Association of serum 25-hydroxyvitamin D with the incidence of 16 cancers, cancer mortality, and all-cause mortality among individuals with metabolic syndrome: a prospective cohort study

Abstract

Purpose
The relationship between vitamin D levels and cancer incidence and mortality in individuals with metabolic syndrome (MetS) remains poorly explored. Herein, we aimed to determine the association between 25-hydroxyvitamin D [25(OH)D] concentrations and the risk of 16 cancer incidence types and cancer/all-cause mortality in patients with MetS.

Methods
We enrolled 97,621 participants with MetS at recruitment from the UK Biobank cohort. The exposure factor was baseline serum 25(OH)D concentrations. The associations were examined using Cox proportional hazards models, which were displayed as hazard ratios (HRs) with 95% confidence intervals (CIs).

Results
Over a median follow-up period of 10.92 years for cancer incidence outcomes, 12,137 new cancer cases were recorded. We observed that 25(OH)D concentrations were inversely related to the risk of colon, lung, and kidney cancer, and HRs (95% CI) for 25(OH)D ≥ 75.0 vs. < 25.0 nmol/L were 0.67 (0.45–0.98), 0.64 (0.45–0.91), and 0.54 (0.31–0.95), respectively. The fully adjusted model revealed a null correlation between 25(OH)D and the incidence of stomach, rectum, liver, pancreas, breast, ovary, bladder, brain, multiple myeloma, leukemia, non-Hodgkin lymphoma, esophagus, and corpus uteri cancer. Over a median follow-up period of 12.72 years for mortality outcomes, 8286 fatalities (including 3210 cancer mortalities) were documented. An “L-shaped” nonlinear dose–response correlation was detected between 25(OH)D and cancer/all-cause mortality; the respective HRs (95% CI) were 0.75 (0.64–0.89) and 0.65 (0.58–0.72).

Conclusion
These findings emphasize the importance of 25(OH)D in cancer prevention and longevity promotion among patients with MetS.

Full text

Vol.:(0123456789)
1 3
European Journal of Nutrition (2023) 62:2581–2592
https://doi.org/10.1007/s00394-023-03169-x
ORIGINAL CONTRIBUTION
Association ofserum 25‑hydroxyvitamin D withtheincidence of16
cancers, cancer mortality, andall‑cause mortality amongindividuals
withmetabolic syndrome: aprospective cohort study
EWu1,2,3· Jun‑PingGuo1· KaiWang4· Hong‑QuanXu2,3· TianXie2,3· LinTao2,3· Jun‑TaoNi5
Received: 24 June 2022 / Accepted: 4 May 2023 / Published online: 20 May 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany 2023
Abstract
Purpose The relationship between vitamin D levels and cancer incidence and mortality in individuals with metabolic
syndrome (MetS) remains poorly explored. Herein, we aimed to determine the association between 25-hydroxyvitamin D
[25(OH)D] concentrations and the risk of 16 cancer incidence types and cancer/all-cause mortality in patients with MetS.
Methods We enrolled 97,621 participants with MetS at recruitment from the UK Biobank cohort. The exposure factor was
baseline serum 25(OH)D concentrations. The associations were examined using Cox proportional hazards models, which
were displayed as hazard ratios (HRs) with 95% confidence intervals (CIs).
Results Over a median follow-up period of 10.92years for cancer incidence outcomes, 12,137 new cancer cases were
recorded. We observed that 25(OH)D concentrations were inversely related to the risk of colon, lung, and kidney cancer,
and HRs (95% CI) for 25(OH)D ≥ 75.0 vs. < 25.0nmol/L were 0.67 (0.45–0.98), 0.64 (0.45–0.91), and 0.54 (0.31–0.95),
respectively. The fully adjusted model revealed a null correlation between 25(OH)D and the incidence of stomach, rectum,
liver, pancreas, breast, ovary, bladder, brain, multiple myeloma, leukemia, non-Hodgkin lymphoma, esophagus, and corpus
uteri cancer. Over a median follow-up period of 12.72years for mortality outcomes, 8286 fatalities (including 3210 cancer
mortalities) were documented. An “L-shaped” nonlinear dose–response correlation was detected between 25(OH)D and
cancer/all-cause mortality; the respective HRs (95% CI) were 0.75 (0.64–0.89) and 0.65 (0.58–0.72).
Conclusion These findings emphasize the importance of 25(OH)D in cancer prevention and longevity promotion among
patients with MetS.
Keywords 25-Hydroxyvitamin D· Metabolic syndrome· Cancer· All-cause mortality· Prevention
Abbreviations
MetS Metabolic syndrome
25(OH)D 25-Hydroxyvitamin D
HR Hazard ratio
CI Confidence interval
VD Vitamin D
CVD Cardiovascular disease
BP Blood pressure
FG Fasting glucose
TG Triglycerides
E. Wu, Jun-Ping Guo, Kai Wang and Jun-Tao Ni are equal
contributors.
* Tian Xie
xbs@hznu.edu.cn
* Lin Tao
taolin@hznu.edu.cn
* Jun-Tao Ni
njt1992@zju.edu.cn
1 Rehabilitation andNursing School, Hangzhou Vocational &
Technical College, Hangzhou310018, Zhejiang, China
2 School ofPharmacy, Hangzhou Normal University,
Hangzhou311121, Zhejiang, China
3 Key Laboratory ofElemene Class Anti-Cancer Chinese
Medicines, Engineering Laboratory ofDevelopment
andApplication ofTraditional Chinese Medicines,
Collaborative Innovation Center ofTraditional Chinese
Medicines ofZhejiang Province, Hangzhou Normal
University, Hangzhou311121, Zhejiang, China
4 Department ofacupuncture andmassage, 2nd Affiliated
Hospital, School ofMedicine, Zhejiang University,
Hangzhou310009, Zhejiang, China
5 Women’s Hospital School ofMedicine Zhejiang University,
Hangzhou310006, Zhejiang, China

