| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatrics (G.I.B., G.S.), University of Pisa, 56126 Pisa, Italy; Department of Pediatrics (A.B.), Division of Pediatric Endocrinology, Marmara University School of Medicine, 34722 Istanbul, Turkey; Department of Pediatrics (M.E.K., M.R.), Ain Shams University, 11566 Cairo, Egypt; Pediatric Endocrinology Research Unit (L.A., M.F.-C.), Hospital Vall d’Hebron, Autonomous University of Barcelona, 08035 Barcelona, Spain; Department of Pediatrics (Y.C.), Division of Pediatric Endocrinology, Yuzuncu Yil University, School of Medicine, 65080 Van, Turkey; Department of Pediatrics (B.O.), Division of Pediatric Endocrinology, Ataturk University, School of Medicine, 25240 Erzurum, Turkey; Dana Children’s Hospital and Tel Aviv University (Y.W.), Tel Aviv 69978, Israel; and Meyer Children’s Hospital and Technion-Israel Institute of Technology (Z.H.), Haifa 31096, Israel
Address all correspondence and requests for reprints to: Dr. Giampiero I. Baroncelli, Department of Pediatrics, University of Pisa, Via Roma 67, 56126 Pisa, Italy. E-mail: g.baroncelli{at}med.unipi.it.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective of the study was to explore the mechanisms leading to rickets in children of the Middle East.
Design and Setting: We conducted a prospective study in 98 rachitic and 50 controls (aged 6 months to 4 yr) from university and community outpatient hospitals in Egypt and Turkey.
Main Outcome Measures: We collected epidemiological, maternal, nutritional, radiographic, and biochemical parameters; markers of bone turnover; and vitamin D receptor (VDR) gene polymorphisms.
Results: Epidemiological factors had a key role in pursuit of rickets; Egyptian and Turkish patients had lower (P < 0.01) dietary calcium intake than controls and the recommended dietary intakes, and serum 25-hydroxyvitamin D levels were reduced in patients, the difference with controls being significant (P < 0.001) only in Turkey, although rickets was more severe in Egypt as determined by the x-ray score (P < 0.05). In Turkey, the F VDR allele frequency was significantly (P < 0.05) increased in patients. The BB VDR genotype was associated with lower serum 25-hydroxyvitamin D levels in both patients and controls and with severity of rickets.
Conclusions: In Turkey most patients had vitamin D deficiency, whereas in Egypt they had mostly calcium insufficiency combined with vitamin D deficiency. In this environ, VDR genotypes may predispose to rickets by increased frequency of the F allele. The unique environs and genetic predisposition have to be accounted for in the design of preventive measures, rather than using European or American recommended dietary intake for calcium and vitamin D.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The present study was a prospective multicenter collaboration, designed and executed by the European Society for Pediatric Endocrinology (ESPE) Bone Club. The aim of study was to explore the mechanisms leading to rickets in children of the Middle East, with an intention to provide a scientific basis for large-scale prevention. The study was designed to compare affected against unaffected children in Egypt and Turkey, in which rickets incidence is as high as 13 and 6%, respectively (18, 19). Characterization included epidemiological and nutritional aspects, clinical and radiographic findings, calcium-phosphate metabolism, biochemical markers of bone turnover, and genetic polymorphism of the vitamin D receptor (VDR).
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A total of 98 rachitic (63 males and 35 females, aged 11.7 ± 8.2 months) and 50 controls (30 males and 20 females, aged 14.2 ± 8.3 months) from Egypt and Turkey were enrolled in the study, examined, and blood drawn simultaneously throughout the year to exclude seasonal variation from January to December 2004. Both patients and controls were recruited from university and community outpatient hospitals in Cairo, Egypt (latitude 30.01 N, corresponding to New Orleans, LA), and Erzurum and Van, Turkey (latitude 39.57 N, corresponding to Philadelphia, PA, and 38.30 N, corresponding to Palermo, Sicily, Italy, respectively). Selection criteria for patients and controls participating to the study are given in Table 1
. Exclusion criteria were based on the history, physical examination, and laboratory testing. Controls were brothers or sisters of nonrachitic patients coming to the outpatient clinic or children with minor illnesses as upper respiratory infections.
