A prospective study of dietary and supplemental zinc intake and risk of type 2 diabetes depending on genetic variation in SLC30A8
Drake et al. Genes & Nutrition
A prospective study of dietary and supplemental zinc intake and risk of type 2 diabetes depending on genetic variation in SLC30A8
Isabel Drake 0
George Hindy 0
Ulrika Ericson 0
Marju Orho-Melander 0
0 Diabetes and Cardiovascular Disease - Genetic Epidemiology, Lund University Diabetes Centre, Department of Clinical Sciences in Malmö, Lund University, Clinical Research Center 60:13, Jan Waldenströms gata 35 , SE-205 02 Malmö , Sweden
Background: The solute carrier family 30 member 8 gene (SLC30A8) encodes a zinc transporter in the pancreatic beta cells and the major C-allele of a missense variant (rs13266634; C/T; R325W) in SLC30A8 is associated with an increased risk of type 2 diabetes (T2D). We hypothesized that the association between zinc intake and T2D may be modified by the SLC30A8 genotype. Results: We carried out a prospective study among subjects with no history cardio-metabolic diseases in the Malmö Diet and Cancer Study cohort (N = 26,132, 38% men; 86% with genotype data). Zinc intake was assessed using a diet questionnaire and food record. During a median follow-up of 19 years, 3676 T2D cases occurred. A BMI-stratified Cox proportional hazards regression model with attained age as the time scale was used to model the association between total and dietary zinc intake, zinc supplement use, zinc to iron ratio, and risk of T2D adjusting for putative confounding factors. The median total zinc intake was 11.4 mg/day, and the median dietary zinc intake was 10.7 mg/day. Zinc supplement users (17%) had a median total zinc intake of 22.4 mg/day. Dietary zinc intake was associated with increased risk of T2D (Ptrend < 0.0001). In contrast, we observed a lower risk of T2D among zinc supplement users (HR = 0.79, 95% CI 0.70-0.89). The SLC30A8 CC genotype was associated with a higher risk of T2D (HR = 1.16, 95% CI 1.07-1.24), and the effect was stronger among subjects with higher BMI (Pinteraction = 0.007). We observed no significant modification of the zinc-T2D associations by SLC30A8 genotype. However, a three-way interaction between SLC30A8 genotype, BMI, and zinc to iron ratio was observed (Pinteraction = 0.007). A high zinc to iron ratio conferred a protective associated effect on T2D risk among obese subjects, and the effect was significantly more pronounced among T-allele carriers. Conclusions: Zinc supplementation and a high zinc to iron intake ratio may lower the risk of T2D, but these associations could be modified by obesity and the SLC30A8 genotype. The findings implicate that when considering zinc supplementation for T2D prevention, both obesity status and SLC30A8 genotype may need to be accounted for.
Gene-nutrient interaction; Zinc; Solute carrier family 30 member 8 gene; Single nucleotide polymorphism; Cohort; Body mass index
Prevalence of type 2 diabetes (T2D) is increasing
worldwide attributable to aging populations and increasing
prevalence of obesity due to obesogenic environmental
]. While both genetic and environmental factors
are considered to increase risk, the most important targets
for prevention of obesity and T2D are modifiable factors
including diet and physical activity [
]. Zinc is an essential
trace element with both structural and functional roles in
many cellular proteins and enzymes. It plays a vital role in
β-cell physiology and insulin action [
]. In animal
studies, zinc intake has been suggested to be protective of
T2D . Two studies in women examining the association
between zinc intake and prospective risk of T2D indicated
that high dietary zinc intake and/or a high zinc to iron
ratio was associated with a lower risk of T2D [
Placebo-controlled randomized trials have found that zinc
supplementation modestly reduces fasting glucose and
hemoglobin A1c (HbA1c), particularly in patients with
T2D . Due to limited number of controlled trials, there
is to date no overall strong evidence supporting that zinc
supplementation may lower the risk of T2D in humans
]. One of the most consistently replicated
singlenucleotide variants increasing the risk of T2D is
rs13266634 (C/T) in the SLC30A8 gene that encodes zinc
transporter 8, which is mainly expressed in the pancreatic
β-cells. The major C-allele of rs13266634 is associated
with a lower early insulin response to glucose and a higher
risk of T2D [
]. As the rs13266634 is a
nonsynonymous variant, the effect of zinc on T2D risk could
plausibly vary across this genotype. Indeed, a recent
casecontrol study observed an interaction between plasma
zinc levels and rs13266634 on T2D risk [
prospective studies investigating whether the association
between zinc intake and risk of T2D differ depending on
rs13266634 genotype are lacking. Given the biological
plausibility for a potential interaction between zinc intake
and SLC30A8 rs13266634 on T2D risk, and the scarcity of
large prospective studies examining the role of zinc intake
in relation to T2D risk, we wanted to examine these
associations in a large population-based prospective cohort of
middle-aged Swedish men and women. As supplemental
zinc is more bioavailable than zinc from food sources [
and as iron has been shown to inhibit the absorption of
], we set out to examine zinc intake from both
food sources and supplements, the zinc to iron ratio in
relation to T2D, and particularly, potential effect
modification by the SLC30A8 rs13266634 genotype.
