Relationship between vitamin D and post-transplant diabetes mellitus in kidney transplant recipients

Abstract

BACKGROUND

Post-transplant diabetes mellitus (PTDM) adversely affects graft survival and is an independent predictor of adverse cardiovascular events. Observational studies in the general population as well as the post-transplant setting suggest an association between the plasma 25-hydroxyvitamin D [25(OH)D] level and onset of type 2 diabetes mellitus. Literature is very limited in the context of PTDM.

AIM

To study the relationship between vitamin D deficiency at the time of kidney transplant and PTDM in the post-transplant period.

METHODS

In this single center study, 72 patients who underwent kidney transplant were included. Blood samples for serum vitamin D level were collected on the day of transplant and analyzed at the end of study. The LIAISON^® 25(OH)D assay was used for quantitative estimation of total 25(OH)D in the serum. PTDM was diagnosed by either fasting plasma glucose (> 126 mg/dL), 2-h post prandial plasma glucose (> 200 mg/dL), or 2-h oral glucose tolerance test after 45 days post-transplant. Hemoglobin A1c was not used to diagnose PTDM. Vitamin D levels were labeled as sufficient (≥ 30.0 ng/mL), insufficient (20.0-29.9 ng/mL), and deficient (< 20.0 ng/mL). Patients were reviewed at 45 days and 1 year post transplant for the occurrence of PTDM.

RESULTS

In our study cohort 72 patients completed the study. Overall, 32 (43.8%) patients developed PTDM during the follow up of 1 year, 44 (61.1%) patients had deficient (< 20 ng/mL) 25(OH)D levels. Twenty-six (81.2%) patients with PTDM had deficient vitamin D levels as compared with 18 (45.0%) patients without PTDM (P = 0.007). This association was also significant when univariable [odds ratio (OR) = 5.3, 95% confidence interval (CI): 1.79-15.67, P = 0.003)] and multivariable (OR = 8.21, 95%CI: 2.19-30.75, P = 0.002) regression analysis was performed. A higher proportion of subjects having PTDM (15.6%) had a positive family history of diabetes mellitus than the controls (2.5%) (P = 0.045). However, this association did not persist in the multivariable regression analysis (OR = 12.6, 95%CI: 0.86-185.4, P = 0.065).

CONCLUSION

Deficient vitamin D levels (< 20 ng/mL) were significantly associated with PTDM in the post kidney transplant setting. Further studies are needed to see the effect of vitamin D replacement on PTDM.

Key Words: Vitamin D deficiency; Vitamin D insufficiency; Post-transplant diabetes mellitus; Kidney transplant recipients; Diabetes mellitus

Core Tip: In this study a significant number of patients who underwent kidney transplant developed post-transplant diabetes mellitus (PTDM) in the 1-year follow up. This is the first study from a tropical region showing a statistically significant association of vitamin D deficiency and PTDM. This finding is important since PTDM adversely affects graft survival and is an independent predictor of adverse cardiovascular events. It will be interesting to study the effect of vitamin D replacement on PTDM to potentially reduce the morbidity associated with PTDM.

INTRODUCTION

Post-transplant diabetes mellitus (PTDM) is a common metabolic abnormality after kidney transplant, amounting to 10%-45% of recipients[1]. PTDM occurs most commonly in the first year after transplantation but does not affect the patient and graft survival in the initial post-transplant period. However, it does impact long-term outcomes[2]. The advances in immunosuppression have led to improvement in immediate survival in terms of acute rejection. However, the cardiovascular morbidity and mortality and infectious complications continue to effect survival. PTDM is an independent predictor of cardiovascular disease[2]. Valderhaug et al[3] demonstrated that PTDM caused a 1.8-fold increase in cardiovascular mortality and 1.5-fold in overall mortality[3]. Understanding the predictors of PTDM may improve transplant outcomes.