|
Study design
Children were examined, a semiquantitative food frequency questionnaire was administered (20), and a parent or guardians were interviewed. In all children, nutritional status based on weight and presence or absence of edema according to the Wellcome Trust classification (21) was evaluated. Epidemiological questionnaire included living conditions, family data, socioeconomic status by parental education and family income, and sunlight exposure by clothing type and time spent outdoors. A maternal questionnaire included interlude to next pregnancy, prenatal and nursing period care and nutrition, clothing type and time spent outdoor during pregnancy, and calcium and vitamin D supplementation. Nutritional data included type of milk fed during infancy, (breast, cow, goat, or formula), dietary calcium intake, and calcium and vitamin D supplementation.
In both patients and controls, dietary calcium intake by a 3-d nutrition diary was estimated and compared with the recommended dietary intakes (RDI) by the National Institute of Health (NIH; Bethesda, MD) (22) and the ESPE Bone Club (23). Calcium insufficiency was arbitrarily defined as less than 50% of the RDI, and suboptimal intake was defined as between 51 and 85% of the RDI. A dietary calcium intake higher than 86% was considered adequate. A value of 25(OH)D under or above 15 ng/ml (37.5 nmol/liter) was defined as vitamin D deficiency or sufficiency, respectively (24, 25). Serum 25(OH)D levels were also assessed in the mothers of Turkish patients.
Radiographic analysis
In all patients, radiographic assessment of the severity of rickets was examined by standard x-rays of both wrist and knee and scored from 0 (normal) to 10 points (severe) by three observers independently according to Thacher et al. (26), and the mean values were used for analyses. Wrist was scored for both radius and ulna and knee for both distal femur and proximal tibia separately.
Assays
Serum calcium and phosphate levels were measured by standard methods. Serum intact PTH, osteocalcin, and bone alkaline phosphatase isoenzyme (BAP) levels were measured in a single laboratory by a two-site immunoradiometric assay (Allegro and human osteocalcin, Nichols Institute, San Juan Capistrano, CA; and Tandem-R Ostase; Hybritech Europe, Liege, Belgium, respectively). Serum 25(OH)D, 1,25(OH)2D, and 24,25-dihydroxyvitamin D [24,25(OH)2D] levels were measured in a single laboratory by a competitive binding RIA (DiaSorin, Stillwater, MN) as previously described (27). 1
-Hydroxylase and 24-hydroxylase activity were estimated from the ratio of the product/substrate [1,25(OH)2D to 25(OH)D and 24,25(OH)2D to 25(OH)D, respectively] and expressed as percent (27). Serum aminoterminal propeptide of type I procollagen (PINP) and cross-linked carboxyterminal telopeptide of type I collagen (ICTP) levels were measured by RIA (Orion Diagnostic, Espoo, Finland). For all measurements, interassay variability was less than 9% and intraassay variability less than 7%. All blood samples were measured in duplicates.
VDR gene polymorphisms
In all patients and controls and the mothers of patients, VDR gene polymorphisms at intron 8 (BsmI) and exon 2 (FokI) were determined for 296 alleles in DNA extracted from leukocytes by direct PCR fragment sequencing (28). Alleles were designated b or B for intron 8 polymorphism when the restriction enzyme site for BsmI was present or absent, and f or F for exon 2 polymorphism when the restriction enzyme site for FokI was present or absent, respectively.
