55–57 5. Mithal A, Wahl DA, Bonjour JP et al (2009) IOF Committee of Scientific Advisors (CSA) Nutrition Working Group. Global vitamin D status and determinants of hypovitaminosis D. Osteoporos Int 20:1807–1820 6. Wahl DA, Cooper C, Ebeling PR et al (2012) A global representation of vitamin D status in healthy populations. Arch Osteoporos 7:155–172 7. Lips P, Hosking D, Lippuner K et al (2006) The prevalence of vitamin D inadequacy amongst women with osteoporosis: an international epidemiological investigation. J Intern Med 260:245–254 8. Baroncelli GI, Bereket A, El Kholy M et al (2008) Rickets in the Middle East: role of environment and genetic predisposition. J Clin Endocrinol Metab 93:1743–1750 9. Fuleihan G El-Hajj, Nabulsi M, Choucair M et al (2001) Hypovitaminosis D in healthy schoolchildren. Pediatrics 107:E53 10. Gannage´-Yared MH, Chedid R, Khalife´ S, Azzi E, Zoghbi F, Halaby G (2009) Vitamin D in relation to metabolic risk factors, insulin sensitivity and adiponectin in a young Middle-Eastern population. Eur J Endocrinol 160:965–971 11. Gannage´-Yared MH, Chemali R, Yaacoub N, Halaby G (2000) Hypovitaminosis D in a sunny country relation to lifestyle and bone markers. J Bone Mine Res 15:1856–1862 12. Gannage´-Yared MH, Maalouf G, Khalife´ S, Challita S, Yaghi Y, Ziade´ N (2009) Prevalence and predictors of vitamin D
123
546
13.
14. 15.
16.
17.
18.
19.
20. 21.
22.
23.
J Endocrinol Invest (2014) 37:541–546 inadequacy among Lebanese osteoporotic women. Br J Nutr 101:487–491 Mendell MJ, Fisk WJ, Kreiss K et al (2002) Improving the health of workers in indoor environments: priority research needs for a national occupational research agenda. Am J Public Health 92:1430–1440 Holick MF (2003) Vitamin D: a millenium perspective. J Cell Biochem 88:296–300 Sullivan SS, Rosen CJ, Halteman WA, Chen TC, Holick MF (2005) Adolescent girls in Maine are at risk for vitamin D insufficiency. J Am Diet Assoc 105:971–974 Growdon AS, Camargo CA Jr, Clark S, Hannon M, Mansbach JM (2012) Serum 25-hydroxyvitamin D levels among Boston trainee doctors in winter. Nutrients 4:197–207 Gonza´lez-Padilla E, Soria Lo´pez A, Gonza´lez-Rodrı´guez E et al (2011) High prevalence of hypovitaminosis D in medical students in Gran Canaria, Canary Islands (Spain). Endocrinol Nutr 58:267–273 Multani SK, Sarathi V, Shivane V, Bandgar TR, Menon PS, Shah NS (2010) Study of bone mineral density in resident doctors working at a teaching hospital. J Postgrad Med 56:65–70 Al-Elq AH (2012) The status of Vitamin D in medical students in the preclerkship years of a Saudi medical school. J Family Community Med 19:100–104 MacLaughlin J, Holick MF (1985) Aging decreases the capacity of human skin to produce vitamin D3. J Clin Invest 76:1536–1538 Mitchell DM, Henao MP, Finkelstein JS, Burnett-Bowie SA (2012) Prevalence and predictors of vitamin D deficiency in healthy adults. Endocr Pract 18:914–923 Gagnon C, Baillargeon JP, Desmarais G, Fink GD (2010) Prevalence and predictors of vitamin D insufficiency in women of reproductive age living in northern latitude. Eur J Endocrinol 163:19–24 Maeda SS, Saraiva GL, Kunii IS et al (2013) Factors affecting vitamin D status in different populations in the city Sao Paulo, Brazil: the Sao PAulo vitamin D Evaluation Study (SPADES). BMC Endocr Disord 13:14
123