2582 European Journal of Nutrition (2023) 62:2581–2592
1 3
HDL-C High-density lipoprotein cholesterol
ICD International classification of diseases
TDI Townsend deprivation index
BMI Body mass index
DM Diabetes mellitus
IQR Interquartile range
RCS Restricted cubic spline
PAR% Population-attributable risk percentage
NHL Non-Hodgkin lymphoma
MM Multiple myeloma
RCT Randomized controlled trial
Introduction
Metabolic syndrome (MetS) has been considered a serious
issue for public health, with one-quarter of the global popu-
lation estimated to be living with MetS in 2018 [1]. Individ-
uals with MetS are at a 1.5 times higher risk of death and a
two-fold higher risk of overall cardiovascular disease (CVD)
than the general population [2]. MetS can also increase the
risk of several types of cancer, including kidney and colon
cancer, and over two-tenths of cancers in the UK have been
associated with MetS [3]. Therefore, identifying modifiable
factors is essential for preventing or delaying MetS compli-
cations and premature death.
Numerous researchers have suggested that a substantial
proportion of individuals suffering from MetS tend to have
insufficient levels of vitamin D (VD), and VD deficiency
has been proposed as a potential risk factor for MetS. In
addition, high VD levels may afford a 43% reduction in car-
diovascular and metabolic disorders [4]. Moreover, given
that central obesity is one of the key factors related to MetS,
it can be expected that this population has low VD levels,
with an intervention trial also suggesting a link between
MetS and VD deficiency [5]. 25(OH)D has been related to
numerous cancers [6], such as colon, breast, and prostate
cancer [7]. Various observational studies have also revealed
that higher VD levels can be correlated with lower cancer
mortality and total mortality risk [8, 9]. For example, a
cohort study assessing 1801 patients with MetS revealed
a protective correlation between elevated 25(OH)D levels
and all-cause mortality (hazard ratio [HR]: 0.36) [10]; how-
ever, the small sample size of the study limited the ability
to analyze multiple causes of death, such as cancer-related
death. Furthermore, prospective studies examining the cor-
relation between VD and multiple cancer types in popula-
tions with MetS are rare. Considering the growing epidemic
of MetS worldwide and potential links between MetS, can-
cer, and 25(OH)D, exploring the effect of VD in individu-
als with MetS could help the public gain a comprehensive
understanding of the correlation between VD and cancer,
and potentially identify strategies for preventing or treating
cancers. Therefore, extensive cohort studies are required to
fill this gap in knowledge.
In the present study, we utilized data from the UK
Biobank to examine the relationships between 25(OH)D and
the risk of 16 types of cancer incidence, cancer mortality,
and all-cause mortality among patients with MetS.
Materials andmethods
Study participants
The UK Biobank is an ongoing cohort that has enrolled > 0.5
million individuals (37–73years old) between 2006 and
2010 [11]. All participants finished the interview and ques-
tionnaire, and biological samples were collected across 22
centers in England, Wales, and Scotland [12]. In the pre-
sent study, participants who were aged < 40 or > 70years
(n = 15), had missing 25(OH)D values (n = 54,144), did not
have MetS (n = 334,488), failed to follow-up (n = 266), and
had cancer at recruitment (n = 9722) were removed. In addi-
tion, we deleted patients who were dead or diagnosed with
cancer within the initial 3-year follow-up period (n = 6067)
and had self-reported cancer during follow-up (n = 91).
Overall, the study comprised 97,621 patients with baseline
MetS (Supplementary Fig. S1).
Ascertaining MetS
We defined MetS as the existence of at least three of the
risk factors listed below, based on a well-accepted consensus
from a joint interim statement [13]: ① Abdominal obesity:
measured by waist circumference thresholds that are ethnic
and gender-group specific; ② Elevated blood pressure (BP):
systolic BP of 130mmHg or higher, diastolic BP at or more
than 85mmHg, or lower hypertension therapy; ③ Elevated
fasting glucose (FG): FG ≥ 100mg/dL, or drug therapy to
reduce elevated FG; notably, blood biochemistry was exam-
ined in a random and non-fasting state among UK Biobank
participants, and we referred to the method employed by
Eastwood, where glucose ≥ 11.1mmol/L (200mg/dL) or
HbA1c ≥ 48mmol/mol (6.5%) captured hyperglycemia even
in the non-fasted state [14]. ④ Elevated triglycerides (TG):
TG ≥ 150mg/dL or drug therapy to reduce raised TG; ⑤
Reduced high-density lipoprotein cholesterol (HDL-C):
HDL-C levels below 40/50mg/dL in males/females.
Exposure assessment
The UK Biobank collected a range of biological samples,
including blood, urine, and saliva, at baseline (2006–2010).
These samples were collected into different containers
and divided into multiple aliquots, with half preserved in

2583European Journal of Nutrition (2023) 62:2581–2592
1 3
Table 1 Baseline characteristics by serum 25(OH)D concentrations
Characteristic Serum 25(OH)D concentration, nmol/L
Total (n = 97, 621) < 25.0 (n = 18, 938) 25.0–49.9 (n = 46,
453)
50.0–74.9 (n = 26,
564)
≥ 75.0 (n = 5666)
Age, mean (SD), years 57.4 (7.8) 55.5 (8.0) 57.2 (7.8) 58.6(7.5) 59.0 (7.4)
Male, n (%) 50,445 (51.7) 9741 (51.4) 23,952 (51.6) 13,736(51.7) 3016 (53.2)
BMI, mean (SD) 31.5 (4.9) 32.3 (5.6) 31.8 (4.8) 30.9(4.3) 30.2 (4.1)
White ethnicity, n (%) 90,047 (92.2) 15,366 (81.1) 43,259 (93.1) 25,835 (97.3) 5587 (98.6)
Employment, n (%)
Employed 51,150 (52.4) 10,592 (55.9) 25,410 (54.7) 12,664 (47.7) 2484 (43.8)
Retired 35,119 (36.0) 5008 (26.4) 15,854 (34.1) 11,548 (43.5) 2709 (47.8)
Unemployed 2010 (2.1) 641 (3.4) 907 (2.0) 402 (1.5) 60 (1.1)
Other 9342 (9.6) 2697 (14.2) 4282 (9.2) 1950 (7.3) 413 (7.3)
Education, n (%)
Higher 24,385 (25.0) 5342 (28.2) 12,075 (26.0) 5860 (22.1) 1108 (19.6)
Middle 30,958 (31.7) 5778 (30.5) 14,709 (31.7) 8599 (32.4) 1872 (33.0)
Lower 13,374 (13.7) 2469 (13.0) 6364 (13.7) 3711 (14.0) 830 (14.6)
Vocational 5468 (5.6) 962 (5.1) 2538 (5.5) 1608 (6.1) 360 (6.4)
Other 23,436 (24.0) 4387 (23.2) 10,767 (23.2) 6786 (25.5) 1496 (26.4)
TDI, n (%)
Least deprived 19,559 (20.0) 2757 (14.6) 9148 (19.7) 6300 (23.7) 1354 (23.9)
Intermediate deprived 58,538 (60.0) 10,624 (56.1) 27,929 (60.1) 16,462 (62.0) 3523 (62.2)
Most deprived 19,524 (20.0) 5557 (29.3) 9376 (20.2) 3802 (14.3) 789 (13.9)
Never smoking, n (%) 49,416 (50.6) 9729 (51.4) 23,639 (50.9) 13,400 (50.4) 2648 (46.7)
Alcohol consumption, n (%)
Low 32,486 (33.3) 5251 (27.7) 15,210 (32.7) 9837 (37.0) 2188 (38.6)
Moderate 38,485 (39.4) 6607 (34.9) 18,625 (40.1) 10,939 (41.2) 2314 (40.8)
High 26,650 (27.3) 7080 (37.4) 12,618 (27.2) 5788 (21.8) 1164 (20.5)
Physical activity, n (%)
Low 27,939 (28.6) 6813 (36.0) 13,530 (29.1) 6348 (23.9) 1248 (22.0)
Moderate 36,341 (37.2) 6429 (33.9) 17,393 (37.4) 10,315 (38.8) 2204 (38.9)
High 33,341 (34.2) 5696 (30.1) 15,530 (33.4) 9901 (37.3) 2214 (39.1)
Medication, n (%)
Cholesterol-lowering drug 30,625 (31.4) 5864 (31.0) 14,167 (30.5) 8393 (31.6) 2201 (38.8)
Antihypertensive drug 36,471 (37.4) 6861 (36.2) 17,036 (36.7) 10,145 (38.2) 2429 (42.9)
Insulin 2956 (3.0) 740 (3.9) 1367 (2.9) 680 (2.6) 169 (3.0)
VD supplementation 3214 (3.3) 334 (1.8) 1354 (2.9) 1234 (4.6) 292 (5.2)