Statistical analysis
Comparison of clinical and biochemical data between patients and controls was determined by a nonparametric Mann-Whitney rank-sum test and the comparison of proportions by z test with Yates correction. Differences for VDR gene alleles, genotypes, and combined genotype distribution between patients and controls and mothers of patients within each population, as well as between the two populations, were analyzed by the 
2 test for three-by-three and two-by-two tables. Simple regression analyses were carried out among the biochemical parameters, dietary calcium intake, or radiographic score of rickets. All statistical analyses were performed by Statview 4.5 program (Abacus Concepts, Inc., Berkley, CA). Data are expressed as mean ± SD unless otherwise stated. A value of P < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Epidemiological, maternal, and nutritional data are summarized in Table 2. In Turkey, family size, overcrowding, income, and socioeconomic status were worse, and the time spent outdoors was lower in patients, compared with controls. Among the maternal factors, interlude to the next pregnancy, prenatal care, physician visits, nutrition during nursing, and exposed body surface were lower in the mothers of patients. In Egypt, epidemiological data did not differ between patients and controls other than maternal education, which was unexpectedly higher in patients than controls. However, the time spent by the child outdoors and proper maternal nutrition during nursing were lower in patients, compared with controls. Nearly all mothers of patients and controls did not receive calcium or vitamin D supplements during pregnancy or lactation. The type of milk fed during infancy did not differ between patients and controls (data not shown) as well as nutritional status, and only a third of Egyptian and Turkish patients were classified as underweight (60–80% of the expected weight for age but no edema). The proportion of breast-fed infants did not differ among Egyptians, whereas it was higher in Turkish patients than controls. The proportion of Turkish patients who received prophylactic vitamin D supplements (400 IU/d) during the first year of life was lower than that of controls, whereas none of the Egyptian patients and only a small proportion of Egyptian controls received vitamin D supplements. No patient or control received calcium supplements. Both Egyptian and Turkish patients had reduced mean dietary calcium intake in comparison with their respective controls (Table 2); however, mean calcium intake was markedly below the NIH and ESPE RDI in Egyptian (46 and 60%, respectively) and Turkish (60 and 74%, respectively) patients as well as in Egyptian controls (60 and 67%, respectively), whereas it was somewhat below the NIH RDI (75%) but normal against ESPE RDI in Turkish controls (96%).
|
Biochemical data are reported in Table 3. Both Egyptian and Turkish patients had marked hypocalcemia and hypophosphatemia and higher serum levels of PTH, BAP, PINP, and ICTP as well as 1,25(OH)2D to 25(OH)D and 24,25(OH)2D to 25(OH)D ratio, in comparison with controls. Serum levels of calcium, phosphate, and PTH were lower in Egyptian than Turkish patients. Serum 25(OH)D and 24,25(OH)2D levels were lower in Turkish patients, compared with controls, whereas serum 25(OH)D levels did not differ (P = 0.074) between Egyptian patients and controls possibly due to the smaller number of subjects and to a wider variation of the levels in patients. Serum 1,25(OH)2D levels did not differ between patients and controls in both countries. Serum osteocalcin levels were lower in Egyptian patients, compared with their own controls and Turkish patients. Serum 25(OH)D levels were very low in the mothers of Turkish patients (4.7 ± 2.6 ng/ml, 11.8 ± 6.5 nmol/liter).
|
Rickets was severe in all patients, with scores being greater in Egyptian patients than Turkish patients (8.1 ± 2.2 and 7.0 ± 2.6, P < 0.05, respectively). The severity of rickets, as manifested by the x-ray score, was predicted by serum levels of phosphate (r = –0.30, P < 0.01), BAP (r = 0.31, P < 0.01), and ICTP (r = 0.22, P < 0.05).
Biochemical parameters and severity of rickets did not differ between malnourished and well-nourished patients (data not shown).
Stratification of patients and controls according to vitamin D status and calcium intake
With regard to serum 25(OH)D levels and dietary calcium intake, children were classified into five groups, defined as pure vitamin D deficiency, pure calcium insufficiency, combined vitamin D deficiency and calcium insufficiency, vitamin D sufficiency with suboptimal calcium intake, and vitamin D sufficiency with adequate calcium intake (Fig. 1). The proportion of these groups was significantly different between patients and controls as well as between Egyptian and Turkish patients but not between Egyptian and Turkish controls. Most Egyptian patients (71%) had calcium insufficiency or vitamin D deficiency, with a 50% of the total having a combination of both situations. As many as 29% of Egyptian patients had normal serum 25(OH)D levels. Most Turkish patients had vitamin D deficiency, pure or combined with calcium insufficiency (52 and 34%, respectively). Vitamin D sufficiency with suboptimal calcium intake was evident in only a small percentage of patients (8 and 5% among Egyptian and Turkish patients, respectively). The severity of rickets was not different among the five groups (data not shown). Moreover, in Egyptian and Turkish controls, 37 and 17% had pure vitamin D deficiency, and 15 and 13% had pure calcium insufficiency, respectively. One patient and three controls in Egypt and nine patients and seven controls in Turkey had an adequate calcium intake; among these, all but one patient and four controls had vitamin D deficiency. Therefore, only one Turkish patient (1%) and six controls (Egypt, n = 1, 5%; Turkey, n = 5, 17%) showed an adequate calcium intake and vitamin D sufficiency.