24. Worstman J, Matsuoka LY, Chen TC, Lu Z, Holick MF (2000) Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr 72:690–693 25. Saneei P, Salehi-Abargouei A, Esmaillzadeh A (2013) Serum 25-hydroxy vitamin D levels in relation to body mass index: a systematic review and meta-analysis. Obes Rev 14:393–404 26. Bassil D, Rahme M, Hoteit M, Gel-H Fuleihan (2013) Hypovitaminosis D in the Middle East and North Africa: prevalence, risk factors and impact on outcomes. Dermatoendocrinol 5:274–298 27. McCarty CA (2008) Sunlight exposure assessment: can we accurately assess vitamin D exposure from sunlight questionnaires? Am J Clin Nutr 87:1097S–1101S 28. Matsuoka LY, Ide L, Wortsman J, MacLaughlin JA, Holick MF (1987) Sunscreens suppress cutaneous vitamin D3 synthesis. J Clin Endocrinol Metab 64:1165–1168 29. Marks R, Foley PA, Jolley D, Knight KR, Harrison J, Thompson SC (1995) The effect of regular sunscreen use on vitamin D levels in an Australian population. Results of a randomized controlled trial. Arch Dermatol 131:415–421 30. Farrerons J, Barnadas M, Rodriguez J et al (1998) Clinically prescribed sunscreen (sun protection factor 15) does not decrease serum vitamin D concentration sufficiently either to induce changes in parathyroid function or in metabolic markers. Br J Dermatol 139:422–427 31. Yu A, Kim J, Kwon O, Oh SY, Kim J, Yang YJ (2013) The association between serum 25-hydroxyvitamin D concentration and consumption frequencies of vitamin d food sources in Korean adolescents. Clin Nutr Res 2:107–114 32. Reis JP, Von Muhlen D, Miller ER (2008) Relation of 25-hydroxyvitamin D and parathyroid hormone levels with metabolic syndrome among US adults. Eur J Endocrinol 159:41–48 33. Holick MF, Binkley NC, Bischoff-Ferrari et al (2011) Evaluation, treatment, and prevention of vitamin D deficiency: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 96:1911–1930
ORIGINAL ARTICLE
Serum 25 hydroxyvitamin D in employees of a Middle Eastern university hospital M.-H. Gannage´-Yared • E. Helou • V. Zaraket • S. Abi Akl • L. Antonios M.-L. Moussalli • S. Wakim
•
Received: 6 December 2013 / Accepted: 6 February 2014 / Published online: 22 March 2014 Ó Italian Society of Endocrinology (SIE) 2014
Abstract Background The Middle East registers the highest rate of vitamin D deficiency worldwide. In Lebanon, previous studies looked at this deficiency in schoolchildren, university students, young adults and postmenopausal women. However, no previous study was performed in hospital workers. The objective of our study was to evaluate vitamin D status in a Beirut hospital center and to look at the potential factors influencing these measurements. Methods This cross-sectional study was performed on hospital employees who came for a regular checkup at the primary health-care department. 25(OH)D measurements were performed using the Dia-Sorin chemiluminescent assay. Results 392 subjects (318 women and 74 men) were included in the study. The mean age of the participants was 41.02 ± 11.3 years. The mean 25(OH)D level was 15.61 ± 7.91 ng/ml, with no significant difference according to gender. There were no significant correlations between 25(OH)D and both BMI and age, but 25(OH)D was significantly associated with educational level (p = 0.03). There was a significant difference in 25(OH)D levels according to season (p \ 0.001) and a significant association between 25(OH)D and the reported weekly hours of sun exposure (r = 0.1, p = 0.032), but not with the reported sunscreen use. Fish consumption was positively M.-H. Gannage´-Yared (&) E. Helou V. Zaraket S. Abi Akl L. Antonios M.-L. Moussalli Department of Endocrinology, Hotel-Dieu de France Hospital, Saint-Joseph University, Beirut, Lebanon e-mail: [email protected] S. Wakim Department of Employees Primary Health Care, Hoˆtel Dieu de France Hospital, Beirut, Lebanon
