Bone mineral density in short bowel syndrome : correlation with BMI and serum vitamins C , E and K

Objective: Bone loss has been established as a major extra-intestinal complication of short bowel syndrome (SBS). The purpose of this study was to correlate bone mineral density (BMD) with body mass index (BMI), serum vitamin and mineral levels in patients with SBS. Material and methods: The study was conducted on 13 patients (8 male and 5 female, 54.7 ± 11.4 years) with SBS (residual small bowel length of 10 to 100 cm). We determined the food ingestion, anthropometry, serum levels of vitamins C, A, D, E and K, as well as serum and urinary levels of phosphorus and calcium. BMD was measured by dual-energy x-ray absorptiometry (DXA). Results: Osteopenia and osteoporosis was diagnosed in all but one SBS patient. Serum levels of vitamin D were low in all volunteers. Sixty-one percent of patients had vitamin E deficiency; hypovitaminosis A and C occurred in one subject. BMI and C, E and K vitamin serum levels correlated with T-score of BMD. Conclusions: Osteopenia and osteoporosis were common in SBS patients. There was a correlation between BMD and the serum levels of vitamins C, E and K, an indicative that such vitamins may influence bone health. Arch Endocrinol Metab. 2015;59(3):252-8


INTRODUCTION
S hort bowel syndrome (SBS) may result from a state of malabsorption secondary to extensive intestinal resection, causing disturbances in the homeostasis of flu ids and electrolytes, chronic diarrhea and weight loss (1).The nutritional therapy aims to replace nutrients and pro vide intestinal rehabilitation and their return to social ac tivities, with improved quality of life (2).Although there are several degrees of anatomical and functional adapta tion in the remaining bowel, some patients are unable to keep a satisfactory nutritional status by oral route, thus becoming dependent on parenteral nutrition (PN) (1).Recently, we showed that SBS patients present deficien cies in vitamins E, D (3,4) and K (3), regardless of venous infusion.Within this context, efforts are primarily devo ted to intestinal adaptation to oral route nutrition (5).Pa tients with severe malabsorption need home PN (5), but it is not a therapeutic option provided by the Brazilian public health system.In our service, these patients receive intermittent PN in periodic hospital admissions (3,4).
In the last decades, specialized nutrition therapy techniques resulted in a greater lifetime after extensive intestinal resection, enabling the onset of complications that were usually undiagnosed.In this context, bone health should be remarked, once bone pain and fractures decrease the quality of life in SBS patients.It is widely known that chronic and severe vitamin D deficiency impairs calcium and phosphorus absorption leading to disturbances in bone mineralization.Slight vitamin D insufficiency has been associated with inability to reach a proper peak bone mass and/or to maintain skeletal homeostasis (6).The impact of other nutritional deficiencies on bone health is not well established.Vitamins K and C, phosphorus, magnesium and protein are nutrients essential to bone health (7), although the role of such nutrients has been scarcely documented in SBS patients.Therefore, the present study was designed to evaluate the correlations between bone mineral density and markers of nutritional status in patients with SBS.

MATERIAL AND METHODS
This transversal study was conducted in a public uni versity hospital after approval by the institutional Ethics Committee (Protocol #29/2007) and all subjects signed an informed consent.Thirteen adults with SBS were included in this study, with clinical characteriza tion showed in table 1. Hemodynamically unstable subjects and individuals with intestinal inflammatory diseases, neoplasia, hepatic failure, diagnosis of active infectious diseases and use of any medications that may impact bone health were excluded from the study.
Among the 13 SBS patients, seven were fed orally.They had received PN during immediate postoperative, but this approach was interrupted after 2 to 6 months due to satisfactory evolution.Six patients were PN de pendent at the time of the study and they were admitted to the hospital to receive intermittent PN, according to the schedule established by the Medical Nutrition Divi sion.During each admission cycle, the patients received PN for 58 days, with an interval of 1040 days between hospitalizations, depending on diarrhea, dehydration, and nutritional status.PN was infused through a totally implantable central venous catheter.The PN mixture contained dextrose (Hiplex ® , Fresenius Kabi, Campi nas, Sao Paulo, Brazil), lipid emulsion (Lipofundin ® MCT/LCT, B. Braun, Melsungen, Germany), crys talline amino acid solutions (Soramin ® , Darrow, Rio de Janeiro, Brazil), electrolytes, vitamins (Cerne ® , Baxter, Florida, USA), and trace elements (Adelement ® , Dar row, Rio de Janeiro, Brazil), according to the current recommendations (8).All patients were followed by a specialized outpatient service, where they were advised to go on an oral diet at home and to take tablets (Cen trum ® , Wyeth, Richmond, VA, USA) containing vita mins and minerals on a daily basis.The composition of the vitamin, mineral and electrolytes prescribed for SBS patients is presented in table 2.  Anthropometric measurements were done using standard techniques.Body impedance analysis was performed using a bioelectrical impedance analyzer (BIA 101A Detroit, MI, USA) after overnight fas ting.Analysis of the composition of the habitual diet was based on the SemiQuantitative Food Frequency Questionnaire, which assesses the food intake for the previous 6 months.The dietetic data was processed by specific software (NutWin 1.5 ® Professional Software; Federal University of São Paulo, São Paulo, Brazil).
Serum phosphorus, as well as urinary calcium and phosphorus were analyzed by atomic absorption flame spectrophotometry.Calcium serum levels were deter mined by colorimetry.Ascorbic acid was measured by colorimetric reaction with 2,4dinitrophenylhydrazine and spectrophotometric detection.Determination of vi tamin D 3 , retinol, and αtocopherol was accomplished by ultraviolet high performance liquid chromatography (HPLCUV), whereas vitamin K was measured on an HPLC device equipped with an electrochemical detector.
BMD was measured by dualenergy xray absorptio metry (DXA) using a Hologic QDR 4500A scanner ® (Hologic Inc, Waltham, MA) in the following sites: to tal femur, femoral neck and lumbar spine.The results were expressed as TScore and analyzed in accordance to the World Health Organization (WHO).Data are reported as mean ± SDs, and by the me dian and range.For correlation between variables, data were analyzed by Spearman's test.Statistical significan ce was set at p < 0.05.Data analyses were performed with the Statistica software (version 8.0, StatSoft Inc, Tulsa, Oklahoma).

