Association of low penetrance vitamin D receptor Tru9I (rs757343) gene polymorphism with risk of premenopausal breast cancer.

Objective The aim of this study was to determine whether a novel polymorphism ( Tru9I) in the low penetrance vitamin D receptor (VDR) gene is associated with risk of premenopausal breast cancer (BC). Methods This case-control study included 228 patients with BC and 503 healthy women living in Pakistan who were analyzed for the VDR Tru9I (rs757343) single nucleotide polymorphism. BC cases were histopathologically confirmed, and all healthy controls were age-matched with patients (age range, 20-45 years). DNA was extracted, and the polymerase chain reaction and restriction fragment length polymorphism assays were performed. Results The VDR Tru9I polymorphism was not significantly associated with premenopausal BC. However, the risk of BC was associated with the 'uu' genotype (odds ratio [OR], 1.141; 95% confidence interval [95% CI], 0.206-6.317). Further, mutant Tru9I was significantly associated with Grade IV carcinoma (OR, 5.36; 95% CI, 1.181-24.338). Conclusion The VDR Tru9I 'uu' genotype may increase the risk of premenopausal BC.

Introduction

Breast cancer (BC) is the most prevalent cancer and a major cause of death among women worldwide. In Pakistan, the incidence of BC is approximately 34.6% in women.[1] Among risk factors, lifestyle, alcoholism, high birth weight, and abdominal adiposity intensify the risk of BC.[2-4] Diet may modify the incidence of BC,[5] although many dietary components such as vitamin D are categorized as “Limited or no conclusion.”[6] Vitamin D plays a major role in mineral homeostasis;[6] however, evidence indicates that its deficiency exacerbates the risk of BC.[7,8] Moreover, higher vitamin D levels are weakly associated with the low BC risk”.[9,10]

Vitamin D derived from diet and sun exposure of the skin undergoes activation within the liver through 25 hydroxylation, and in the kidney through 1-α hydroxylation, to form 1,25(OH)2D3 (active vitamin D). The classical role of vitamin D is to regulate calcium and phosphate metabolism to maintain bone mineral density (BMD). However, there are wide ranges of nonclassical functions of vitamin D within the body. Vitamin D is activated by forming a complex with the vitamin D receptor (VDR) that is involved in the regulation of many target genes that contribute to cell differentiation, cell growth, programmed cell death, angiogenesis, inflammation, and immune responses.[11,12] Vitamin D, in concert with growth factors, arrests the cell cycle by preventing entry into S phase.[13] Vitamin D induces apoptosis through down-regulation of the antiapoptotic protein B-cell lymphoma 2 (BCL2),[13] inhibits the expression of P-cadherin[14] and increases the expression E-cadherin,[15] which may contribute to the anti-invasive or antimetastatic effects of vitamin D. Further, vitamin D induces the expression of the tumor suppressor BRCA1, because the binding sites for the vitamin D/VDR/retinoid X receptor heterodimer complex are present within the BRCA1 promoter region.[16] Although VDR is present in many body tissues, including normal and cancerous breast tissues, its expression is low in mammary cancers cells,[17] which may be caused by variations in VDR.

There are many polymorphic regions in VDR, among which Fok1, Bsm1, Apa1, and Taq1 are the most extensively studied in American (Hispanic white, nonhispanic white, and Caucasian) and Asians (Iranian and Egyptian) populations.[19-24] Attention focuses on the role of single nucleotide polymorphisms (SNPs) in VDR that are associated with the development of a wide range of pathological conditions, including type 2 diabetes mellitus, prostate cancer, and BC[18-20]. However, limited studies are available on the Pakistani population.[25] Studies of Jordanian populations reveal a statistically significant difference between VDR Taq1 variants and 25(OH)D among women with BC.[21] Bsm1 and Cdx2 polymorphisms are significantly associated with women with BC living in Southeastern Iran.[20] The Fok1 SNP in BC plays a protective role, whereas the Bsm1 SNP is not associated with BC among women living in the Iranian city of Urmia.[22] Among Egyptian women, the Bsm1 “B” allele and “Bb” variants of VDR may represent a risk factor for susceptibility to the development of BC.[23] In contrast, a meta-analysis shows no association of VDR SNPs (Fok1, Bsm1, Taq1, and Apa1) with risk of BC in a general population or in a Caucasian population.[24] Among Pakistanis who are BRCA1/2 non-carriers, the Bsm1 “b” allele is associated with an increased risk of BC”.[25] Although the associations of VDR Fok1 and Bsm1 SNPs with the risk of BC have been investigated, no study, to our knowledge, investigated the relationship between the VDR Tru9I polymorphism and the premenopausal risk of BC. Therefore, the present study was designed detect an association between the novel low penetrance VDR Tru9I SNP and BC among premenopausal women of Pakistan.