The common risk factors for PTDM include age, high body mass index (BMI, > 30 kg/m²), ethnicity, insulin resistance, family history of type 2 diabetes mellitus, high triglycerides, chronic hepatitis C infection, immunosuppression medication (calcineurin inhibitors, steroids, mammalian target of rapamycin inhibitors), cytomegalovirus (CMV) infection, human leukocyte antigen (HLA) mismatch, diseased donor, and hypomagnesaemia[1,4]. Prevention of PTDM include identification of modifiable risk factors at the time of transplantation. In 2003 the international expert panel suggested the use of the World Health Organization definition for the diagnosis of diabetes mellitus and impaired glucose tolerance for PTDM. The expert panel in 2014 recommended screening for PTDM by postprandial glucose monitoring and hemoglobin A1c (HbA1c)[1].

25-hydroxyvitamin D [25(OH)D] plays an important role in calcium and bone homeostasis[5]. Many in vitro and in vivo studies have described the extraskeletal pleiotropic effects of vitamin D[6,7]. The pleiotropic effects of vitamin D are caused by extra renal expression of cytochrome P450 family 27 subfamily B member 1, cytochrome P450 family 24 subfamily A member 1, and vitamin D receptor[8]. In the general population few observational studies have shown a relationship between 25(OH)D levels and the onset of type 2 diabetes mellitus[4]. A study by Quach et al[9] involving 442 patients in a post-kidney transplant setting showed that serum 25(OH)D is an independent predictor of PTDM in recipients of a kidney transplant[9]. Meta-analysis suggests a 4% lower risk of PTDM for every 4 ng/mL increase in 25(OH)D level[10]. 25(OH)D replacement leads to improvement in beta cell function and delays the onset of diabetes mellitus[11].

There are no randomized controlled trials to evaluate the optimum level of 25(OH)D to improve clinical outcomes, leading to no unequivocal definition of 25(OH)D deficiency. Serum 25(OH)D levels < 20.0 ng/mL, 20.0-29.9 ng/mL, and ≥ 30.0 ng/mL are considered deficient, insufficient and sufficient, respectively[12]. Optimal vitamin D levels may further be different for specific disease conditions. 25(OH)D levels of more than 10 ng/mL are found to be good enough for prevention of rickets and osteomalacia, and 25(OH)D levels more than 30 ng/mL may be optimum for the prevention of secondary hyperparathyroidism or osteoporosis[13].

Low levels of 25(OH)D are common in the general population but are more frequent in chronic kidney disease. About 30% of recipients of a kidney transplant are 25(OH)D deficient, and 81% are 25(OH)D insufficient[14]. Only 12% of recipients of a kidney transplant have 25(OH)D levels more than 30 ng/mL in the first year after transplant[15,16]. The prevalence of vitamin D deficiency may be different in different geographical areas. Data regarding vitamin D deficiency in the kidney transplant setting and PTDM are very scarce. The present study aimed to investigate the prevalence of 25(OH)D deficiency, prevalence of PTDM, and the relationship between them in patients after undergoing kidney transplant in a tropical country. The scope of the study further includes the analysis of various determinants of PTDM including vitamin D deficiency.

MATERIALS AND METHODS

This was a single center cohort study conducted in the department of nephrology at our hospital. The study was approved by the institute ethics committee (approval No. IECPG-534/29.08.2019). Written informed consent was taken from all participants. All patients with kidney failure who underwent kidney transplantation from August 2019 to March 2020 were included in this study. Those with preexisting diabetes mellitus and pediatric kidney transplant recipients were excluded from this study. Of the 99 patients who underwent kidney transplant during this period, 9 patients with diabetes and 7 pediatric patients were excluded from the study. In addition, 5 patients expired during the study period, 3 patients lost their graft, and 3 patients were lost to follow-up. A total of 72 patients completed the study.

Definitions

PTDM was diagnosed per the World Health Organization guidelines and international expert panel recommendation[17]. PTDM was diagnosed by either fasting plasma glucose (> 126 mg/dL), 2-h post prandial plasma glucose (> 200 mg/dL), or 2-h oral glucose tolerance test after 45 days post-transplant. HbA1c was not used to diagnose PTDM due to various issues like high prevalence of anemia in our patients, and altered kidney function can also interfere with HbA1c estimation. Patients who received anti-diabetic medications in the immediate post-transplant period and continued to take them even 2 months after transplant were also classified as PTDM.