associated with 25(OH)D levels (p = 0.018), while milk, dairy product or egg consumption did not achieve any significant relationship. In a stepwise linear regression analysis, fish consumption and season were the only independent predictors of 25(OH)D levels (p = 0.007 and p = 0.0001 respectively). Conclusion Vitamin D deficiency is common among hospital workers. This finding reinforces the need for vitamin D supplementation in these high-risk populations. Keywords
Vitamin D Hospital workers Lebanese
Introduction Vitamin D deficiency is very common worldwide and continues to be an unrecognized epidemic in many populations [1–3]. With the implementation of food-fortifying programs, it was thought that the deficiency had been eradicated, but its prevalence appears to be increasing [4]. It has been reported in healthy children, young and middleaged adults and the elderly. Despite ample sunshine, the Middle East registers the highest rate of rickets and vitamin D deficiency worldwide [5–8]. In large part, this is explained by limited sun exposure secondary to hot weather, dress codes, culinary habits and prolonged breastfeeding [5, 6]. In Lebanon, the deficiency affects all the subgroups of the population, more particularly schoolchildren [9], university students [10], young adults [11] and postmenopausal women [12]. Because few foods contain vitamin D precursors, exposure to sunlight is the primary determinant of vitamin D status [1]. As a result, in the United States (US) and elsewhere, most of the indoor workers are at risk for low vitamin D levels [13]. At northern latitudes, this risk is further
123
542
J Endocrinol Invest (2014) 37:541–546
compounded by the lack of sufficient UVB rays in the winter to produce vitamin D in the skin [14, 15]. Few studies have looked at the prevalence of vitamin D deficiency in medical communities. These studies were mainly performed in medical students. In the US, 25 % of medical trainees were found to have vitamin D deficiency [25 hydroxyvitamin D (25(OH)D) \20 ng/ml] [16], while this prevalence was higher among medical students from Spain [17] and India [18] (32.6 and 87.5 % respectively). Moreover, in Saudi Arabia, 96 % of medical students have 25(OH)D levels below 30 ng/ml [19]. To our knowledge, no previous study looked at vitamin D status in hospital workers of different socioeconomic levels. The objective of our study was to evaluate vitamin D status in a single hospital center of the Beirut area and to look at the potential environmental and dietetic factors influencing this status.
of dairy products (cheese, yoghurt or other dairy products) (a serving approximates 40 g) or of fish (a serving approximates 150 g) and weekly consumption of eggs. Medical history (such as diabetes, hypertension and dyslipidemia) was recorded. The body mass index (BMI) was calculated as weight divided by the square of height. None of the subjects included in the study was taking vitamin D supplementation.
Materials and methods
Statistical analysis
Subjects
The statistical analysis was performed using the software SPSS (version 13). Data were expressed as mean and standard deviation or as percentage. The Pearson coefficient of correlation was used for linear correlation between variables. ANOVA test was used to compare qualitative variables with three or more categories and Student’s t test to compare means. A stepwise multiple linear regression test was performed to identify the independent variables that affect 25(OH)D levels. A p \ 0.05 was considered to be statistically significant.