Materials and methods

Ethics statement

The study was approved by the Bioethics Committee, Board of Advance Studies & Research, University of Karachi (ethics approval number [10(27) 28032012] and conducted in accordance with the Helsinki Declaration. Each subject provided written informed consent before providing a blood sample.

Study design and study population

The study included 228 patients with BC and 503 control subjects who were premenopausal ethnic Pakistanis aged 20–45 years. The sample size was calculated using OpenEpi (http://www.openepi.com/Menu/OE_Menu.htm) with 80% statistical power and 95% confidence intervals (95% CIs). Patients with BC, who were randomly selected from two tertiary hospitals in Sindh, Pakistan from 2012 to 2015, were histopathologically diagnosed. The medical information (types, grades, estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2, and TNM classification) of patients with BC was obtained from histopathological reports and hospital records. All healthy volunteers were age-matched (1:2) to premenopausal healthy subjects acquired from the general population. Whole blood (3 mL) was collected from each subject. We used a self-structured questionnaire to acquire along with information about personal demographics, detailed history, and BC-related risk factors.

DNA isolation

DNA was extracted using a DNA isolation kit (Gene JET Genomic DNA Purification Kit, Thermo Fisher Scientific Baltics UAB, and Vilnius, Lithuania) following the manufacturer’s protocols. Samples were stored at –86°C.

Primer selection and genotyping

VDR Tru9I (rs757343) genotyping was performed using polymerase chain reaction-restriction fragment length polymorphism analysis. The forward primer F: 5’-GCAGGGTACAAAACTTTGGAG-3' and the reverse primer R: 5’-CCTCATCACCGACATCATGTC-3' were used[26] to amplify the isolated DNA. Amplification was performed using a thermal cycler (Veriti; Applied Biosystems, Foster City, CA, USA) in a final reaction volume (50 µL) containing 25 µL of 2X GoTaq Green Master Mix (Promega, Madison, WI, USA), 4 µL of 12 pmol of both primers, 5 µL of DNA, and 12 µL of nuclease-free water. Polymerase chain reaction analysis was performed as follows: 94°C, initial denaturation (5 minutes); 40 cycles of denaturation at 94°C (30 s); annealing at 59°C (30 s), and extension at 72°C (30 s). The final extension was performed at 72°C for 7 minutes. The 177-bp amplicons were digested using 5 units of Mse1 (37°C for 1 hour). Amplicons and restriction fragment length polymorphism fragments, which were separated using 2.5% agarose gel electrophoresis, were detected using ethidium bromide and visualized using a ChemiDoc-It2 imaging system with VisionWorks LS Analysis Software (version 7.1) (UVP, Cambridge, UK). The results were coded as U-u, where the Upper case letter is denoted as the absence and lowercase letter is designated as the presence of the restriction site, respectively. The call and concordance rates of the VDR genotypes were 98% and 100%, respectively. The accuracy of the genotypes was determined by subjecting samples ( 15 cases and 15 controls) to nucleotide sequence analysis.

Data analysis

Statistical analysis was performed using SPSS software version 22 (IBM Corporation, Armonk, NY, USA) with significance indicated by P < 0.05. The Shapiro–Wilk and Kolmogorov–Smirnov tests were used to determine whether the variables were normally distributed. The goodness-of-fit (χ2) test was used to test whether the controls exhibited Hardy–Weinberg equilibrium to determine deviation of their genotype distribution. Comparisons between groups of VDR Tru9I genotypic and allelic distributions were calculated using Person’s χ2 test. Odds ratios (ORs) along with 95% CIs were determined using binary and multinomial logistic regression analysis of the Tru9I SNP to determine the risk of BC. The significance of ORs was determined using Wald’s statistics and defined as P < 0.05.

Results

Characteristics of patients with BC

Patients’ characteristics are presented in Table 1, which shows that the parameters were not normally distributed among patients according to the Shapiro–Wilk and Kolmogorov–Smirnov tests (P < 0.01).

Table 1

Characteristics of patients with BC (N = 228).