|
No difference was observed between patients and controls for BsmI genotype or allele frequencies (B allele frequencies: Egypt, 0.37 and 0.42; Turkey, 0.43 and 0.42; b allele frequencies: Egypt, 0.63 and 0.58, Turkey, 0.57 and 0.58, respectively). In both countries a tendency to increased FF genotype frequency in patients against controls was observed (Egypt, 63 vs. 55%; Turkey, 53 vs. 47%), although it did not reach significance; the ff genotype was absent in Egyptians. In Turkey, the frequency of the F allele was increased and that of the f allele was decreased in patients against controls (0.75 vs. 0.65, and 0.25 vs. 0.35, respectively, P = 0.024); similar results were observed for the entire rickets group, compared with the entire group of controls (0.77 vs. 0.70, and 0.23 vs. 0.30, respectively, P = 0.04). There was no difference in allele frequencies between patients and their mothers for both BsmI and FokI genotypes (data not shown).
Figure 2 shows the associations of VDR BsmI and FokI genotypes with biochemical parameters, dietary calcium intake, and severity of rickets in patients and/or controls. The BB genotype in both patients (Fig. 2A) and controls (Fig. 2B), as well as the Bb genotype in patients only (Fig. 2A), were associated with lower serum 25(OH)D levels; in patients, the BB genotype was also associated with higher x-ray score (Fig. 2A), and the bb genotype was associated with the lowest calcium intake (Fig. 2C). In controls, the BB genotype was associated with the highest 1,25(OH)2D to 25(OH)D ratio (Fig. 2B). In Egypt, but not in Turkey, patients with the FF genotype had lower serum 1,25(OH)2D levels and 1,25(OH)2D to 25(OH)D ratio, compared with patients with the Ff genotype (Fig. 2D), but otherwise there was no difference in biochemistry or severity of rickets (data not shown).
|
Patients’ stratification according to serum 25(OH)D levels and dietary calcium intake was different among the VDR BsmI genotypes (Fig. 3) but not among the FokI genotypes (data not shown). A combined vitamin D deficiency with calcium insufficiency and a pure vitamin D deficiency were the main risk factors in homozygous BB and heterozygous Bb patients, respectively. No patient with the BB genotype had vitamin D sufficiency with suboptimal or adequate calcium intake or a pure calcium insufficiency; the proportion of patients with pure calcium insufficiency was the highest in the bb genotype. Controls’ stratification did not show any significant difference among the VDR BsmI genotypes (data not shown); only one control with vitamin D deficiency and suboptimal calcium intake carried the BB genotype.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In Turkey, rickets is a disease of the underprivileged, strongly correlated with negative social background and insufficient exposure to sunlight; a lack of vitamin D supplementation appears to be decisive for the development of the disease. During the first year of life only 9% of patients received vitamin D supplements against 83% of controls, and the estimated time spent outdoor was less in patients than in controls. Exposed body surface by the mothers of patients during pregnancy was smaller and the number of children was higher than that of the mothers of controls. Moreover, serum 25(OH)D levels were severely reduced in the mothers of patients, as was also found by other studies (1, 3, 9, 23), suggesting that reduced vitamin D stores likely contributed to the development of rickets in their children and strengthening the need for vitamin D supplement’s recommendation for mother and baby.