The study population consisted of 392 subjects recruited among the 900 hospital employees of both sexes working at the Hoˆtel-Dieu de France Hospital, one of biggest university hospitals of the Beirut area. The study was approved by the Ethics Committee of our hospital (File number CEHDF 527). All the subjects signed their agreement when they filled the questionnaire before participating in the study. A 25(OH)D measurement was performed for all hospital employees who came for a regular checkup that was performed approximately every 2 years with the in-house primary care physician. The first 397 subjects who presented between 2009 and 2010 during an approximate 1 year period were recruited. Five osteoporotic women treated with biphosphonates were excluded from the study. None of the subjects were taking vitamin D supplements. Hospital employees included all kind of professions (nurses, medical assistants, technicians, secretaries, medical engineers, administrative employees, etc.). Doctors and medical students were excluded from the study because they are not hospital employees. A questionnaire was filled by all the subjects. The following data were collected: age (in years), sex, weight (in kilograms), height (in centimeters), level of education (primary, secondary or university), number of weekly hours of sun exposure and use of facial sunscreen (never, rarely, most of the time or always). The period of sampling was categorized according to the four seasons (winter: January, February, March; spring: April, May, June; summer: July, August, September and fall: October, November, December). The following vitamin D-rich food were recorded: weekly cups of consumed milk, weekly servings
123
Biological parameters 25(OH)D measurements were performed using the DiaSorin chemiluminescent assay which detects both vitamin D2 and vitamin D3. The sensitivity of the assay is around 1.5 ng/ml, and both the intra- and intercoefficient of variation of the assay are \12 %. Calcium, phosphorus and creatinine were measured using a Kodak automate.
Results Baseline anthropometric, sociocultural and nutritional characteristics of the population are shown in Table 1. The study population consisted of 392 subjects (318 women and 74 men). Men were statistically older than women; they also had higher BMI and lower educational level than their female counterparts (p = 0.014, p \ 0.001 and p \ 0.001 respectively). They also used less sunscreen and reported a higher exposure to sun (p \ 0.001 and p = 0.006 respectively). No differences in dietary habits were observed according to gender. Men were more often hypertensive, dyslipidemic and diabetic than women (p \ 0.001, p = 0.001 and p = 0.001, respectively. 25(OH)D levels and their relation to gender, anthropometric parameters and educational level The mean 25(OH)D level in the whole population was of 15.61 ± 7.91 ng/ml (median 14.75, 25th percentile 4–9.4, 75th percentile 9.6–42.1); in women it was
J Endocrinol Invest (2014) 37:541–546
543
Table 1 Demographic characteristics of the study population (data are expressed as mean ± standard deviation or as percentage)
Table 2 Biological charactersitics of the study population (data are expressed as mean ± standard deviation or as percentage)
Variable
Variable
Total (n = 392)
Men (n = 74)
Women (n = 318) 15.76 ± 8.19
Total (n = 392)
Men (n = 74)
Women (n = 318)
Age (years)*
41.02 ± 11.3
43.97 ± 12.11
40.36 ± 11.04
25(OH)D (ng/ml)***
15.61 ± 7.91
14.98 ± 6.99
Weight (kg)
67.88 ± 13.8
82.35 ± 11.85
64.61 ± 11.99
Calcium (mmol/l)
2.38 ± 0.09
2.38 ± 0.09
2.39 ± 0.09
BMI (kg/m2)*** Waist circumference (cm)***
25.23 ± 4.49 84.92 ± 10.6
27.12 ± 3.5 100.7 ± 6.6
24.8 ± 4.6 81.83 ± 8.28
Phosphorus (mmol/l)***
1.21 ± 0.17
1.11 ± 0.18
1.23 ± 0.16
Creatinine (lmol/l)
62.6 ± 12.5
75.53 ± 13.6
58.52 ± 8.87
The difference of means between men and women is significant at the 0.001 level (***)
Educational level*** Primary (%)
21.5
35.6
18.3
Secondary (%)
22.8
31.5
20.8
University (%)
55.7
32.9
60.9
Milk (number of glasses/week)
1.15 ± 1.42
0.94 ± 1.19
1.2 ± 1.47
Dairy products (weekly servings)
5.32 ± 4.06
5.51 ± 4.8
5.34 ± 3.88
Fish (servings per week)
0.91 ± 0.973
0.89 ± 0.7
0.92 ± 1.02
Eggs (yellow egg/wk.)