Characteristics	Patients
	n (%)
Affected area
Right*	123 (53.9)
Left*	99 (43.2)
Both*	6 (2.6)
Types of BC
Invasive ductal carcinoma (IDC)*	197 (93.4)
Invasive lobular carcinoma (ILC)*	13 (5.7)
Ductal carcinoma in situ (DCIS)*	6 (2.6)
Others*	12 (5.2)
Grades of BC
I*	6 (2.6)
II*	164 (71.9)
III*	45 (19.7)
IV*	7 (3)
Both II &III*	2 (1.3)
None*	4 (1.7)
Immunohistochemical classification
Estrogen receptor status
Positive (0 to +9)*	126 (55.3)
Negative (−1 to −2)*	34 (14.9)
Unknown	68 (29.8)
Progesterone receptor status
Positive (0 to +8)*	114 (5)
Negative (−1 to −2)*	38 (16.6)
Unknown	76 (33.33)
Human epidermal growth factor receptor 2 status
Positive (0 to +3)*	98 (42.9)
Negative (−1 to −2)*	20 (8.7)
Unknown	110 (48.2)
TNM classification
(For invasive cases only)
Size of primary tumor (T stage)
Tx*	2 (0.8)
Tis*	4 (1.7)
T1*	0 (0)
T2*	35 (15.3)
T3*	31 (13.5)
T4*	26 (11.4)
No involvement*	199 (56.5)
Metastasis to regional lymph node
(N stage)
Nx*	6 (2.6)
N0*	23 (10)
N1*	36 (15.7)
N2*	19 (8.3)
N3*	15 (6.5)
No involvement*	129 (56.5)
Distant metastasis (M stage)
Mx*	39 (17.1)
M0*	48 (21.0)
M1*	12 (5.2)
Unknown	129 (56.5)

BC, breast cancer. *Statistically significant difference (P < 0.05, Shapiro–Wilk or Kolmogorov–Smirnov tests).

Association of the VDR Tru9I SNP with BC

The genotype distributions of Tru9I among patients (228/228) and healthy controls (497/503) are presented in Table 2. The distribution of Tru9I in the control population did not deviate (χ2 = 0.416, P > 0.05) from the Hardy–Weinberg equilibrium. There was no significant difference (χ2 = 0.238) between the VDR Tru9I genotype among patients and controls. However, an 11.2% increase in OR (1.112; 95% CI, 0.202–6.125) was observed among ‘uu’ carriers compared with ‘UU and Uu’ carriers. After adjusting for age, the VDR Tru9I ‘uu’ genotype exhibited a 14.1% increase in the OR (1.141; 95% CI, 0.206–6.317) (Table 2).

Table 2

VDR Tru9I genotype frequencies and associations with the risk of premenopausal BC.

Genotypes	Alternate designation	BC casesn (%)	Controlsn (%)	Χ2	P value	UOR (95% CI)	AOR (95% CI)
GG	UU	179 (78.5)	398 (80)	0.238	0.888	Ref	Ref
GA	Uu	47 (20.6)	95 (19.1)			1.1 (0.744–1.627)	1.085 (0.731–1.611)
AA	uu	2 (0.88)	4 (0.8)			1.112 (0.202–6.125)	1.141 (0.206–6.317)
Total		228	497
GG	UU	179 (78.5)	398 (80)			Ref	Ref
GA+AA	Uu+uu	49 (21.4)	99 (19.6)			1.101 (0.749–1.617)	1.087 (0.737–1.604)
Total		228	497
G allele	U allele	405 (88.8)	891 (89.6)	0.223	0.637	Ref	Ref
A allele	u allele	51 (11.1)	103 (10.3)			1.089 (0.763–1.554)	1.086 (0.76–1.55)
Total		456	994

DNA was amplified from 228/228 patients’ samples and 497/503 control subjects’ samples. BC, breast cancer; UOR, unadjusted odds ratio; AOR, odds ratio adjusted for age and time of blood collection, CI, confidence interval.

Association of the VDR Tru9I SNP with specific pathological characteristics of patients with BC

We next investigated the association of the VDR Tru9I SNP with BC risk through immunohistochemical subtyping of BC tissues and TNM classification (Table 3). A direct association with risk of BC was found for at least one ‘u’ allele, which was more evident and significant for G-IV stage carcinoma (OR, 5.36; 95% CI, 1.181–24.338).

Table 3

Associations of the VDR Tru9I single nucleotide polymorphism and risk of premenopausal BC according to the immunohistochemical and pathological characteristics of patients with BC.