Screening for PTDM was conducted at 45 days post-transplant and at 1 year (end of study). Graft outcomes were assessed by blood urea, serum creatinine, urine routine examination, and morning spot urine protein creatinine ratio. The LIAISON^® 25(OH)D assay was used for quantitative estimation of total 25(OH)D in the serum. On the day of transplantation, a 5-mL blood sample was collected and stored in the freezer for vitamin D level estimation at the end of study.

Immunosuppression

All patients received triple drug immunosuppression therapy, which included steroid, tacrolimus, and mycophenolate mofetil. Tacrolimus was started at a fixed dose of 0.1 mg/kg/day, and the subsequent dose was adjusted per therapeutic drug monitoring to achieve desired therapeutic levels. Tacrolimus exposure was comparable in both cohorts. Mycophenolate mofetil was also started at a fixed dose of 30 mg/kg/day, and the dose was reduced at fixed intervals per standard protocols without therapeutic drug monitoring. All patients received a fixed dose of prednisolone, which was subsequently reduced to 7.5 mg per day at 3 months. Patients who required induction therapy received either interleukin-2 receptor antagonist (basiliximab) or rabbit-antithymocyte globulin per the immunological risk profile of the recipient. The majority of patients received no induction other than methylprednisolone pulse. Some patients received basiliximab injection 20 mg on day 1 and day 4. Few patients received rabbit-antithymocyte globulin at a dose of 3 mg/kg over 3 days. The patients were followed up prospectively, and demographic and clinical data were recorded including recipient’s age, blood group, native kidney disease, form of dialysis before transplant, dialysis vintage, induction agent, serum calcium, serum phosphate, and spot urine protein creatinine ratio. Details of anti-rejection therapy, CMV infection, and hepatitis C virus (HCV) infection was also noted. Donor details including age, sex, relation, blood group, and HLA mismatch were recorded. Patients were screened for the development of PTDM at 45 days and at 1 year after transplantation. At the end of study at 1 year when vitamin D levels were available, those who had deficient or insufficient levels were offered treatment.

Statistical analysis

Data were analyzed by STATA 14. Categorical variables were expressed as n (%). Quantitative variables that followed normal distribution were expressed as mean ± SD. Those variables that did not follow normality were expressed as median (minimum-maximum). Independent t test rank-sum test was used to compare quantitative variables between two cohorts as appropriate. Categorical variables were analyzed using χ² and Fisher exact tests. Univariable odds were calculated. Multivariable logistic regression was carried out to find independent risk factors for PTDM. Those variables that were found to be statistically significant in univariable analysis (P < 0.10) and clinically important were included in the logistic regression. P < 0.05 was considered statistically significant.

RESULTS

A total of 72 patients were included in the study for analysis. Among the 72 patients, 32 (43.8%) patients developed PTDM during the follow-up of 1 year (Table 1). Twenty-three (71.9%) patients required oral hypoglycemic drugs while 9 (28.1%) patients required insulin therapy.

Table 1 Baseline characteristics of the patients.