RESULTS
Four patients presented BMI below 18.5 kg/m 2 and re duced fat and lean body mass (Table 3).The subjects had an adequate intake of nutrients related to bone health, except for calcium (Table 3).However, considering the addition of daily mineral and vitamin supplement tablets, all nutrients exceeded the nutritional recommendations.Table 4  Bone mineral density in short bowel syndrome mal BMD; osteopenia was observed in 6 patients (46%) and osteoporosis in 6 individuals (46%).There was no correlation between the Tscore and variables such as patient age, the length of remaining small bowel and the time elapsed after intestinal resection.
There were positive correlations between Tscore and BMI in the three study sites (Figure 1), although we could not establish any correlations between Tscore and lean/fat body masses.Moreover, there were no correlations between Tscore and serum lev els of vitamins D and A in the three study sites.Serum vitamins C (Figure 2) and E (Figure 3) associated with the Tscores of neck and total femur.Vitamin K serum levels correlated with neck femoral Tscore (Figure 4).There was no correlation between Tscores and serum and urinary calcium and phosphorus.

DISCUSSION
The frequency of alterations in bone mass observed in the present study was greater than those documented in previous investigations.Osteomalacia occurred in 36% of SBS subjects even with normal calcium and phos phorus serum levels (9) and with daily vitamin D sup plementation (10).Bone disease occurred in 67% (11) and 84% (12) of patients who were receiving PN over a long period.Osteoporosis was documented in 46% in the lumbar spine and 12.7% in the femural head; os Arch Endocrinol Metab.2015;59/3 Bone mineral density in short bowel syndrome teopenia occurred in 18% in the lumbar spine and 18% in the femoral head among patients receiving PN (13).Disturbances in nutrient absorption may be partially responsible for the high frequency of osteoporosis/ osteopenia.Even after small intestinal resections (jeju noileal bypass) there was a decrease in the serum levels of calcium, phosphorus, 25OH vitamin D 3 , and 1,25 (OH)2 vitamin D 3 (10) .Shortterm PN was associated with an improvement of bone metabolism, which evol ved from a hyperkinetic turnover pattern to a positive balance on bone remodeling, whereas longterm PN was associated with decreased bone formation (12).Except for patients whose intestinal disease started prior to reaching the peak bone mass, longterm PN had no deleterious effect on cortical bone and impro ved trabecular bone (11).Metabolic acidosis, excessive aminoacid infusion, insufficient vitamin D and mag nesium supply are some other factors ascribed to the onset of bone disease (13).Vitamin D is not effective in increasing calcium absorption in patients with intes tinal failure, and its excess may be associated with bone disease (14).Hypomagnesemia is common in SBS pa tients, and may decrease the parathormone secretion and action, resulting in an insufficient production of 1,25hydroxyvitamin D, thus contributing to lower calcium bone storage (15).
Although we did not measure the serum levels of PTH, it can be hypothesized that secondary hyperpa rathyroidism kept circulatory calcium levels within the normal range (16) in our SBS patients.Hypercalciuria (observed in 3 of our SBS patients) cannot be ascribed to excessive supply of calcium, protein (15) or glucose and sodium (17).It is possible that deficiencies of vita mins D and K, observed in our patients may have been responsible for increased bone reabsorption and impai red calcium incorporation into bone, thus resulting in hypercalciuria.
In this study, there was a positive correlation between BMI and Tscore obtained from bone densitometry.Metaanalysis by Laet and cols.(18) showed nearly twofold increase in risk ratio for hip fracture in adults with a BMI of 20 kg/m² when compared with others with 25 kg/m².Subjects with low BMI present reduced muscle mass and nutritional deficiencies, which implies in a substantial risk for all fractures, largely indepen dent on age and gender, but dependent on BMD (18).Furthermore, women with lower body weight present reduced peripheral conversion of gonadal hormones, resulting in deficiency of estrogens and its consequent adverse effects on bone (19).Our data are in accordan ce with those of Raman and cols.(13), who did not find any correlations between Tscore and venous su pply of aminoacid, calcium, magnesium and fosphate, apart from the duration of PN.In women undergoing PN, multiple regression showed negative correlation between Zscore and age when PN was initiated (20).