	Tru9I genotype
	UU (GG)		Uu+uu (GA+AA)
BC characteristics	Cases/Controls	OR (95% CI)	Cases/Controls	OR (95% CI)
Types of BC
IDC	152/398	Ref	45/99	1.19 (0.799–1.774)
ILC	13/398	Ref	0/99	1.67 × 10−9 (1.67 × 10−9–1.67 × 10−9)
DCIS	5/398	Ref	1/99	0.804 (0.93–6.96)
Other	9/398	Ref	3/99	1.34 (0.356–5.042)
Grades of BC
I	5/398	Ref	1/99	0.804 (0.093–6.96)
II	131/398	Ref	33/99	1.02 (0.66–1.578)
III	35/398	Ref	10/99	1.149 (0.55–2.399)
IV	3/398	Ref	4/99	5.36 (1.181–24.338)*
II & III	2/398	Ref	0/99	12 × 10−9 (12 × 10−9–12 × 10−9)
Unknown	3/398	Ref	1/99	1.34 (0.138–13.02)
Immunohistochemical classification
Estrogen receptor status
Positive (0 to +9)	103/398	Ref	23/99	0.898 (0.543–1.484)
Negative (−1 to −2)	25/398	Ref	9/99	1.447 (0.655–3.199)
None	51/398	Ref	17/99	1.34 (0.742–2.421)
Progesterone receptor status
Positive (0 to +8)	92/398	Ref	22/99	0.961 (0.575–1.608)
Positive (−1 to −2)	30/398	Ref	8/99	1.072 (0.477–2.411)
None	57/398	Ref	19/99	1.34 (0.762–2.355)
Human epidermal growth factor receptor 2 status
Positive (0 to +3)	78/398	Ref	20/99	1.031 (0.602–1.766)
Negative (−1 to −2)	18/398	Ref	2/99	0.447 (0.102–1.957)
None	83/398	Ref	27/99	1.308 (0.804–2.128)
TNM classification (For invasive cases)
Size of primary tumor (T stage)
Tx	1/398	Ref	1/99	4.681 (0.277–79.215)
Tis	3/398	Ref	1/99	1.281 (0.13–12.598)
T1	0/398	Ref	0/99	NC
T2	27/398	Ref	9/99	1.412 (0.639–3.121)
T3	25/398	Ref	6/99	0.896 (0.354–2.267)
T4	23/398	Ref	3/99	0.521 (0.152–1.783)
No involvement	100/398	Ref	29/99	1.135 (0.706–1.823)
Metastasize to regional lymph node (N stage)
Nx	4/398	Ref	2/99	1.821 (0.325–10.218)
N0	20/398	Ref	3/99	0.593 (0.172–2.048)
N1	25/398	Ref	11/99	1.796 (0.847–3.806)
N2	16/398	Ref	3/99	0.737 (0.209–2.598)
N3	14/398	Ref	1/99	0.297 (0.038–2.304)
No involvement	100/398	Ref	29/99	1.134 (0.706–1.822)
Distant metastasis (M stage)
Mx	29/398		10/99	1.386 (0.654–2.94)
M0	39/398	Ref	9/99	0.928 (0.435–1.979)
M1	11/398	Ref	1/99	0.365 (0.047–2.864)
No involvement	100/398	Ref	29/99	1.166 (0.73–1.862)

BC, breast cancer; CI, confidence interval; OR, odds ratio. VDR Tru9I “Uu” and “uu” genotypes were combined, because there were very few subjects with “uu.” ORs were calculated using multinomial logistic regression analysis adjusted for age and time of blood collection. *P < 0.05.

Relationship of the VDR Tru9I SNP and risk factors associated with BC

The possible relationship of the Tru9I SNP and other contributing factors to the risk of BC are presented in Table 4. Those with the mutant (Uu+uu) genotype along with Sindhi and Baluchi ethnicity and marital age greater than 20 years were at significantly increased risk of developing BC; however, there were no significant associations with other contributing factors and BC.

Table 4

Combined associations of the VDR Tru9I single nucleotide polymorphism and factors contributing to the risk of premenopausal BC.