1.07 ± 1.26
1.33 ± 1.7
1.02 ± 1.22
Sun exposure (number of hours/week)**
4.71 ± 4.3
6.2 ± 7.2
4.2 ± 5.18
Dietary habits
Use of sunscreen*** Never (%)
58.7
91.8
51.2
Rarely (%)
23.3
5.5
27.3
Most of the time (%)
9.9
2.7
11.5
Daily (%)
8.1
0
9.9
Hypertension (%)***
10.6
23.3
7.8
Dyslipidemia (%)**
24.1
38.4
20.8
Diabetes (%)**
3.3
9.6
1.9
Fig. 1 Repartition of 25(OH)D levels in quartiles according to gender
Comorbidities
The difference of means between men and women is significant at the 0.05 level (*), 0.01 level (**) or at the 0.001 level (***)
15.76 ± 8.19 ng/ml (median 14.8, 25th percentile 4–9.4, 75th percentile 21.8–42.1) and in men 14.98 ± 6.99 ng/ml (median 14.4, 25th percentile 4–8.67, 75th percentile 20.7–31.9) (Table 2, Fig. 1). There was no significant difference according to gender, even after adjustment for age, BMI and educational level. 25(OH)D levels were not significantly correlated with BMI or age (p = 0.38 and p = 0.57, respectively) even after adjustment for sun exposure and nutritional factors. However, 25(OH)D levels were found to be significantly associated with educational level (p = 0.03).
25(OH)D levels and their relation to season, sun exposure and nutritional factors 25.3 % of the 25(OH)D measurements were performed during the winter season, 32.1 % during the spring, 24.5 % during the summer and 18.1 % during the fall. The respective average values of 25(OH)D during winter, spring, summer and fall were 13.14, 13.53, 20.4 and 16.35 ng/ml (Fig. 2). 26.78 % of the population had 25(OH)D levels below 10 ng/ ml, 44.64 % had levels C10 and \20 ng/ml, 23.47 % C20 and \30 ng/ml and 5.1 % C30 ng/ml. There was a statistically significant difference in 25(OH)D levels according to season (p \ 0.001). Accordingly, there was a significant association between 25(OH)D and the reported weekly hours of sun exposure (r = 0.11, p = 0.032). The reported use of sunscreen was not significantly associated with 25(OH)D levels (p = 0.13). Finally, among the nutritional factors, only fish consumption was positively correlated with 25(OH)D levels (r = 0.12, p = 0.018), while milk, dairy
123
544
J Endocrinol Invest (2014) 37:541–546
Discussion
Fig. 2 25(OH)D repartition according to season
products or eggs were not (p = 0.12, p = 0.109 and p = 0.97, respectively). 25(OH)D levels in relation to other biological parameters No significant correlations were found between 25(OH)D levels and calcium, phosphorus or creatinine levels (p = 0.94, p = 0.31 and p = 0.62, respectively). 25(OH)D levels in relation to concomitant medical disorders There were no significant differences in 25(OH)D levels in hypertensive, dyslipidemic or diabetic subjects compared to their counterpart unaffected subjects (p = 0.33, p = 0.35, and p = 0.31 respectively). These results are still nonsignificant after adjustment for age, sun exposure, season and nutritional factors. Stepwise linear regression analysis A stepwise linear regression analysis was carried out to find the factors that may independently affect 25(OH)D levels. Fish consumption and season were the only independent predictors of 25(OH)D levels (p = 0.007 and p = 0.0001 respectively) (Table 3).