2584 European Journal of Nutrition (2023) 62:2581–2592
1 3
a working archive at –80°C and the other half stored in
a nitrogen-vapor backup archive. The average time from
venipuncture to sample storage was 24 ± 2.5h. The 25(OH)
D concentrations were detected utilizing a direct competi-
tive chemiluminescent immunoassay. Detected levels ranged
between 10 and 375 (nmol/L). Specific details have been
described elsewhere [15]. The 25(OH)D levels were sub-
sequently divided into four categories based on the Guide-
lines of Endocrine Society Clinical Practice: ① Severe
deficiency: < 25.0; ② Moderate deficiency: 25.0–49.9; ③
Insufficient: 50.0–74.9; and ④ Sufficient ≥ 75.0nmol/L [16].
Ascertainment ofoutcomes
The primary outcome was the initial incidence of cancer.
Information regarding cancer incidence was obtained using
cancer registries linked to the NHS Information Centre
(residents of England/Wales) and NHS Scotland (residents
of Scotland). Herein, we recorded the incidence of 16 can-
cer types with sample sizes of more than 100 new cases,
including cancer of the esophagus, non-Hodgkin lymphoma
(NHL), stomach, liver, multiple myeloma, colon, rectum,
breast, leukemia, pancreas, lung, corpus uteri, ovary, as
well as kidney, bladder, and brain [17]. Cancer diagnoses
were coded using the ICD–10 and ICD–9 (Supplementary
TableS1). According to the different censoring dates of each
database, we used the first occurrence of cancer, death, Feb-
ruary 29, 2020 (for residents of England/Wales), or October
31, 2015 (for residents of Scotland) as the endpoint (which-
ever occurred first).
Table 1 (continued)
Characteristic Serum 25(OH)D concentration, nmol/L
Total (n = 97, 621) < 25.0 (n = 18, 938) 25.0–49.9 (n = 46,
453)
50.0–74.9 (n = 26,
564)
≥ 75.0 (n = 5666)
Family history, n (%)
Hypertension 49,240 (50.4) 9650 (51.0) 23,540 (50.7) 13,225 (49.8) 2825 (49.9)
DM 27,948 (28.6) 6023 (31.8) 13,452 (29.0) 7006 (26.4) 1467 (25.9)
Cancer 33,821 (34.6) 6022 (31.8) 16,217 (34.9) 9466 (35.6) 2166 (37.3)
Season of vitamin D assessment, n (%)
Spring 29,228 (29.9) 5684 (30.0) 13,966 (30.1) 7988 (30.1) 1590 (28.1)
Summer 29,308 (30.0) 5464 (28.9) 13,967 (30.1) 8057 (30.3) 1820 (32.1)
Autumn 17,416 (17.8) 3432 (18.1) 8249 (17.8) 4706 (17.7) 1029 (18.2)
Winter 21,669 (22.2) 4358 (23.0) 10,271 (22.1) 5813 (21.9) 1227 (21.7)
Due to rounding, the sum of percentage may not be 100%. Binary variables show only one of the options
SD standard deviation, IQR interquartile range, VD vitamin D,DM diabetes mellitus
All pvalue < 0.001, except for sex (pvalue = 0.068)
Fig. 1 HRs (95% CIs) for 16 site-specific cancer incidence according
to serum 25(OH)D categories. NHL non-Hodgkin lymphoma. MM
Multiple myeloma. It is ranked by effect size. The associations were
examined in the fully adjusted Cox proportional hazard regression
models, 25(OH)D concentration < 25.0nmol/L was considered refer-
ence group

2585European Journal of Nutrition (2023) 62:2581–2592
1 3
The secondary outcome was mortality, including death
from cancer and all-cause. We obtained data regarding
deaths from the “Data Portal”, which was connected to
the NHS Digital Centre (residents of England/Wales) and
NHS central register (residents of Scotland) [18]. Cancer
mortality was recorded using ICD–10 codes: C00–C97. For
analyses with death as the outcome, according to the differ-
ent censoring dates of each database, either the death date,
or September 30, 2021 (for patients of England/Wales), or
October 31, 2021 (for patients of Scotland) was utilized as
the endpoint.
Assessment ofcovariates
Covariates were selected according to previous studies and
priori knowledge [19], including age at recruitment (years),
sex (male,female), ethnicity (white, other), education (higher,
middle, lower, vocational, other), employment (employed,
retired, unemployed, other), Townsend deprivation index
(TDI; least deprived, intermediate deprived, most deprived),
smoking (never, former/current), frequency of alcohol con-
sumption (low, middle, high), frequency of physical activ-
ity (low, moderate, high), body mass index (BMI; < 25,
25–30, ≥ 30kg/m2) [20], food intake, including vegetables,
fruit, fish, and processed meat, VD assessment season (spring,
summer, autumn, winter), history (yes/no) of anti-cholesterol
drugs, antihypertensive drugs and insulin, VD supplementa-
tion (yes/no), and family history (yes/no) of hypertension, dia-
betes mellitus (DM), and cancer. TDI measures the degree of
deprivation in an individual's residential area at recruitment,
incorporating baseline information on employment, social
class, housing, and car ownership [21]. The smaller the TDI
value, the more relatively wealthy the examined population
[22]. Details regarding covariates are provided in Supplemen-
tary Methods.
Statistical analysis
Baseline characteristics are summarized by the 25(OH)D
concentrations category; continuous data are presented as the
median (interquartile range [IQR]) or mean ± standard devia-
tion (SD), and categorical data are presented as frequency
(percentage). Generalized linear models were utilized to deter-
mine correlations between 25(OH)D and baseline character-
istics. The reverse Kaplan–Meier method was employed to
determine the median follow-up person-years. Missing values
Fig. 2 Multivariable-adjusted
dose–response associations
between 25(OH)D and the
incidence of 16 cancers. a Leu-
kemia cancer, b Liver cancer, c
Rectal cancer, d non-Hodgkin
lymphoma cancer, e Breast can-
cer, f Brain cancer, g Multiple
myeloma cancer, h Stomach
cancer, i Bladder cancer, j
Corpus uteri cancer, k Pancreas
cancer, l Colon cancer, m Lung
cancer, n Kidney cancer, o
Esophagus cancer, p Ovary
cancer. The dose–response asso-
ciations were examined in the
fully adjusted Cox proportional
hazard regression models based
on restricted cubic splines with
three knots, and the shaded area
represents the 95% CI for the
dose–response curve. 25(OH)
D 25-hydroxyvitamin D, NHL
non-Hodgkin lymphoma, MM
Multiple myeloma. All p for
nonlinearity < 0.05, except for
ovary cancer and brain cancer

2586 European Journal of Nutrition (2023) 62:2581–2592
1 3
Table 2 HRs (95%CIs) for cancer mortality and all-cause mortality according to serum 25(OH)D categories
Model 1: adjusted for age at recruitment, sex, ethnicity, employment, education, and TDI
Model 2: adjusted for model one and BMI, smoking, alcohol consumption, physical activity, and intake of vegetables, fruit, fish, and processed meat
Model 3: adjusted for model two and VD assessment season, VD supplementation, medication history of anti-cholesterol, antihypertensive, insulin, family history of high BP, DM, and cancer
Outcomes Serum 25(OH)D concentration, nmol/L
< 25 (n = 18,938) 25–49.9 (n = 46,453) 50.0–74.9 (n = 26,564) ≥ 75.0 (n = 5666) p for trend
Cancer mortality (n = 3210)
No. of cases/person-years 641/192158 1536/477364 854/273603 179/58395
Model 1 1.00 (ref) 0.83(0.76-0.91) 0.72(0.65-0.80) 0.67(0.57-0.80) <0.001
Model 2 1.00 (ref) 0.89(0.81-0.98) 0.80(0.72-0.89) 0.76(0.64-0.90) <0.001
Model 3 1.00 (ref) 0.89(0.81-0.98) 0.81(0.72-0.90) 0.75(0.64-0.89) <0.001
All-cause mortality (n = 8286)
No. of cases/person-years 1922/233507 3861/575736 2060/329978 443/70400
Model 1 1.00 (ref) 0.73(0.69-0.77) 0.60(0.56-0.64) 0.57(0.51-0.63) <0.001
Mode 2 1.00 (ref) 0.78(0.74-0.83) 0.69(0.64-0.73) 0.66(0.60-0.74) <0.001
Model 3 1.00 (ref) 0.78(0.74-0.83) 0.69(0.65-0.74) 0.65(0.58-0.72) <0.001