	Total	PTDM	Controls	P value
Total patients	72	32	40
Male gender	61 (84.7)	30 (93.7)	31 (77.5)	0.057
Age (years), mean ± SD	32.4 ± 9.2	34.6 ± 9.6	30.7 ± 8.5	0.076
> 40	55 (76.3)	34 (85.0)	21 (65.0)	0.092
< 40	17 (23.6)	6 (15.0)	11 (35.0)
Dialysis vintage, months, median (minimum-maximum)	17.0 (12.0-25.5)	14.5 (10.0-24.5)	17.0 (13.0-27.0)	0.189
BMI, kg/m2, mean ± SD	19.9 ± 3.8	20.5 ± 3.7	19.4 ± 3.9	0.266
BMI grading				0.338
0	28 (38.8)	9 (28.1)	19 (47.5)
1	26 (36.1)	19 (59.3)	17 (42.5)
2	5 (6.9)	3 (9.3)	2 (5.0)
3	3 (4.1)	1 (3.1)	2 (5.0)
BMI grading at 12 months
0	20 (27.7)	7 (21.8)	13 (32.5)	0.294
1	44 (61.1)	21 (65.6)	23 (57.5)
2	6 (8.3)	4 (12.5)	2 (5.0)
3	2 (2.7)	0	2 (5.0)
Recipient blood cohort
O	16 (22.2)	7 (21.8)	9 (22.5)	0.371
AB	5 (6.9)	1 (3.1)	4 (10.0)
A	20 (27.0)	7 (21.8)	13 (32.5)
B	30 (41.6)	17 (53.1)	13 (32.5)
Native kidney disease
Unclassified	43 (59.7)	16 (50.0)	27 (67.5)	0.714
Chronic TID	13 (18.0)	7 (21.8)	6 (15.0)
IgA	9 (12.5)	5 (15.6)	4 (10.0)
FSGS	4 (5.5)	2 (6.2)	2 (5.0)
MPGN	1 (1.3)	1 (3.1)	0
ADPKD	2 (2.7)	1 (3.1)	1 (2.5)
Family history of DM	6 (8.3)	5 (15.6)	1 (2.5)	0.045
History of smoking	2 (2.7)	2 (6.2)	0	0.113

Data are presented as n (%). PTDM: Post-transplant diabetes mellitus; BMI: Body mass index; TID: Tubulointerstitial disease; IgA: Immunoglobulin A nephropathy; FSGS: Focal segmental glomerulosclerosis; MPGN: Membranoproliferative glomerulonephritis; ADPKD: Autosomal dominant polycystic kidney disease; DM: Diabetes mellitus; SD: Standard deviation.

Demography

Patients were divided into two groups: With and without PTDM. There was a trend towards a male predisposition in patients with PTDM (n = 30, 93.7%) as compared with patients without PTDM (n = 31, 77.5%) (P = 0.057). In the univariable analysis by logistic regression, female gender appeared to have non-significant protection from PTDM compared with males [odds ratio (OR) = 0.23; 95% confidence interval (CI): 0.04-1.30, P = 0.105]. A similar trend was seen in multivariable analysis (OR = 0.21, 95%CI: 0.04-1.17). Our study showed a trend towards relative protection of the female sex and a male predisposition for the development of PTDM. The mean age of the patients was 32.4 ± 9.2 years. It was comparable in patients with PTDM (34.6 ± 9.6 years) and without PTDM (30.7 ± 8.5 years) (P = 0.076). The median duration of dialysis vintage was also comparable in patients with PTDM (14.5, 10.0-24.5 months) and patients without PTDM (17.0, 13.0-27.0 months, P = 0.189).

There was no statistical difference in the BMI of the patients in the two groups (Table 1). Most patients with PTDM (n = 19, 59.3%) and without PTDM (n = 17, 42.5%) had normal BMI. Even at 12 months post-transplant there was no significant difference in the BMI in the two groups. Blood group B was the most common blood group seen in 17 (53.1%) patients with PTDM and 17 (32.5%) patients without PTDM. The most common native kidney disease was unclassified, and it was comparable in patients with PTDM (n = 16, 50.0%) and patients without PTDM (n = 27, 67.2%) (P = 0.714). The most common biopsy finding among the patients who underwent kidney biopsy was immunoglobulin A nephropathy, and it was comparable in patients with PTDM (n = 5, 15.6%) and without PTDM (n = 4, 10.0%) (P = 0.714). Family history of diabetes mellitus was significantly associated with the incidence of PTDM in recipients of a kidney transplant. About 15.6% of patients who developed PTDM had a family history of diabetes mellitus as compared with 2.5% in controls (P = 0.045) (Table 1). In the univariable regression analysis, a family history of diabetes mellitus showed a trend towards association with PTDM (OR = 6.96; 95%CI: 0.77-62.9, P = 0.084). Similar results were also obtained in the multivariable regression analysis (OR = 12.6; 95%CI: 0.86-185.4, P = 0.065). Most donors were live donors (n = 66, 91.6%, Table 2).