The Malmö Diet and Cancer Study (MDCS) is a
population-based prospective cohort set in the south of
]. Baseline examinations were carried out
between March 1991 and October 1996. The source
population included all persons living in the city of Malmö
born between 1926 and 1945 and with sufficient Swedish
reading and writing skills. With a participation rate of
40%, 28,098 participants completed all baseline
]. From this study population, we excluded
subjects with a history of prevalent diabetes mellitus or
cardiovascular disease, resulting in a total study
population of 26,132. A flow chart of the analytical study
population, including information on covariates with missing
data, is shown in Additional file 1: Figure S1.
Follow-up and case ascertainment
All subjects were followed from entry into the study
until their date of diabetes mellitus diagnosis, date of
death from any cause, emigration, or end of follow-up
(31 December 2014), whichever came first. Information
on vital and emigration status of the study participants
was obtained from the Swedish Cause of Death Registry
and the Swedish Tax Agency. Diabetes status at baseline
and during follow-up and information on date of
diabetes diagnosis were identified from seven registers as
well as baseline and re-examination screenings of the
MDCS and the Malmö Preventive Project [
National Diabetes Register [
] and the regional
Diabetes 2000 Register [
] required a proven diagnosis
by a physician at the hospital based on international
standards for diagnosis (i.e., fasting plasma glucose
concentration ≥ 7.0 mmol/l measured twice). For cases not
diagnosed at a hospital, the local HbA1c register from
the Department of Clinical Chemistry, Skåne University
Hospital, Malmö, was used [
]. Other registries
used to identify diabetes cases included the National
Patient Register, the Swedish Cause of Death Register
(ICD10 codes E10-E14 and O244-O249), and the
Prescribed Drug Register (ATC code A10). The different
sources of case ascertainment were overlapping. Of the
total 3831 diabetes diagnoses identified during
followup, 2397 cases were captured by the National Diabetes
Register. Information on diabetes type was lacking for
most of the cases (54%). For subjects where type was
specified, we censored cases of type 1 diabetes (n = 136),
LADA (n = 9), secondary diabetes (n = 1), and other
(n = 9) at the date of diagnosis. For the remaining cases
(n = 3676), given the age distribution at diagnosis, we
assumed that they were T2D.
Genotyping was performed at the Clinical Research
Centre, Malmö, Sweden, using Sequenom MassARRAY
iPLEX (Sequenom, San Diego, CA, USA) according to
the manufacturers’ instructions. The concordance rate
was >99% in a subset of 5500 samples which were
additionally genotyped using Human Omni Express Exome
Bead Chip Kit (Illumina, San Diego, CA, USA).
Genotyping success rate was 97.3%. No deviation from the
Hardy-Weinberg equilibrium was observed (P = 0.82). In
total, 20,929 subjects in the current study population had
genotype data for the SLC30A8 rs13266634 variant.
Dietary assessment was conducted using a modified diet
history method including a 168-item diet questionnaire
(using exact frequencies and pictures to assess portion
sizes), a 7-day food record (in which descriptions of
prepared meals, nutrient supplements, and cold
beverages were collected), and a 1-h dietary interview. Data
on the validity [
] and reproducibility  of the
method have been published. Energy and nutrient
intakes were computed using the MDC Food and
Nutrient Database, mainly originating from the PC
Kost2-93 food database of the Swedish National Food
Administration. Use of dietary supplements (including
type and amount) was registered in the 7-day food
record and categorized as any diet supplement use (yes/no)
and zinc supplement use (yes/no). Total zinc and iron
intakes (mg/day) were estimated by combining intake
from both food sources and supplements. The zinc to
iron ratio was calculated as the ratio of total
energyadjusted zinc to energy-adjusted total iron intake. Other
dietary variables included in the main analyses were total
energy intake, alcohol, fiber, fruit and vegetables,
processed meat, sugar-sweetened beverages, and coffee
intake. Further, in sensitivity analysis, we examined
additional adjustments for saturated fatty acids, fish and
shellfish, red meat, total protein (as percentage of total
energy intake), iron, and calcium. Food and nutrient
intakes were energy-adjusted using the residual method
] and participants ranked into quintiles to reduce the
influence of outliers and handle highly skewed intake
levels. Participants were classified as misreporters of
energy intake if the ratio of energy intake to basal
metabolic rate was outside the 95% confidence interval of the
calculated physical activity level [
]. Participants with
potentially unstable food habits over time were identified
using the questionnaire item “Have you previously
changed your food habits substantially due to illness or other
reason?” (yes/no) [
]. Season of dietary data collection
(January–March, April–June, July–September, October–
December) was adjusted for as a categorical variable to
account for seasonal variation in reported food intakes.