BC risk factors	Combined		UU		Uu+uu
	Cases/Controls	OR (95% CI)	Cases/Controls	OR (95% CI)	Cases/Controls	OR (95% CI)
Age (years)#
20–26 (Younger)	20/58	Ref	16/51	Ref	4/5	Ref
27–33	34/78	1.264 (0.661–2.418)	23/53	1.383 (0.667–2.914)	11/23	0.598 (0.134–2.675)
34–40	110/225	1.418 (0.812–2.475)	83/179	1.478 (0.796–2.745)	27/44	0.767 (0.189–3.109)
41–45 (Older)	64/142	1.307 (0.726–2.352)	57/115	1.58 (0.829–3.012)	7/27	0.324 (0.068–1.535)
Total	228/503		179/398		49/99
Ethnicity/Race
Urdu-Speaking	94/252	Ref	78/199	Ref	16/51	Ref
Sindhi	41/12	9.41 (4.719–18.767)*	29/9	8.456 (3.809–18.772)*	12/3	14.934 (3.573–62.418)*
Panjabi	20/104	0.522 (0.305–0.891)*	16/86	0.493 (0.272–0.897)*	4/17	0.675 (0.193–2.335)
Pukhtoon	18/47	0.996 (0.549–1.806)	15/39	0.946 (0.492–1.821)	3/8	0.97 (0.22–4.283)
Baluchi	19/4	12.926 (4.265–39.171)*	12/3.	10.875 (2.963–39.908)*	7/1	19.323 (2.15–173.641)*
Others	36/84	1.17 (0.738–1.854)	29/62	1.188 (0.709–1.992)	7/19	1.179 (0.408–3.41)
Total	228/503		179/398		49/99
Marital status
Yes	208/363	Ref	162/286	Ref	45/75	Ref
No	20/140	0.139 (0.071–0.273)*	16/112	0.15 (0.071–0.318)*	4/24.	0.111 (0.024–0.524)*
Total	228/503		179/398		49/99
Marital age (years)
<20	105/244	Ref	84/186	Ref	21/56	Ref
20 to 30	95/117	3.343 (2.157–5.18)*	71/98	2.48 (1.525–4.033)*	24/19	10.645 (3.473–32.623)*
>30	8/2.	20.996 (4.15–106.236)*	8/2.	15.914 (3.138–80.698)*	0/0	NC
Total	208/363		163/286		45/75
Parity
Parous	191/342	Ref	149/267	Ref	40/71	Ref
Nulliparous	37/161	0.347 (0.216–0.557)*	30/131	0.366 (0.216–0.619)*	9/28	0.395 (0.145–1.071)
Total	228/503		179/398		45/99
Age at first child birth (years)
Below 20	49/63	Ref	41/55	Ref	8/8	Ref
20 to 24	81/118	0.528 (0.33–0.845)*	64/87	0.515 (0.302–0.879)*	17/30	0.582 (0.208–1.627)
25 to 29	39/115	0.621 (0.328–1.179)	30/88	0.458 (0.215–0.977)*	9/26	1.533 (0.41–5.731)
>29	17/43	1.218 (0.753–1.969)	11/36	1.086 (0.639–1.846)	6/7	2.006 (0.604–6.661)
Total	186/339		146/266		40/71
History of breast feeding
Yes	186/339	Ref	146/266	Ref	40/71	Ref
No	22/24	1.674 (0.901–3.107)	17/20	1.54 (0.773–3.068)	5/4	2.269 (0.528–9.748)
Total	208/363		163/286		45/75
Age at menarche (years)
12 to 14	186/423	Ref	148/334	Ref	38/83	Ref
>14	28/45	0.913 (0.479–1.74)	13/29	0.995 (0.5–1.978)	1/6	0.306 (0.035–2.662)
<12	14/35	1.389 (0.838–2.303)	18/35	1.111 (0.606–2.035)	10/10	2.567 (0.946–6.966)
Total	228/503		179/398		49/99
Use of hormone replacement therapy
No	209/481	Ref	168/382	Ref	41/93	Ref
Yes	19/22	1.954 (1.034–3.691)*	11/16.	1.547 (0.702–3.412)	8/6	3.114 (0.99–9.798)
Total	228/503		179/398		49/99
Use of oral contraceptive
No	197/454	Ref	155/361	Ref	42/87	Ref
Yes	31/49	1.473 (0.91–2.383)	24/37	1.521 (0.878–2.635)	7/12	1.17 (0.423–3.234)
Total	228/503		179/398		49/99
Waist to hip ratio
Good (<0.8)	27/92	Ref	20/71	Ref	7/19	Ref
Average (0.8–0.85)	15/173	0.296 (0.15–0.585)*	12/134	0.317 (0.146–0.685)*	3/37	0.252 (0.057–1.12)
High (>0.85)	186/238	2.687 (1.672–4.317)*	147/193	2.727 (1.58–4.707)*	39/43	2.568 (0.949–6.944)
Total	228/503		179/398		49/99
BMI (kg/m2)
Normal/healthy weight (18.6–23)	56/157	Ref	43/127	Ref	13/29	Ref
Underweight (>18.5)	22/64	0.966 (0.544–1.717)	19/48	1.196 (0.633–2.261)	3/14	0.383 (0.09–1.637)
Overweight (23.1–27)	79/158	1.381 (0.909–2.097)	63/132	1.374 (0.858–2.2)	16/23	1.689 (0.654–4.364)
Obese (>27)	71/124	1.589 (1.028–2.456)	54/91	1.708 (1.036–2.815)*	17/33	1.183 (0.474–2.953)
Total	228/503		179/398		49/99
Family history of BC
No	191/455		149/361	Ref	42/88	Ref
Yes	37/48	1.83 (1.153–2.906)*	30/37	1.947 (1.156–3.28)*	7//11	1.348 (0.473–3.841)
Total	228/503		179/398		49/99