Table 2 Donor and immunology details.

	Total	PTDM	Control	P value
Total patients	72	32	40
Type of donor
Live donor	66 (91.6)	29 (90.6)	37 (92.5)	0.858
Deceased donor	6 (8.3)	3 (9.3)	3 (7.5)
Distribution of live donor
Parent	42 (58.3)	17 (53.1)	25 (62.5)	0.865
Spouse	18 (25.0)	9 (28.1)	9 (22.5)
Sibling	3 (4.1)	2 (6.2)	1 (2.5)
Unrelated	3 (4.1)	1 (3.1)	2 (2.5)
Number of HLA mismatch
1	3 (4.1)	2 (6.2)	1 (2.5)	0.304
2	11 (15.2)	5 (15.6)	6 (15.0)
3	33 (45.8)	11 (34.3)	22 (40.0)
4	6 (8.3)	3 (9.3)	3 (7.5)
5	7 (9.7)	3 (9.3)	4 (10.0)
6	6 (8.3)	5 (15.6)	1 (2.5)
High PRA	4 (5.5)	2 (5.0)	2 (5.0)	0.815
Induction agent
No induction	35 (48.6)	13 (40.6)	22 (55.0)	0.297
Basiliximab	26 (36.0)	12 (37.5)	14 (35.0)
ATG	11 (15.2)	7 (21.8)	4 (10.0)

Data are presented as n (%). PTDM: Post-transplant diabetes mellitus; HLA: Human leukocyte antigen; PRA: Panel reactive antibody; ATG: Antithymocyte globulin.

We found that 11 (34.3%) transplant recipients with PTDM as compared with 22 (40%) without PTDM were haplo-identical with their donors. There was no significant difference in the HLA mismatch in the two groups (P = 0.304). Most of the patients received no induction other than methylprednisolone pulse. Thirteen (40.6%) patients with PTDM and 22 (55%) without PTDM received no induction. Seven (21.8%) patients with PTDM received antithymocyte globulin induction as compared with 4 (10.0%) patients receiving antithymocyte globulin induction, but the difference was not significant (P = 0.297). The number of acute rejection episodes and subsequent use of methylprednisolone was not statistically different in patients with and without PTDM. There were 9 (22.5%) acute rejections in patients without PTDM as compared with 3 (9.3%) in patients with PTDM (P = 0.138). The use of methylprednisolone pulse was also higher in the control group (n = 8, 20%) as compared with the PTDM group (n = 2, 6.2%), but the results were not statistically significant (P = 0.094). There were a greater number of CMV infections in the PTDM group (n = 4, 12.5%) as compared with the control group (2, 5%), but the difference was not significant (P = 0.253). The rate of HCV infection was comparable in the control group (n = 7, 17.5%) and PTDM group (n = 6, 18.7%) (P = 0.891).

Vitamin D deficiency and PTDM

Low serum vitamin D levels at the time of transplant were associated with subsequent development of PTDM. We classified the serum vitamin D levels according to standard definitions. In patients with PTDM sufficient, insufficient, and deficient vitamin D levels were seen in 2 (6.2%), 4 (12.5%), and 26 (81.2%) recipients, respectively, as compared with controls in which 8 (20.0%), 14 (35.0%), and 18 (45.0%) patients were sufficient, insufficient, and deficient, respectively (P = 0.007) (Table 3). After the regression model statistical significance of association between low vitamin D level (< 30 ng/mL) and PTDM was lost, but there was a trend towards a positive association (P = 0.065) (Table 4). However, when we analyzed the relationship between vitamin D deficiency (vitamin D level < 20 ng/mL) excluding insufficiency with PTDM using both univariable (P = 0.002), and multivariable (P = 0.003) regression analysis, the association was significant.