2587European Journal of Nutrition (2023) 62:2581–2592
1 3
of continuous variables were imputed with median values, and
mode values were imputed for categorical variables (Supple-
mentary Methods) [23].
Cox proportional hazard regression was performed to assess
the associations between 25(OH)D and the risk of developing
16 cancer types, mortality from cancer and all-cause among
patients with MetS. In model one, we adjusted for sociode-
mographic characteristics, including age at recruitment, sex,
ethnicity, employment, education, and TDI. In model two, we
additionally adjusted lifestyle factors, including BMI, smok-
ing, alcohol consumption, physical activity, and intake of veg-
etables, fruit, fish, and processed meat. We further adjusted
for factors associated with 25(OH)D, medication, and family
history in model three, i.e., VD assessment season, VD sup-
plementation, medication history of anti-cholesterol, antihy-
pertensive, insulin, family history of high BP, DM, and cancer
to account for potential confounding factors. The proportional
hazards assumption was assessed utilizing Schoenfeld residu-
als, and null significant evidence of non-proportionality was
detected (p > 0.05).
A restricted cubic spline (RCS), adjusted for model three,
was used to assess dose–response correlations. Moreover, we
calculated the population-attributable risk percentage (PAR%)
to determine the proportion that may be attributed to 25(OH)
D deficiency (< 75nmol/L) using the following formula [24]:
where Ρe represents the proportion of participants exposed to
deficiency of 25(OH)D, with HR obtained from Cox regres-
sions adjusted for model three (considering the 25(OH)D
sufficient group as reference). Stratified analyses were per-
formed based on baseline characteristics. The interaction
term beta from Cox models was used to determine multipli-
cative interactions.
PAR
%=
P
e
(HR −1)
P
e
(HR −1)+1×100%
,
Sensitivity analyses were performed to test the robust-
ness and validity of obtained results. First, we removed
patients with cancer at baseline and those who died or had
cancer within the initial 3year follow-up period to exclude
the possibility of reverse causality. Second, we used quar-
tiles to reclassify 25(OH)D levels. Statistical analyses and
plot graphs were generated using Stata (version 15.0) and R
(version 4.1.3). Statistical significance was determined by a
two-sided pvalue < 0.05.
Results
Baseline characteristics
Table1 presents the baseline characteristics of patients
with MetS. Among 97,621 participants, 19.4, 47.6, 27.2,
and 5.8% reported having severe deficiency, moderate defi-
ciency, insufficient levels, and sufficient levels of 25(OH)
D, respectively. Participants with elevated 25(OH)D lev-
els have a higher likelihood of being male and older, have
lower BMI values, and use antihypertensive drugs and VD
supplements.
Association between25(OH)D andtheincidence
of16 cancer types
Over a median follow-up period of approximately
10.92years (a total of 1,001,520 person-years) for can-
cer incidence outcomes, 12,137 new cancer cases were
recorded, and new cases of 16 site-specific cancers are
shown in Fig.1. Based on the multivariable-adjusted
model, a protective correlation was detected between
25(OH)D levels and colon, lung, and kidney cancers,
and respective HRs (95% CIs) for 25(OH)D concentra-
tions ≥ 75.0 vs. < 25.0nmol/L were 0.67 (0.45–0.98), 0.64
(0.45–0.91), and 0.54 (0.31–0.95) (Fig.1). The multivari-
able-adjusted RCS revealed inverse relationships between
25(OH)D levels and cancer of the colon, lung, and kidney
(p for nonlinearity < 0.05) (Fig.2). Moreover, we found
a null relationship between the VD status and cancer of
the stomach, rectum, liver, pancreas, breast, ovary, blad-
der, brain, multiple myeloma, along with an increased risk
of leukemia and NHL. Notably, although we discovered
links between 25(OH)D and cancer of the esophagus and
corpus uteri in model 1, the respective HRs (95% CIs) for
concentrations 50.0–74.9 vs. < 25.0 (nmol/L) were 0.60
(0.39–0.91) and 0.66 (0.48–0.91), and the association
disappeared after full adjustment in models 2 and 3 (Sup-
plementary TableS2).
Fig. 3 Multivariable-adjusted dose–response associations between
25(OH)D and cancer mortality and all-cause mortality. a Cancer
mortality, b All-cause mortality. The dose–response associations
were examined in the fully adjusted Cox proportional hazard regres-
sion models based on restricted cubic splines with three knots, the
gray shaded area represents the 95% CI for the dose–response curve.
25(OH)D 25-hydroxyvitamin D. All p for nonlinearity < 0.05

2588 European Journal of Nutrition (2023) 62:2581–2592
1 3
Association between25(OH)D andcancer/all‑cause
mortality
Over a median follow-up period of approximately
12.72years (a total of 1,209,621 person-years) for mor-
tality outcomes, 8286 fatalities (3210 cancer mortali-
ties) were documented (Table2). After total adjustment
for model three, 25(OH) D was inversely correlated with
death from cancer and all-cause, and the HRs (95% CIs)
for 25(OH)D concentrations ≥ 75.0 vs. < 25.0nmol/L were
0.75 (0.64–0.89) and 0.65 (0.58–0.72), respectively. Simi-
lar results were obtained in models one and two. Figure3
presents the “L-shaped” nonlinear dose–response relation-
ships between 25(OH)D and cancer/all-cause mortality (p
for nonlinearity < 0.001), with HRs approaching zero in the
fully adjusted model. Moreover, PAR% analysis indicated
that 18.25% (95% CI 10.46–25.48%) of all-cause mortality
incidents were theoretically attributable to serum 25(OH)D
deficiency (< 75nmol/L).
Fig. 4 Multivariable-adjusted dose–response associations between
25(OH)D and all-cause mortality stratified by baseline characteristics.
a Age at baseline (40–49, 50–59, 60–70years), b Sex (female, male),
c Education (Higher, others), d Employment (employed, retired,
unemployed, other), e TDI (least deprived, intermediate deprived,
most deprived), f BMI category (< 25kg/m2, 25–30kg/m2, ≥ 30 kg/
m2), g smoking status (never smoke, current/former smoker), h
physical activity (low, moderate, high), i Cholesterol-lowering drug
(yes, no), j Antihypertensive drug (yes, no), k Insulin (yes, no), l
Vitamin D supplementation (yes, no), (m) Family history of hyper-
tension (yes, no), (n) Family history of diabetes mellitus (yes, no),
(o) Family history of cancer (yes, no). All the dose–response asso-
ciations were examined in the fully adjusted Cox proportional haz-
ard regression models based on restricted cubic splines with 3 knots,
except for when stratified by age baseline, age group (40–49, 50–59,
60–70 years) was included, but not baseline age (years) in the full
model. All p for interactions were < 0.001. The shaded area represents
the 95% CI for the dose–response curve. 25(OH)D 25-hydroxyvita-
min D, TDI Townsend deprivation index, InterM Intermediate, BMI
body mass index, DM diabetes mellitus. All p for interactions < 0.05