Table 3 Stepwise multiple regression analysis with 25(OH)D as a dependent variable Beta
Standard error
Significance
Constant
12.02
0.75
\0.0001
Season
1.90
0.37
\0.0001
Fish consumption
1.09
0.40
0.007
123
This study is the first to be performed in a sample of the Lebanese population including subjects with a wide age range (20–63 years) and from different socioeconomic and educational levels, thus representing several subgroups of our population. The mean 25(OH)D level observed was 15.61 ng/ml. Only 5.1 % of the population had levels higher than the recommended threshold of 30 ng/ml. In addition, 26.78 % of the population had severe vitamin D deficiency [25(OH)D \ 10 ng/ml]. It has been previously reported that worldwide, the Lebanese population is one of the most severely affected by vitamin D deficiency since 85 % of postmenopausal osteoporotic women [7] and 51.7 % of university students [10] have 25(OH)D levels below 30 ng/ml. In addition, 52 % of schoolchildren [9] and 72.8 % of young adults aged 30–50 years [11] have levels below 20 and 12 ng/ml, respectively. The mean reported 25(OH)D levels in the present study is between the one observed in the university students’ study, which is 31 ng/ml [10], and in the young adults’ study, which is 9.71 ng/ml [11]. The difference we observed with the university students can be explained by the higher students’ educational levels or by the different ages of the two populations. Conversely, the study performed in young adults [11] recruited subjects from lower socioeconomic level, with different culinary habits and during the winter season explaining their lower 25(OH)D levels. We did not observe an association between gender, age or BMI and 25(OH)D levels. A gender difference was observed in some subgroups of the Lebanese population [9, 11], but not in the university students’ study [10]. The lack of gender difference observed in our study can be explained by the fact that schedules are equally distributed in both sexes. It is also possible that the relatively small number of recruited men precludes any observed gender difference. In addition, it should be noted that ours is the first Lebanese study recruiting subjects with a wider age range compared to other studies, thus allowing studying the effect of age on vitamin D status. We did not observe an association between age and 25(OH)D levels. This could be related to the exclusion of elderly subjects older than 65 years, since aging is associated with a decline in the capacity of human skin to produce vitamin D3 [20]. Similarly to our results, studies from the US [21] or Canada [22] did not find such an association, while in contrast age was a predictor of 25(OH)D in other populations such as the Brazilian one [23]. Surprisingly, BMI was not found to be significantly correlated with 25(OH)D levels, unlike most of the other Lebanese studies [9, 12]. The lack of such relationship could be different clinical characteristics of our population, more particularly its relatively lower BMI.
J Endocrinol Invest (2014) 37:541–546
Similarly, in the students’ Lebanese study [10], the inverse relationship between BMI and 25(OH)D was nonsignificant in females. The association between vitamin D and obesity is attributed to the reduced 25(OH)D bioavailability when stored in the adipose tissue [24]. A recent metaanalysis shows significant heterogeneity between studies; the weak negative association between both parameters remained mainly significant in Western countries, but not in developing nations [25]. On the other hand, similarly to previous Lebanese studies [9, 12], our study reinforces the importance of low socioeconomic level as a risk factor for 25(OH)D deficiency. An important significant seasonal variation in 25(OH)D levels was observed in our study. This finding was also observed in the studies conducted among both Lebanese university students [10] and schoolchildren [9]. However, this influence was not reported in postmenopausal osteoporotic women probably due to the fact that most of these subjects were taking vitamin supplements. The importance of season as a risk factor for hypovitaminosis D in the Middle East was also recently highlighted [26]. Similarly, the number of weekly reported hours of sun exposure was correlated with 25(OH)D levels. Even if self-report sun exposure using questionnaires poorly correlates with objective measures, such as dosimetry [27], our finding reinforces the importance of sun exposure as a source of vitamin D. On the other side, and contrarily to what would be anticipated, sunscreen use was not found in our study to be associated with 25(OH)D levels. Even if the application of sunscreen was shown to interfere with the cutaneous synthesis of vitamin D [28], ‘real-life’ trials have failed to document such a relationship, perhaps due to inappropriate application of sunscreen [29, 30]. Vitamin D-rich food (milk, fish, eggs) consumption is limited in our population. Milk consumption (the only vitamin D-fortified food in our country) is very low, probably due to our high lactose intolerance prevalence. Dairy products, even if much more consumed, are not vitamin D fortified, explaining the absence of relationship between their consumption and 25(OH)D levels. Fish consumption was independently associated with 25(OH)D values, probably due to the fact that it is the only naturally rich vitamin D food. Similarly to our study, a recent Korean study [31] showed that serum 25(OH)D was associated with fish consumption. Finally, no relationship has been demonstrated between 25(OH)D levels and the presence in our patients of diabetes, hypertension or dyslipidemia, in contrast with other studies that found an association between vitamin D deficiency and the components of metabolic syndrome, including insulin resistance [32]. This could be explained by the younger age of our population or by the small number of subjects with these comorbidities.