No correlation was found between bone density and serum levels of vitamin D 3 , a finding which is in contrast with that observed in others studies (10,17).In our study, all patients presented low serum levels of vitamin D, which makes it impossible to perform cor relation analysis of Tscores.On the other hand, there was correlation between BMD and serum levels of vita mins C and K. Vitamin C takes part in the process of ly sine and proline hydroxylation, necessary to form bone collagen triple helix structure (21).Apart from that, as corbic acid stimulates alkaline phosphatase, which plays a role in the synthesis of type I collagen matrix, in mi neralization and in the expression of several osteoblast proteins (22).It has been documented since the 1980s that patients with vertebral compression fractures pre sent lower vitamin K serum levels (23).In addition to malabsorption, vitamin supplements used by our SBS patients does not contain vitamin K.The association of vitamin K with BMD has been recently take into con sideration in patients under home PN (24).Vitamin K is essential for osteocalcin production by osteoblasts (25) and the stimulation of bone collagen surplus (26) by reducing calcium urinary excretion and inhibiting the production of PGE2 and IL6, compounds which stimulate bone reabsorption (26).
In the present study, we could also observe a posi tive correlation between BMD and vitamin E serum levels.Tocopherol inhibits epiphyseal cartilage reab sorption and protects the membranes of chondrocytes by reducing free radical generation and lipidic peroxi dation (27).In addition, alphatocopherol may reduce osteoclast formation and bone loss, by inhibiting the induction of the receptor activator of nuclear factorκB (RANK) (28), which is an oxidative stressresponsive transcription factor.
The absence of correlation between BMD and se rum vitamin A is in accordance with the literature, al though researchers have not reached a consensus (29).Melhus and cols.(30) showed that excessive vitamin A supply associates negatively with BMD, even after adjustments in the other risk factors for osteoporosis.Experimental studies showed that vitamin A deficiency Arch Endocrinol Metab.2015;59/3 Bone mineral density in short bowel syndrome relates to alterations in osteoclast and osteoblast activity (31).On the other hand, excessive vitamin A reduces bone collagen synthesis (32).Inadequate vitamin A su pply is associated with compromised bone health, by mechanisms that are still unknown (33).
Limitations of our study include the small sample.However, once SBS does not have a high prevalence, we could consider our sample size as appropriate.Also, we did not perform the measurements of bone health related hormones.Multiple hormones, including sex steroids, are considered regulators of bone metabo lism and have been associated with bone density (34).Considering the scarcity of studies about BMD in SBS patients, this work is relevant for it assesses possible fac tors associated to metabolic bone disease simultaneou sly, including vitamin K.
In conclusion, osteopenia and osteoporosis were common in SBS patients and there was a correlation between BMD and the serum levels of vitamins C, E and K, an indicative that such vitamins may influence bone health.Continuous monitoring of Tscore and serum vitamin levels and supplementation with vita mins C, K and E in SBS patients are recommended for a timely intervention.Further investigations involving larger samples could assess the effectiveness of oral and venous supply of vitamins C, K and E in reducing the progression of bone disease in SBS.
shows serum concentrations of vitamins and serum and urinary concentrations of calcium and phosphorus in SBS patients.All patients presented low vitamin D 3 level.Vitamin E deficiency affected eight patients and hypovita minosis A and C occurred in one subject.Furthermore, low calcium serum levels were documented in five patients, whereas hypercalciuria was observed in three individuals.The Tscore values were: 1.40 (range: 3.70 to 0.20) in the total femur, 1.90 (range: 4.0 to 0.30) in the femoral neck and 2.20 (range: 4.20 to 0.60) in the lumbar spine.Only one patient presented a nor Arch Endocrinol Metab.2015;59/3

Figure 4 .Figure 1 .
Figure 4. Correlation between serum vitamin K and T-score of femoral neck in short bowel syndrome patients.

Figure 2 .
Figure 2. Correlation between plasma vitamin C and T-scores of femoral neck in short bowel syndrome patients.

Figure 3 .
Figure 3. Correlation between plasma vitamin E and T-scores of total femur and femoral neck in short bowel syndrome patients.

Table 1 .
Clinical characteristics of SBS patients

Table 2 .
Supply of micronutrients in oral and venous route * According to the individual requirement.Arch Endocrinol Metab.2015;59/3 Bone mineral density in short bowel syndrome

Table 3 .
Body composition and daily oral nutritional intake in 13 patients with short bowel syndrome BMI: body mass index; M: male; F: female; EAR: estimated average requirement; AI: adequate Intake.

Table 4 .
Serum concentration of vitamins and serum and urinary concentration of calcium and phosphorus in 13 patients with short bowel syndrome