2589European Journal of Nutrition (2023) 62:2581–2592
1 3
Subgroup analysis andsensitivity analyses
Based on the findings of stratified analyses, correlations
between 25(OH)D and all-cause mortality were stronger
among those who were younger, female, employed, had
better socioeconomic status, and had normal weight (p for
interaction < 0.001). However, we found a null associa-
tion between VD supplementation and all-cause mortality
(HR 0.86; 95% CI 0.45–1.62) (Fig.4 and Supplementary
TableS3). Similar results were detected in the sensitivity
analyses on categorizing 25(OH)D concentrations into four
groups based on quartiles (Supplementary TableS4–S5).
Discussion
Among 97,621 patients with MetS, we found an “L-shaped”
dose–response correlation between 25(OH)D levels and
cancer/all-cause mortality. However, the subgroup analysis
revealed that VD supplementation was unlikely to reduce the
risk of all-cause mortality. Moreover, we noted an inverse
dose–response correlation between 25(OH)D and the risk of
colon, lung, and kidney cancer (p for nonlinearity < 0.05).
Notably, the risk of colon, lung, and kidney cancer and
death from cancer and all-cause decreased with increasing
25(OH)D concentrations, leveling off at approximately 50
(nmol/L). This suggests that the increased risk is primarily
related to low 25(OH)D status. Accordingly, our study has
potential implications for cancer prevention and increasing
life expectancy and contributes to our understanding of dis-
ease etiology.
Although several previous studies have observed an
inverse correlation between 25(OH)D and the death risk
among the ordinary population [9, 25, 26] or individuals
with CVD [27] diabetes [19], or prediabetes [28], to our
knowledge, this is the first study to focus on patients with
MetS and outcomes of 16 incident cancer types and cancer
mortality. A recent analysis has indicated that individuals
with prediabetes (HR: 0.66) or DM (HR: 0.60) and sufficient
25(OH)D levels had a lower risk of mortality than those with
severe 25(OH)D deficiency [15]. This is similar to the HR
value of 0.65 obtained in our study among the MetS popula-
tion. Moreover, several observational studies have indicated
that 25(OH)D deficiency carries a two-fold higher risk of
mortality among aging male subjects [29]. Notably, several
randomized controlled trials have revealed null evidence
regarding VD supplementation reducing the death risk [30,
31]. The null associations may be attributed to inadequate
dose and duration of VD supplementation, poor compliance,
and residual confounding in observational studies.
Numerous previous studies have documented a protec-
tive correlation between VD and the risk of colon [32], lung
[33], and kidney cancer [34]. We found similar results in
the MetS population and provided novel evidence of inverse
linear dose–response associations. It has been known for
years that VD deficiency is related to obesity and insulin
resistance; hence, the protective effect of VD in MetS and
related diseases can plausibly be expected. Additionally,
although several researchers have documented the role of
VD in colorectal cancer prevention [8], the association may
vary depending on the anatomical site. Herein, we observed
an inverse correlation between 25(OH)D and colon cancer
but not rectal cancer risk. A nested case–control study from
Europe also supports this finding [35]; this may be attributed
to the distinct carcinogenic mechanisms in the colorectum,
warranting further research.
Diverse mechanisms have been proposed regarding the
protective effect of high 25(OH)D levels in increasing life
expectancy and preventing several cancer types [36, 37]. For
instance, VD is related to better endothelial cell function and
glucocorticoid response, as well as anti-inflammatory prop-
erties [38–40]. Regarding potential antitumor mechanisms,
it has been found that VD can downregulate proliferation-
related genes, such as the JUN and JUND proto-oncogenes
[41], suppress the long non-coding RNA CCAT2 [42], and
promote MYC protein degradation to inhibit tumor cell pro-
liferation in several systems [43]. Furthermore, although VD
does not directly induce apoptosis [44], it is known to be
involved in the expression of apoptosis-related genes. For
example, pro-apoptotic proteins, such as BAX and BGA, are
upregulated in cancer cells, including colon cancer cells
[45, 46]. VD has also been found to activate and induce
autophagy in a healthy state to maintain homeostasis and
protect cells against damage from inflammation and cancer
[47]. For example, 1,25-(OH)2D3 was shown to promote
autophagic cell death in a VD receptor- and the p53-depend-
ent manner in lung cancer cells [48]. Remarkably, VD may
induce differentiation in certain cancer cells by upregulating
epithelial genes or inhibiting key epithelial–mesenchymal
transition-inducing transcription factors. In colon, lung, and
other cell lines, VD was found to induce E-cadherin, which
is associated with increased epithelial differentiation [49].
Additionally, VD plays a role in inhibiting tumor angiogen-
esis [50], suppressing cancer cell migration and invasion
[51], and enhancing the immune response against tumor
cells [52].
The strength of the present study lies in its large sample
size and the rich data resources of the UK Biobank, which
allowed us to analyze multiple causes of death and the risk
of cancer incidence in the MetS population. However, this
study has several limitations. First, although we adjusted for
potential confounders and followed up over 10 median years,
the possibility of unadjusted confounders, such as additional
details on VD supplementation and anti-glycemic drugs, the
latitude of residence, time spent outside in the sun, use of
sunscreen, and skin type, and inappropriate categorization