545
Our study may present some limitations. It is a single center study, thus it does not reflect vitamin D statues among all hospital workers. In addition, it does not include subjects older than 65 years, for the simple reason that they are retired. In conclusion, our study shows that the prevalence of vitamin D deficiency in hospital workers is very high, since 94.9 % of the subjects had 25(OH)D values below 30 ng/ ml. In addition, 25(OH)D levels were not significantly associated with age, BMI, gender or dairy products consumption. Conversely, education level, season sampling, degree of sun exposure and fish consumption were associated with vitamin D status, while season and fish consumption were the only independent predictors. Our study is the first one to be performed in a medical community population, regrouping different professions. It reinforces the need for regular screening for vitamin D deficiency in people who work indoors and for vitamin D supplementation of this high-risk population [33]. Conflict of interest The authors declare that they have no financial competing interests.
References 1. Holick MF (2007) Vitamin D deficiency. New Eng J Med 357:266–281 2. Holick MF (2006) High prevalence of vitamin D inadequacy and implications for health. Mayo Clin Proc 81:353–373 3. Holick MF, Chen TC (2008) Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr 87:1080S– 1086S 4. Lowdon J (2008) Low vitamin D status: on the increase? J Fam Health Care 18:55–57 5. Mithal A, Wahl DA, Bonjour JP et al (2009) IOF Committee of Scientific Advisors (CSA) Nutrition Working Group. Global vitamin D status and determinants of hypovitaminosis D. Osteoporos Int 20:1807–1820 6. Wahl DA, Cooper C, Ebeling PR et al (2012) A global representation of vitamin D status in healthy populations. Arch Osteoporos 7:155–172 7. Lips P, Hosking D, Lippuner K et al (2006) The prevalence of vitamin D inadequacy amongst women with osteoporosis: an international epidemiological investigation. J Intern Med 260:245–254 8. Baroncelli GI, Bereket A, El Kholy M et al (2008) Rickets in the Middle East: role of environment and genetic predisposition. J Clin Endocrinol Metab 93:1743–1750 9. Fuleihan G El-Hajj, Nabulsi M, Choucair M et al (2001) Hypovitaminosis D in healthy schoolchildren. Pediatrics 107:E53 10. Gannage´-Yared MH, Chedid R, Khalife´ S, Azzi E, Zoghbi F, Halaby G (2009) Vitamin D in relation to metabolic risk factors, insulin sensitivity and adiponectin in a young Middle-Eastern population. Eur J Endocrinol 160:965–971 11. Gannage´-Yared MH, Chemali R, Yaacoub N, Halaby G (2000) Hypovitaminosis D in a sunny country relation to lifestyle and bone markers. J Bone Mine Res 15:1856–1862 12. Gannage´-Yared MH, Maalouf G, Khalife´ S, Challita S, Yaghi Y, Ziade´ N (2009) Prevalence and predictors of vitamin D
123
546
13.
14. 15.
16.
17.
18.
19.
20. 21.
22.
23.