Table 3 Patient outcomes and lab parameters during follow-up.

	Total	PTDM	Controls	P value
Number of cases	72	32	40
Duration of follow-up (weeks)	46.9 ± 4.8	47.9 ± 0.2	46.1 ± 6.3	0.116
Acute rejection	12 (16.6)	3 (9.3)	9 (22.5)	0.138
MP pulse given	10 (13.8)	2 (6.2)	8 (20.0)	0.094
CMV infection	6 (8.3)	4 (12.5)	2 (5.0)	0.253
HCV infection	13 (18.0)	6 (18.7)	7 (21.0)	0.891
Serum vitamin D levels
> 30 ng/mL (sufficient)	10 (13.8)	2 (6.2)	8 (20.0)	0.007
20-30 ng/mL (insufficient)	18 (25.0)	4 (12.5)	14 (43.7)
< 20 ng/mL (deficient)	44 (61.0)	26 (81.2)	18 (56.2)
Serum calcium levels (mg/dL)	8.8 ± 0.08	9.0 ± 0.70	8.7 ± 0.60	0.080
Serum phosphate levels (mg/dL)	2.9 ± 0.7	3.1 ± 0.9	2.8 ± 0.5	0.231
Proteinuria(g/dL), median (minimum-maximum)	0.3 (0.2-0.4)	0.5 (0.3-0.7)	0.2 (0.2-0.3)	0.023
Serum creatinine (mg/dL), median (minimum-maximum)	1.2 (0.6-7.0)	1.3 (1.0-1.5)	1.1 (1.0-1.7)	0.367

Data are presented as n (%) or mean ± SD. PTDM: Post-transplant diabetes mellitus; MP: Methylprednisolone; CMV: Cytomegalovirus; HCV: Hepatitis C virus.

Table 4 Multivariable logistic regression for post-transplant diabetes mellitus.

Variable	Univariable OR (95%CI)	P value	Multivariable OR (95%CI)	P value
Female	0.22 (0.04-1.10)	0.067	0.28 (0.04-1.92)	0.194
Family history of diabetes mellitus	6.96 (0.77-62.93)	0.084	12.60 (0.86-185.40)	0.065
Vitamin D levels < 30 ng/mL	3.74 (0.73-19.09)	0.065	4.65 (0.75- 28.80)
Vitamin D levels < 20 ng/mL	5.30 (1.79-15.67)	0.003	8.21 (2.19-30.75)	0.002
Dialysis vintage	0.99 (0.97-1.02)	0.568	NA
Age > 40 years	3.24 (1.06-9.93)	0.040	3.96 (0.92-16.93)	0.064
Serum Ca2+	1.80 (0.92-3.51)	0.087	3.07 (1.22-7.72)	0.017
Induction therapy
Basiliximab	1.57 (0.57-4.36)	0.385	NA
ATG	2.96 (0.73-12.09)	0.130	NA

OR: Odds ratio; CI: Confidence interval; ATG: Antithymocyte globulin; NA: Not available.

DISCUSSION

Among the 72 patients who completed the study, 32 (43.8%) patients developed PTDM during the follow-up of 1 year. There was a trend towards male predisposition in patients with PTDM. BMI was comparable in both groups at the start and end of the study. The median duration of dialysis was 14.5 month in both groups. The most common native kidney disease was unclassified. Among those who underwent native kidney biopsy, immunoglobulin A nephropathy was the most common diagnosis. A higher proportion of subjects with PTDM (15.6%) had a positive family history of diabetes mellitus than the controls (2.5%). However, this association did not persist in multivariable regression. Most donors were live donors (91.6%). Both groups had comparable HLA matching and induction therapy. HCV and CMV infections were also comparable in the two groups. Acute rejections and methylprednisolone pulse use was similar in both groups. Vitamin D deficiency was quite common: Seen in 81.2% of patients with PTDM as compared with 45.0% in controls. There was a clear relationship between vitamin D deficiency (vitamin D level < 20 ng/mL) with PTDM using both univariable (P = 0.002), and multivariable (P = 0.003) regression analysis. There was a trend towards a positive association (P = 0.065) between low vitamin D level (< 30 ng/mL) and PTDM.