In Egypt, rickets was not related to living conditions or maternal clothing during pregnancy, and paradoxically, maternal education was higher in patients than controls, suggesting that it is not a social disease. Only maternal nutrition during nursing and the estimated time spent outdoors were less in patients than in controls, but in the sun-flooded Egypt, it is expected to be sufficient to maintain a normal vitamin D status. In fact, infants require approximately 2 h of sunlight per week if they are fully clothed with no hat to reach vitamin D sufficiency (29), and the mean time spent outdoors was above this cutoff in Egyptian patients.
Malnutrition was not a primary cause of rickets in both countries. The mean dietary calcium intake was lower in patients than in controls, and in Egypt it was well below the NIH (22) and ESPE RDI (23) but not low enough to be labeled calcium deficiency (daily intake < 200 mg) (2, 3, 14, 16, 17). Yet no reliable data on the lowest calcium intake that would cause rickets are evident (30). In the absence of reliable indicators of nutritional adequacy for calcium, estimates of calcium insufficiency are based largely on adequacy of dietary intake related to the estimated requirements, but this approach may be complicated by the fact that RDIs for calcium vary with expert authorities, and they could differ between Egypt and Turkey populations. Moreover, because the adequate intake for calcium varies with age, we estimated it as percent of some RDIs (22, 23), and a threshold value of –50% was used to separate the children with a condition of possible calcium insufficiency from those with a possible condition of suboptimal calcium intake. Only a few Egyptian and Turkish children had an adequate calcium intake. Likely, a pure calcium insufficiency was not the sole cause of rickets in some of our patients (up to 21% in Egyptian patients and up to 8% in Turkish patients). By contrast to patients with calcium deficiency rickets in whom serum 1,25(OH)2D levels are elevated (14, 17, 24, 31), our patients had normal serum levels, and calcium deficiency rickets has been usually observed at an older age than that of our patients (12, 14, 16).
Definition of vitamin D deficiency has varied, depending on the study and the age of patients, but it has been shown that serum 25(OH)D levels less than 15 ng/ml (37.5 nmol/liter) are usually associated with rickets (25). Vitamin D deficiency may be exacerbated further by increased catabolism of 25(OH)D as a consequence of secondary hyperparathyroidism due to calcium deficiency/insufficiency (3, 24). The negative correlation of serum 25(OH)D levels with serum PTH levels and the positive correlation with serum calcium levels support this mechanism in our patients.
Although the definition of children by a different calcium intake standard and/or by a different serum 25(OH)D levels threshold could affect their stratification, our results suggest that Egyptian and Turkish patients have, at least in part, different mechanisms causing rickets (3, 24). In fact, the stratification of patients showed that vitamin D deficiency, pure or associated with calcium insufficiency, was found to be frequent in Turkish patients (up to 86%), whereas in Egyptians the most frequent situation (50%) was defined as calcium insufficiency associated with vitamin D deficiency. The cause of rickets in 8% of Egyptian and 5% of Turkish patients showing a vitamin D sufficiency with a suboptimal calcium intake and in only one Turkish patient with vitamin D sufficiency and adequate calcium intake was not clearly defined. This may suggest that additional factors, including genetic factors, may have a key role in the pathogenesis of the disease in some patients.
The higher BAP levels, despite the lower osteocalcin levels, were compatible with an arrest of the osteoblast phenotype in the developmental phase immediately preceding matrix mineralization induced by vitamin D deficiency (32) and/or calcium insufficiency according to the positive correlation between serum calcium and osteocalcin levels. Severity of rickets was predicted by serum BAP, phosphate, and ICTP levels, reflecting the development of secondary hyperparathyroidism (32).