2590 European Journal of Nutrition (2023) 62:2581–2592
1 3
of confounders may persist. Second, 25(OH)D levels were
tested only once at recruitment; although a single measure-
ment of the 25(OH)D was considered a reasonable indicator
of VD status [53], non-differential misclassification bias may
exist. Third, although we adjusted the season in the model
to address seasonal variations in 25(OH)D concentrations.
Seasonal variation in 25(OH)D is complex, warranting more
sophisticated modeling in the future to convert 25OHD lev-
els to “year-round” concentrations [54, 55]. Fourth, as the
study was limited to individuals with MetS aged ≥ 40years,
it was impossible to directly generalize our findings to the
general population. Finally, we excluded participants with-
out 25(OH)D values and those lost to follow-up, which may
have affected the generalizability of the findings.
Conclusions
In the present study assessing participants with MetS, we
found an inverse association between 25(OH)D and colon,
lung, and kidney cancer. Null correlation between VD status
and cancer of the stomach, rectum, liver, pancreas, ovary,
bladder, brain, multiple myeloma, NHL, leukemia, and even
an increased risk of breast cancer was observed. Notably,
the fully adjusted model revealed no relationship between
25(OH)D and esophageal and corpus uteri cancer. Addition-
ally, we detected an inverse correlation between 25(OH)D
and cancer/all-cause mortality. These findings emphasize the
importance of 25(OH)D in cancer prevention and longevity.
Supplementary Information The online version contains supplemen-
tary material available at https:// doi. org/ 10. 1007/ s00394- 023- 03169-x.
Acknowledgements We thank the participants and staff of the UK
Biobank for their dedication and contribution to the research. We
appreciate the UK Biobank for giving us the opportunity to access the
database through the Access Management System.
Author contributions EW: performed the statistical analysis and wrote
the manuscript, J-PG and KW: provided consultation in their areas of
expertise, H-QX: were responsible for technical support, TX and LT:
designed the study, J-TN: revised the manuscript. All authors read and
approved the final manuscript. All authors have read and agreed to the
published version of the manuscript.
Funding This research was funded by the High Level Talent Research
Launch Project of Hangzhou Vocational & Technical College
(RCXY202242) (to E Wu). The funding sources had no role in study
design; in the collection, analysis, or interpretation of data; in the writ-
ing of the report; or in the decision to submit the article for publication.
Data availability The UK Biobank datasets are openly available by
submitting a data request proposal from https:// www. ukbio bank. ac.
uk/ (We accessed on 9 April 2022). We are authorized to access the
database through the Access Management System (AMS) (Application
number: 78563).
Declarations
Conflict of interest The authors declare no conflict of interest.
Ethics approval The UK Biobank was approved by the Research Eth-
ics Committees of the NorthWest Multi-Centre (reference no. 21/
NW/0157). All the study participants signed an informed consent form.
References
1. Saklayen MG (2018) The global epidemic of the metabolic syn-
drome. Curr Hypertens Rep 20(2):12. https:// doi. org/ 10. 1007/
s11906- 018- 0812-z
2. Peiris CL, van Namen M, O’Donoghue G (2021) Education-
based, lifestyle intervention programs with unsupervised exer-
cise improve outcomes in adults with metabolic syndrome. A
systematic review and meta-analysis. Rev Endocr Metab Disord
22(4):877–890. https:// doi. org/ 10. 1007/ s11154- 021- 09644-2
3. Micucci C, Valli D, Matacchione G, Catalano A (2016) Current
perspectives between metabolic syndrome and cancer. Oncotarget
7(25):38959–38972
4. Maroufi NF, Pezeshgi P, Mortezania Z, Pourmohammad P,
Eftekhari R, Moradzadeh M, Vahedian V, Nouri M (2020)
Association between vitamin D deficiency and prevalence of
metabolic syndrome in female population: a systematic review.
Hormone Mole Biol Clin Investigat. https:// doi. org/ 10. 1515/
hmbci- 2020- 0033
5. Prasad P, Kochhar A (2016) Interplay of vitamin D and metabolic
syndrome: a review. Diabet Meta Synd 10(2):105–112. https:// doi.
org/ 10. 1016/j. dsx. 2015. 02. 014
6. Yuan S, Baron JA, Michaëlsson K, Larsson SC (2021) Serum cal-
cium and 25-hydroxyvitamin D in relation to longevity, cardiovas-
cular disease and cancer: a Mendelian randomization study. NPJ
Genom Med 6(1):86. https:// doi. org/ 10. 1038/ s41525- 021- 00250-4
7. Carlberg C, Muñoz A (2022) An update on vitamin D signaling
and cancer. Semin Cancer Biol 79:217–230. https:// doi. org/ 10.
1016/j. semca ncer. 2020. 05. 018
8. Zhou J, Ge X, Fan X, Wang J, Miao L, Hang D (2021) Associa-
tions of vitamin D status with colorectal cancer risk and survival.
Int J Cancer 149(3):606–614. https:// doi. org/ 10. 1002/ ijc. 33580
9. Fan X, Wang J, Song M, Giovannucci EL, Ma H, Jin G, Hu Z,
Shen H, Hang D (2020) Vitamin D status and risk of all-cause
and cause-specific mortality in a large cohort: results from the
UK biobank. J Clin Endocrinol Metab 105(10):dgaa432. https://
doi. org/ 10. 1210/ clinem/ dgaa4 32
10. Thomas GN, Briain OH, Bosch JA, Pilz S, Loerbroks A, Kleber
ME, Fischer JE, Grammer TB, Böhm BO, März W (2012) Vita-
min D levels predict all-cause and cardiovascular disease mortal-
ity in subjects with the metabolic syndrome: the Ludwigshafen
risk and Cardiovascular health (LURIC) study. Diabetes Care
35(5):1158–1164. https:// doi. org/ 10. 2337/ dc11- 1714
11. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp
K, Motyer A, Vukcevic D, Delaneau O, O’Connell J, Cortes A,
Welsh S, Young A, Effingham M, McVean G, Leslie S, Allen N,
Donnelly P, Marchini J (2018) The UK Biobank resource with
deep phenotyping and genomic data. Nature 562(7726):203–209.
https:// doi. org/ 10. 1038/ s41586- 018- 0579-z
12. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J,
Downey P, Elliott P, Green J, Landray M, Liu B, Matthews P, Ong
G, Pell J, Silman A, Young A, Sprosen T, Peakman T, Collins R
(2015) UK biobank: an open access resource for identifying the