The overall prevalence of PTDM in our study was 43.8%. There was a trend towards a male predisposition in patients with PTDM. The mean age of the patients was 32.4 ± 9.2 years and was comparable in the two groups. In a similar study by Dedinska et al[18], 167 kidney transplant recipients enrolled in the study, and PTDM was observed in 64 (38.3%) patients[17]. Patients in the PTDM group were older with a mean age of 50 ± 9.6 years as compared with the controls (43 ± 12.0 years, P = 0.001). Patients enrolled in our study had a similar prevalence of PTDM and were relatively younger. In our study more patients with PTDM (15.6%) had a family history of diabetes mellitus as compared with controls (2.5%) (P = 0.045). This finding was also seen by Dedinska et al[18] who documented a much higher positive family history of diabetes in 78% of patients with PTDM.

Vitamin D deficiency was found to be the most important factor associated with the incidence of PTDM in our study. The majority of patients with PTDM were vitamin D deficient. However, in the multivariable analysis by cox regression, a significant association of vitamin D levels with the incidence of PTDM was seen only with deficient levels (< 20 ng/mL) and not with insufficient levels (< 30 ng/mL). There are very few other studies in the literature on this subject. In the study by Le Fur et al[19] with a total of 444 patients, 58 (13%) patients had developed PTDM. Multivariable cox analysis indicated an increased incidence of PTDM in patients with vitamin D deficiency (< 10 ng/mL) (P = 0.048), but serum vitamin D levels < 30 ng/mL only showed a trend towards an association (P = 0.073). Another study by Chaudhary and Patil[20] including 468 subjects found a 12.6% incidence of PTDM. This study showed a positive correlation between vitamin D deficiency (< 10 ng/mL) and incidence of PTDM (P < 0.001). Unlike these studies, our study was able to establish a clear association with a higher cutoff vitamin D level (< 20 ng/mL) and PTDM and a trend towards association with a vitamin D level < 30 ng/mL. A larger sample size may have shown a significant relationship even with levels < 30 ng/mL.

Our study consolidated the available evidence and generated fresh evidence from a tropical country where vitamin D deficiency is less rampant due to adequate sun exposure. This evidence supported and highlighted the positive association and possible therapeutic intervention to prevent development of PTDM and reduce the morbidity associated with this condition in a post-transplant setting.

Strengths and limitations

Strengths of the study included the prospective nature of the study, and sufficient duration of follow-up to assess the development of PTDM and exclude transient glycemic fluctuation due to immunosuppression medication in the immediate post-transplant period. Limitations of the study included the single center design, small sample size, and single blood sampling for vitamin D level estimation at the time of transplant. It also lacked an intervention arm in which supplementation of vitamin D and its subsequent effect on the development of PTDM could have provided high quality evidence for direct causal association between vitamin D deficiency and the development of PTDM.

CONCLUSION

PTDM is one of the common complications seen in the post-transplant period. PTDM had a high prevalence of 43.8% of our kidney transplant recipients. Vitamin D deficiency was also quite common in the post-transplant setting. Deficient and insufficient vitamin D levels were found in 61.1% and 25.0% of patients, respectively. Low vitamin D levels < 20 ng/mL were significantly associated with the development of PTDM. A larger sample size will be more useful for assessing the strength of the association of low vitamin D levels (< 30 ng/mL) with PTDM in kidney transplant settings. It will be interesting to see the effect of vitamin D replacement at the time of transplant and its effect on the prevalence of PTDM. This study strengthened the available evidence for the role of vitamin D deficiency in the development of PTDM and suggested a possible therapeutic intervention.

ACKNOWLEDGEMENTS

We acknowledge the contribution of Dr Ashish Datt Upadhyay for statistical analysis.