J Endocrinol Invest (2014) 37:541–546 inadequacy among Lebanese osteoporotic women. Br J Nutr 101:487–491 Mendell MJ, Fisk WJ, Kreiss K et al (2002) Improving the health of workers in indoor environments: priority research needs for a national occupational research agenda. Am J Public Health 92:1430–1440 Holick MF (2003) Vitamin D: a millenium perspective. J Cell Biochem 88:296–300 Sullivan SS, Rosen CJ, Halteman WA, Chen TC, Holick MF (2005) Adolescent girls in Maine are at risk for vitamin D insufficiency. J Am Diet Assoc 105:971–974 Growdon AS, Camargo CA Jr, Clark S, Hannon M, Mansbach JM (2012) Serum 25-hydroxyvitamin D levels among Boston trainee doctors in winter. Nutrients 4:197–207 Gonza´lez-Padilla E, Soria Lo´pez A, Gonza´lez-Rodrı´guez E et al (2011) High prevalence of hypovitaminosis D in medical students in Gran Canaria, Canary Islands (Spain). Endocrinol Nutr 58:267–273 Multani SK, Sarathi V, Shivane V, Bandgar TR, Menon PS, Shah NS (2010) Study of bone mineral density in resident doctors working at a teaching hospital. J Postgrad Med 56:65–70 Al-Elq AH (2012) The status of Vitamin D in medical students in the preclerkship years of a Saudi medical school. J Family Community Med 19:100–104 MacLaughlin J, Holick MF (1985) Aging decreases the capacity of human skin to produce vitamin D3. J Clin Invest 76:1536–1538 Mitchell DM, Henao MP, Finkelstein JS, Burnett-Bowie SA (2012) Prevalence and predictors of vitamin D deficiency in healthy adults. Endocr Pract 18:914–923 Gagnon C, Baillargeon JP, Desmarais G, Fink GD (2010) Prevalence and predictors of vitamin D insufficiency in women of reproductive age living in northern latitude. Eur J Endocrinol 163:19–24 Maeda SS, Saraiva GL, Kunii IS et al (2013) Factors affecting vitamin D status in different populations in the city Sao Paulo, Brazil: the Sao PAulo vitamin D Evaluation Study (SPADES). BMC Endocr Disord 13:14
123
24. Worstman J, Matsuoka LY, Chen TC, Lu Z, Holick MF (2000) Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr 72:690–693 25. Saneei P, Salehi-Abargouei A, Esmaillzadeh A (2013) Serum 25-hydroxy vitamin D levels in relation to body mass index: a systematic review and meta-analysis. Obes Rev 14:393–404 26. Bassil D, Rahme M, Hoteit M, Gel-H Fuleihan (2013) Hypovitaminosis D in the Middle East and North Africa: prevalence, risk factors and impact on outcomes. Dermatoendocrinol 5:274–298 27. McCarty CA (2008) Sunlight exposure assessment: can we accurately assess vitamin D exposure from sunlight questionnaires? Am J Clin Nutr 87:1097S–1101S 28. Matsuoka LY, Ide L, Wortsman J, MacLaughlin JA, Holick MF (1987) Sunscreens suppress cutaneous vitamin D3 synthesis. J Clin Endocrinol Metab 64:1165–1168 29. Marks R, Foley PA, Jolley D, Knight KR, Harrison J, Thompson SC (1995) The effect of regular sunscreen use on vitamin D levels in an Australian population. Results of a randomized controlled trial. Arch Dermatol 131:415–421 30. Farrerons J, Barnadas M, Rodriguez J et al (1998) Clinically prescribed sunscreen (sun protection factor 15) does not decrease serum vitamin D concentration sufficiently either to induce changes in parathyroid function or in metabolic markers. Br J Dermatol 139:422–427 31. Yu A, Kim J, Kwon O, Oh SY, Kim J, Yang YJ (2013) The association between serum 25-hydroxyvitamin D concentration and consumption frequencies of vitamin d food sources in Korean adolescents. Clin Nutr Res 2:107–114 32. Reis JP, Von Muhlen D, Miller ER (2008) Relation of 25-hydroxyvitamin D and parathyroid hormone levels with metabolic syndrome among US adults. Eur J Endocrinol 159:41–48 33. Holick MF, Binkley NC, Bischoff-Ferrari et al (2011) Evaluation, treatment, and prevention of vitamin D deficiency: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 96:1911–1930