2591European Journal of Nutrition (2023) 62:2581–2592
1 3
causes of a wide range of complex diseases of middle and old age.
PLoS Med 12(3):e1001779. https:// doi. org/ 10. 1371/ journ al. pmed.
10017 79
13. Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI,
Donato KA, Fruchart JC, James WP, Loria CM, Smith SC Jr
(2009) Harmonizing the metabolic syndrome: a joint interim
statement of the International diabetes federation task force on
epidemiology and prevention; National heart, lung, and blood
institute; American heart association; World heart federation;
International atherosclerosis society; and International association
for the study of obesity. Circulation 120(16):1640–1645. https://
doi. org/ 10. 1161/ circu latio naha. 109. 192644
14. Eastwood SV, Mathur R, Atkinson M, Brophy S, Sudlow C, Flaig
R, de Lusignan S, Allen N, Chaturvedi N (2016) Algorithms for
the capture and adjudication of prevalent and incident diabetes in
UK biobank. PLoS ONE 11(9):e0162388. https:// doi. org/ 10. 1371/
journ al. pone. 01623 88
15. Zhang P, Guo D, Xu B, Huang C, Yang S, Wang W, Liu W, Deng
Y, Li K, Liu D, Lin J, Wei X, Huang Y, Zhang H (2022) Associa-
tion of serum 25-hydroxyvitamin D With cardiovascular outcomes
and all-cause mortality in individuals with prediabetes and diabe-
tes: results from the UK biobank prospective cohort study. Diabe-
tes Care 45(5):1219–1229. https:// doi. org/ 10. 2337/ dc21- 2193
16. Holick MF, Binkley NC, Bischoff-Ferrari HA, Gordon CM, Han-
ley DA, Heaney RP, Murad MH, Weaver CM (2011) Evaluation,
treatment, and prevention of vitamin D deficiency: an Endocrine
Society clinical practice guideline. J Clin Endocrinol Metab
96(7):1911–1930. https:// doi. org/ 10. 1210/ jc. 2011- 0385
17. Knuppel A, Fensom GK, Watts EL, Gunter MJ, Murphy N, Papier
K, Perez-Cornago A, Schmidt JA, Smith Byrne K, Travis RC,
Key TJ (2020) Circulating insulin-like growth factor-I concentra-
tions and risk of 30 cancers: prospective analyses in UK biobank.
Can Res 80(18):4014–4021. https:// doi. org/ 10. 1158/ 0008- 5472.
Can- 20- 1281
18. Zhang YB, Chen C, Pan XF, Guo J, Li Y, Franco OH, Liu G,
Pan A (2021) Associations of healthy lifestyle and socioeconomic
status with mortality and incident cardiovascular disease: two pro-
spective cohort studies. BMJ 373:n604. https:// doi. org/ 10. 1136/
bmj. n604
19. Wan Z, Guo J, Pan A, Chen C, Liu L, Liu G (2021) Association
of serum 25-Hydroxyvitamin D concentrations with all-cause and
cause-specific mortality among individuals with diabetes. Diabe-
tes Care 44(2):350–357. https:// doi. org/ 10. 2337/ dc20- 1485
20. Lin XJ, Wang CP, Liu XD, Yan KK, Li S, Bao HH, Zhao LY, Liu
X (2014) Body mass index and risk of gastric cancer: a meta-
analysis. Jpn J Clin Oncol 44(9):783–791. https:// doi. org/ 10. 1093/
jjco/ hyu082
21. Lourida I, Hannon E, Littlejohns TJ, Langa KM, Hyppönen
E, Kuzma E, Llewellyn DJ (2019) Association of lifestyle and
genetic risk with incidence of dementia. JAMA 322(5):430–437.
https:// doi. org/ 10. 1001/ jama. 2019. 9879
22. Pei YF, Zhang L (2021) Is the townsend deprivation index a relia-
ble predictor of psychiatric disorders? Biol Psychiatry 89(9):839–
841. https:// doi. org/ 10. 1016/j. biops ych. 2021. 02. 006
23. Han H, Cao Y, Feng C, Zheng Y, Dhana K, Zhu S, Shang C,
Yuan C, Zong G (2022) Association of a healthy lifestyle with all-
cause and cause-specific mortality among individuals with type
2 diabetes: a prospective study in UK biobank. Diabetes Care
45(2):319–329. https:// doi. org/ 10. 2337/ dc21- 1512
24. Song Z, Yang R, Wang W, Huang N, Zhuang Z, Han Y, Qi L, Xu
M, Tang YD, Huang T (2021) Association of healthy lifestyle
including a healthy sleep pattern with incident type 2 diabetes
mellitus among individuals with hypertension. Cardiovasc Dia-
betol 20(1):239. https:// doi. org/ 10. 1186/ s12933- 021- 01434-z
25. Melamed ML, Michos ED, Post W, Astor B (2008) 25-hydroxyvi-
tamin D levels and the risk of mortality in the general population.
Arch Intern Med 168(15):1629–1637. https:// doi. org/ 10. 1001/
archi nte. 168. 15. 1629
26. Dudenkov DV, Mara KC, Petterson TM, Maxson JA, Thacher TD
(2018) Serum 25-hydroxyvitamin d values and risk of all-cause
and cause-specific mortality: a population-based cohort study.
Mayo Clin Proc 93(6):721–730. https:// doi. org/ 10. 1016/j. ma yocp.
2018. 03. 006
27. Dai L, Liu M, Chen L (2021) Association of serum 25-Hydroxy-
vitamin D concentrations with all-cause and cause-specific mor-
tality among adult patients with existing cardiovascular disease.
Front Nutri 8:740855. https:// doi. org/ 10. 3389/ fnut. 2021. 740855
28. Lu Q, Wan Z, Guo J, Liu L, Pan A, Liu G (2021) Association
between serum 25-hydroxyvitamin D concentrations and mor-
tality among adults with prediabetes. J Clin Endocrinol Metab
106(10):e4039–e4048. https:// doi. org/ 10. 1210/ clinem/ dgab4 02
29. Dejaeger M, Antonio L, Bouillon R, Moors H, Wu FCW, O’Neill
TW, Huhtaniemi IT, Rastrelli G, Forti G, Maggi M, Casanueva
FF, Slowikowska-Hilczer J, Punab M, Gielen E, Tournoy J, Van-
derschueren D (2022) Aging men with insufficient vitamin D
have a higher mortality risk: no added value of its free fractions
or active form. J Clin Endocrinol Metab 107(3):e1212–e1220.
https:// doi. org/ 10. 1210/ clinem/ dgab7 43
30. Manson JE, Cook NR, Lee IM, Christen W, Bassuk SS, Mora S,
Gibson H, Gordon D, Copeland T, D’Agostino D, Friedenberg
G, Ridge C, Bubes V, Giovannucci EL, Willett WC, Buring JE
(2019) Vitamin D supplements and prevention of cancer and car-
diovascular disease. N Engl J Med 380(1):33–44. https:// doi. org/
10. 1056/ NEJMo a1809 944
31. Zittermann A, Ernst JB, Prokop S, Fuchs U, Dreier J, Kuhn
J, Knabbe C, Birschmann I, Schulz U, Berthold HK, Pilz S,
Gouni-Berthold I, Gummert JF, Dittrich M, Börgermann J
(2017) Effect of vitamin D on all-cause mortality in heart fail-
ure (EVITA): a 3-year randomized clinical trial with 4000 IU
vitamin D daily. Eur Heart J 38(29):2279–2286. https:// doi. org/
10. 1093/ eurhe artj/ ehx235
32. Stubbins RE, Hakeem A, Núñez NP (2012) Using components
of the vitamin D pathway to prevent and treat colon cancer. Nutr
Rev 70(12):721–729. https:// doi. org/ 10. 1111/j. 1753- 4887. 2012.
00522.x
33. Feng Q, Zhang H, Dong Z, Zhou Y, Ma J (2017) Circulating
25-hydroxyvitamin D and lung cancer risk and survival: a dose-
response meta-analysis of prospective cohort studies. Medicine
96(45):e8613. https:// doi. org/ 10. 1097/ md. 00000 00000 008613
34. Darling AL, Abar L, Norat T (2016) WCRF-AICR continu-
ous update project: systematic literature review of prospective
studies on circulating 25-hydroxyvitamin D and kidney cancer
risk. J Steroid Biochem Mol Biol 164:85–89. https:// doi. org/ 10.
1016/j. jsbmb. 2015. 10. 001
35. Jenab M, Bueno-de-Mesquita HB, Ferrari P, van Duijnhoven FJ,
Norat T, Pischon T, Jansen EH, Slimani N, Byrnes G, Rinaldi
S, Tjønneland A, Olsen A, Overvad K, Boutron-Ruault MC,
Clavel-Chapelon F, Morois S, Kaaks R, Linseisen J, Boeing
H, Bergmann MM, Trichopoulou A, Misirli G, Trichopou-
los D, Berrino F, Vineis P, Panico S, Palli D, Tumino R, Ros
MM, van Gils CH, Peeters PH, Brustad M, Lund E, Tormo MJ,
Ardanaz E, Rodríguez L, Sánchez MJ, Dorronsoro M, Gonza-
lez CA, Hallmans G, Palmqvist R, Roddam A, Key TJ, Khaw
KT, Autier P, Hainaut P, Riboli E (2010) Association between
pre-diagnostic circulating vitamin D concentration and risk of
colorectal cancer in European populations:a nested case-control
study. BMJ 340:b5500. https:// doi. org/ 10. 1136/ bmj. b5500
36. Sofianopoulou E, Kaptoge SK, Afzal S, Jiang T, Gill D, Gun-
dersen TE, Bolton TR, Allara E, Arnold MG, Mason AM,