VDR Tru9I “Uu” and “uu” genotypes were combined because there were very few subjects in the “uu” category. DNA was amplified from 228/228 patients’ samples and 497/503 control subjects’ samples. ORs were adjusted for age and time of blood collection. *P <0.05.

#Crude/unadjusted ORs for age were calculated. BC, breast cancer; BMI, body mass index; NC, not calculated, OR, odds ratio.

Nine models of contributing factors were prepared in which the Tru9I genotype was adjusted by sequentially adding individual contributing factors (Appendix 1). This analysis shows that the Tru9I ‘uu’ genotype in model 8 significantly increased the risk of BC (OR, 1.288; 95% CI, 0.16–10.398).

Discussion

Among Asian countries, Pakistan has the highest incidence of BC,[27] which is a heterogeneous group of diseases associated with various risk factors. The most common risk factor is exposure to endogenous hormones such as estradiol.[28] Approximately 15% of BC is caused by genetic changes,[29] including those associated with high penetrance genes such as BRCA 1/2,[30] and low-to-moderate penetrance genes such as checkpoint kinase-2 (CHEK2) and VDR.[20,31] VDR contributes to the prevention of BC, and the vitamin D/VDR complex induces apoptosis, reduces angiogenesis, proliferation, and invasion and increases cell differentiation.[32]

Evidence indicates that the VDR mediates most of the activities of vitamin D. Analyses of common VDR SNPs (Fok1, Bsm1, Apa1, Taq1, and Poly-A) show associations between VDR SNPs and BC, although the results are inconsistent among different types of exposure.[20,22,33,34] The Fok1 ‘ff’ genotype may increase the risk of BC.[35,36] However, evidence suggests that Fok1 significantly correlates with human epidermal growth factor receptor 2 status[20] and is significantly associated with estrogen receptor status of patients with BC.[37] One study shows that the Bsm1 ‘b’ allele is associated with risk of BC;[20] although another study reports the opposite results.[22] To our knowledge, the present study is the first to determine the association of the low penetrance VDR Tru9I polymorphism with premenopausal BC. The VDR Tru9I is caused by an A to G substitution that was first identified in 2000.[38] VDR Tru9I resides in intron 8, +443-bp from the end of exon 8, and may be involved in the regulation of gene expression through its affect on enhancer affinity and target area.[39]

The present study shows that the Tru9I mutant ‘uu’ genotype was associated with an increased risk of developing premenopausal BC. There are few association studies of Tru9I in other diseases such as coronary heart disease (CHD), colorectal adenoma, and prostate cancer.[26,39,40] However, Tru9I is not associated with CHD[39] and prostate cancer[40] among the Chinese Han population, whereas Tru9I lowers the risk of colorectal carcinoma among women living in North Carolina.[26] We show further that mutant Tru9I may be associated with the development of Grade IV carcinoma, whereas another study shows the opposite relationship with colorectal adenoma, which is more pronounced for multiple, larger, and sessile adenomas, and perhaps with dysplasia.[26]

This study has certain limitations. First, the sample size was relatively small. Large-scale studies are therefore required to support our conclusions. Second, we did not study environmental factors including diet and lifestyle, which may modify the association between the VDR Tru9I polymorphism and BC. Finally, other VDR SNPs and serum vitamin D levels were not studied. However, further research is required to confirm our findings and explore the possible mechanisms of BC underlying the association of VDR Tru9I with oncogenesis to facilitate the development of more effective treatment.

We conclude that the VDR Tru9I ‘uu’ genotype may predict a high risk for premenopausal BC.