Although social and nutritional factors may explain the high prevalence of rickets in the Middle East, additional genetic factors, interacting with various degrees of vitamin D deficiency, and/or calcium insufficiency may increase predisposition in some children. An Egyptian study reported some differences in palmar dermatoglyphics between rachitic infants and controls (33). Whereas the evolutionary context of VDR polymorphism has not been investigated so far, haplotypes in the VDR locus have been shown to have significant phenotypic association (34). The prevalence of the F allele was increased and that of the f allele was decreased in rickets against controls, in agreement with findings in Nigerian children with calcium deficiency rickets (35). The F allele confers a transcriptionally somewhat more efficient VDR (36), and its increased prevalence suggests an evolutionary adaptation to a vitamin D and calcium insufficient environment or lifestyle (37). In Egyptian patients, serum 1,25(OH)2D levels and 1,25(OH)2D to 25(OH)D ratio were lower in FF homozygotes, as expected to occur in subjects with a more effective VDR. In healthy adolescents, a greater calcium absorption was found in FF homozygotes, compared with those of ff homozygotes and Ff heterozygotes (38); however, the positive effect of the FF genotype is limited whether dietary calcium is severely restricted (39). Egyptian controls also had a poor dietary calcium intake, but it was higher, compared with that of patients, suggesting that it is the interaction of VDR polymorphism with reduced calcium intake and vitamin D status that could determine the individual susceptibility to developing rickets.
In addition to the VDR F allele effect, we showed that the VDR B allele may predispose an individual to vitamin D deficiency, whereas the b allele was more frequent in patients with reduced calcium intake; moreover, rickets with the homozygous BB genotype was more severe. Homozygous BB genotype association with lower serum 25(OH)D levels was evident in both patients and controls; in controls, but not patients, lower serum 25(OH)D levels were associated with an increased 1,25(OH)2D to 25OHD ratio, suggesting that vitamin D status may be regulated by VDR polymorphism. Associations between VDR genotypes and bone and mineral metabolism have been described, in agreement with results found in control and rachitic children; in BB individuals, lower bone mineral density (40, 41) and lower 1,25(OH)2D and higher PTH and BAP in hemodyalized BB patients, compared with the bb group (42), had been reported. Moreover, dietary calcium intake was positively associated with bone mineral density only among persons with the bb genotype (40). In our study, homozygous BB and heterozygous Bb patients were associated mainly with pure vitamin D deficiency and combined vitamin D deficiency with calcium insufficiency, in agreement with reports of apparently greater than usual doses of vitamin D, associated with calcium supplementation, to cure rickets in children of the Middle East (13, 15).
In Mongolian children with rickets, frequencies of VDR genotypes for polymorphisms in intron 8 (BsmI) and exon 9 (TaqI), which are in linkage disequilibrium, were similar in patients and controls, as in our study for BsmI (43).
In conclusion, our study demonstrates that rickets in the Middle East is determined by nutrition but also by the environment, lifestyle, and genetic predisposition. As such, it should be not labeled pure nutritional rickets. The rachitic children suffer mostly from calcium insufficiency combined with vitamin D deficiency in Egypt and from vitamin D deficiency associated with insufficient mother and child care in Turkey.
A nonnegligible percentage of controls of both countries had pure vitamin D deficiency and/or calcium insufficiency, which may suggest that other factors are involved in the pathogenesis of rickets.
Although little is known about the evolutionary adaptive changes of the VDR genotypes, its polymorphisms may predispose to rickets in the given environs and nutrition. The unique environs and genetic predisposition have to be accounted for in the design of preventive measures rather than using European or American RDI for calcium and vitamin D.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online February 19, 2008
1 * G.I.B., A.B., M.E.K., and L.A. contributed equally to this paper, and all four are to be regarded as first author. 
Abbreviations: BAP, Bone alkaline phosphatase isoenzyme; ESPE, European Society for Pediatric Endocrinology; ICTP, carboxyterminal telopeptide of type I collagen; 1,25(OH)2D, 1,25-dihydroxyvitamin D; 24,25(OH)2D, 24,25-dihydroxyvitamin D; 25(OH)D, serum 25-hydroxyvitamin D; PINP, propeptide of type I procollagen; RDI, recommended dietary intake; VDR, vitamin D receptor.
Received June 25, 2007.
Accepted February 12, 2008.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
M, Aksoy H, Tan H, Akda
R 1999 Incidence of rickets among 0-to-3 years-old children in Erzurum. & Ccedil;ocuk Sal ve Hastalklar Dergisi 42:389–396
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, P < 0.001 vs. Bb; 
, P < 0.05 vs. BB; **, P < 0.02 vs. Ff.