2592 European Journal of Nutrition (2023) 62:2581–2592
1 3
Chung R, Pennells LAM, Shi F, Sun L, Willeit P, Forouhi NG,
Langenberg C, Sharp SJ, Panico S, Engström G, Melander O,
Tong TYN, Perez-Cornago A, Norberg M, Johansson I, Katzke
V, Srour B, Sánchez MJ, Redondo-Sánchez D, Olsen A, Dahm
CC, Overvad K, Brustad M, Skeie G, Conchi Moreno-Iribas
N, Onland-Moret C, van der Schouw YT, Tsilidis KK, Heath
AK, Agnoli C, Krogh V, de Boer IH, Kobylecki CJ, Çolak Y,
Zittermann A, Sundström J, Welsh P, Weiderpass E, Aglago
EK, Ferrari P, Clarke R, Boutron M-C, Severi G, MacDon-
ald C, Providencia R, Masala G, Ros RZ, Jolanda Boer WM,
Verschuren M, Cawthon P, Schierbeck LL, Cooper C, Schulze
MB, Bergmann MM, Hannemann A, Kiechl S, Brenner H, van
Schoor NM, Albertorio JR, Sacerdote C, Linneberg A, Kårhus
LL, Huerta JM, Imaz L, Joergensen C, Ben-Shlomo Y, Lun-
dqvist A, Gallacher J, Sattar N, Wood AM, Wareham NJ, Nor-
destgaard BG, Di Angelantonio E, Danesh J, Butterworth AS,
Burgess S (2021) Estimating dose-response relationships for
vitamin D with coronary heart disease, stroke, and all-cause
mortality: observational and Mendelian randomisation analyses.
Lancet Diabet Endocrinol 9(12):837–846
37. Muñoz A, Grant WB (2022) Vitamin D and cancer: an historical
overview of the epidemiology and mechanisms. Nutrients. https://
doi. org/ 10. 3390/ nu140 71448
38. Merke J, Milde P, Lewicka S, Hügel U, Klaus G, Mangelsdorf
DJ, Haussler MR, Rauterberg EW, Ritz E (1989) Identification
and regulation of 1,25-dihydroxyvitamin D3 receptor activity and
biosynthesis of 1,25-dihydroxyvitamin D3. Studies in cultured
bovine aortic endothelial cells and human dermal capillaries. J
Clin Invest 83(6):1903–1915. https:// doi. org/ 10. 1172/ jci11 4097
39. Herr C, Greulich T, Koczulla RA, Meyer S, Zakharkina T, Bran-
scheidt M, Eschmann R, Bals R (2011) The role of vitamin D in
pulmonary disease: COPD, asthma, infection, and cancer. Respir
Res 12(1):31. https:// doi. org/ 10. 1186/ 1465- 9921- 12- 31
40. Yue CY, Ying CM (2020) Sufficience serum vitamin D before
20 weeks of pregnancy reduces the risk of gestational dia-
betes mellitus. Nutr Metab 17:89. https:// doi. org/ 10. 1186/
s12986- 020- 00509-0
41. Zhu C, Wang Z, Cai J, Pan C, Lin S, Zhang Y, Chen Y, Leng M,
He C, Zhou P, Wu C, Fang Y, Li Q, Li A, Liu S, Lai Q (2021)
VDR signaling via the enzyme NAT2 inhibits colorectal cancer
progression. Front Pharmacol 12:727704. https:// doi. org/ 10. 3389/
fphar. 2021. 727704
42. Wang L, Zhou S, Guo B (2020) Vitamin D suppresses ovarian
cancer growth and invasion by targeting long non-coding RNA
CCAT2. Int J Mol Sci. https:// doi. org/ 10. 3390/ ijms2 10723 34
43. Salehi-Tabar R, Memari B, Wong H, Dimitrov V, Rochel N, White
JH (2019) The tumor suppressor FBW7 and the Vitamin D recep-
tor are mutual cofactors in protein turnover and transcriptional
regulation. Mole Cancer Res MCR 17(3):709–719. https:// doi.
org/ 10. 1158/ 1541- 7786. Mcr- 18- 0991
44. Liu J, Liu Y, Li H, Wei C, Mao A, Liu W, Pan G (2022) Poly-
phyllin D induces apoptosis and protective autophagy in breast
cancer cells through JNK1-Bcl-2 pathway. J Ethnopharmacol
282:114591. https:// doi. org/ 10. 1016/j. jep. 2021. 114591
45. Wu X, Hu W, Lu L, Zhao Y, Zhou Y, Xiao Z, Zhang L, Zhang
H, Li X, Li W, Wang S, Cho CH, Shen J, Li M (2019) Repurpos-
ing vitamin D for treatment of human malignancies via targeting
tumor microenvironment. Acta pharmaceutica Sinica B 9(2):203–
219. https:// doi. org/ 10. 1016/j. apsb. 2018. 09. 002
46. Markowska A, Antoszczak M, Kojs Z, Bednarek W, Markowska
J, Huczyński A (2020) Role of vitamin D(3) in selected malig-
nant neoplasms. Nutrition (Burbank, Los Angeles County, Calif)
79-80:110964. https:// doi. org/ 10. 1016/j. nut. 2020. 110964
47. Bhutia SK (2022) Vitamin D in autophagy signaling for health and
diseases: insights on potential mechanisms and future perspec-
tives. J Nutri Biochem 99:108841. https:// doi. org/ 10. 1016/j. jnu tb
io. 2021. 108841
48. Sharma K, Goehe RW, Di X, Hicks MA 2nd, Torti SV, Torti
FM, Harada H, Gewirtz DA (2014) A novel cytostatic form of
autophagy in sensitization of non-small cell lung cancer cells
to radiation by vitamin D and the vitamin D analog, EB 1089.
Autophagy 10(12):2346–2361. https:// doi. org/ 10. 4161/ 15548 627.
2014. 993283
49. Fernández-Barral A, Bustamante-Madrid P, Ferrer-Mayorga G,
Barbáchano A, Larriba MJ, Muñoz A (2020) Vitamin D Effects
on cell differentiation and stemness in cancer. Cancers (Basel)
12(9):2413. https:// doi. org/ 10. 3390/ cance rs120 92413
50. Fernandez-Garcia NI, Palmer HG, Garcia M, Gonzalez-Martin A,
del Rio M, Barettino D, Volpert O, Muñoz A, Jimenez B (2005)
1alpha,25-Dihydroxyvitamin D3 regulates the expression of Id1
and Id2 genes and the angiogenic phenotype of human colon car-
cinoma cells. Oncogene 24(43):6533–6544. https:// doi. org/ 10.
1038/ sj. onc. 12088 01
51. Vanoirbeek E, Eelen G, Verlinden L, Carmeliet G, Mathieu C,
Bouillon R, O’Connor R, Xiao G, Verstuyf A (2014) PDLIM2
expression is driven by vitamin D and is involved in the pro-
adhesion, and anti-migration and -invasion activity of vitamin D.
Oncogene 33(15):1904–1911. https:// doi. org/ 10. 1038/ onc. 2013.
123
52. Chun RF, Liu PT, Modlin RL, Adams JS, Hewison M (2014)
Impact of vitamin D on immune function: lessons learned from
genome-wide analysis. Front Physiol 5:151. https:// doi. org/ 10.
3389/ fphys. 2014. 00151
53. Sun Q, Pan A, Hu FB, Manson JE, Rexrode KM (2012)
25-Hydroxyvitamin D levels and the risk of stroke: a prospective
study and meta-analysis. Stroke 43(6):1470–1477. https:// doi. org/
10. 1161/ strok eaha. 111. 636910
54. Degerud E, Hoff R, Nygård O, Strand E, Nilsen DW, Nordre-
haug JE, Midttun Ø, Ueland PM, de Vogel S, Dierkes J (2016)
Cosinor modelling of seasonal variation in 25-hydroxyvitamin D
concentrations in cardiovascular patients in Norway. Eur J Clin
Nutr 70(4):517–522. https:// doi. org/ 10. 1038/ ejcn. 2015. 200
55. Dan L, Chen X, Xie Y, Sun Y, Hesketh T, Wang X, Chen J (2022)
Nonlinear association between serum 25-Hydroxyvitamin D and
all-cause mortality in adults with inflammatory bowel disease in
a prospective cohort study. J Nutr 152(9):2125–2134. https:// doi.
org/ 10. 1093/ jn/ nxac1 48
Springer Nature or its licensor (e.g. a society or other partner) holds
exclusive rights to this article under a publishing agreement with the
author(s) or other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the terms of
such publishing agreement and applicable law.