Calendar year of study entry was adjusted for as a
categorical variable to account for the recruitment of
slightly older individuals during the last 2 years of
baseline examinations and to account for a minor change in
coding routines for the dietary assessment in September
Lifestyle and other variables
Age and sex were determined by the participants’
Swedish personal identification numbers. Nurses
measured height (cm) and weight (kg) with subjects wearing
light indoor clothing with no shoes. Body mass index
(BMI) was calculated as weight divided by height
squared (kg/m2). BMI was classified as normal-weight
(< 25 kg/m2), overweight (25–29.9 kg/m2), and obese (≥
30 kg/m2). Information on socioeconomic and lifestyle
factors was collected from the MDC baseline
]. Educational level was categorized as
elementary, primary and secondary, upper secondary, further
education without a degree, or university/college degree.
Smoking status was defined as never, former, or current
(including irregular smokers). Leisure time physical
activity was estimated based on an adaptation of the
Minnesota Leisure Time Physical Activity Instrument
]. Participants estimated the number of minutes
per week spent on 17 different physical activities for
each of the four seasons, and a score was calculated by
multiplying an intensity factor with the duration of each
activity. The score was categorized into sex-specific
Baseline characteristics across quintiles of
energyadjusted total, dietary, zinc to iron ratio, and zinc
supplement use (yes/no) were examined. Mean values
(standard deviation, SD) of continuous variables were
calculated for each category of zinc intake. For skewed
dietary variables, geometric mean values (95%
confidence intervals, CIs) were calculated. Further, we
examined the baseline characteristics by T2D status and
SLC30A8 genotype. We used a BMI-stratified Cox
proportional hazards regression models with attained age as
the underlying time metric to estimate hazard ratios
(HRs) and 95% CIs for the association between the
different measures of zinc intake in relation to risk of
incident T2D. We examined the non-linear association
between total zinc intake and risk of T2D by fitting a
restricted cubic spline with three knots chosen according
to Harrell’s recommended percentiles to a Cox
proportional hazards regression model using a continuous
variable for total zinc intake (mg/day). Extreme intakes
(> 50 mg/day) were excluded from this analysis. The
main multivariable model included both dietary (total
energy intake, season of dietary data collection, dietary
fiber, alcohol consumption, fruit and vegetables,
processed meats, sugar-sweetened beverages, coffee intake,
and any diet supplement use) and non-dietary covariates
(age, calendar year of study entry, BMI, sex, educational
level, smoking status, leisure time physical activity).
Sensitivity analyses included exclusion of subjects
classified as energy misreporters and those reporting
substantial dietary changes in the past. For all presented
Cox models, the proportional hazards (PH) assumption
was fulfilled as determined by the Schoenfeld test. Since
including BMI as a covariate violated the PH
assumption, we present the BMI-stratified Cox model for the
main analyses and further examined potential
interactions with BMI. In the MDCS, sufficient statistical power
to examine dietary hypotheses (i.e., 80% and α = 0.05)
assuming a case-control design with three controls per
case and a validity coefficient of the dietary variable of 0.6
was reached when more than 283 cases had accumulated.
Statistical power to detect an interaction between
rs13266634 genotype with zinc intake was examined using
Quanto version 1.2.4 (http://biostats.usc.edu/software).
In this study, we had 80% power to detect an
interaction odds ratio of at least 1.30 for T2D. Stata SE/14.2
for Mac (StataCorp, College Station, TX) was used for
all statistical analyses. All reported tests were
twosided, and P values < 0.05 were considered statistically
Baseline characteristics of the study population
During a median follow-up time of 19 years, we
identified 3676 incident cases of T2D. The mean age at
diagnosis was 69.1 years (SD = 8.1). In total, 4417 subjects
(17%) reported use of supplements containing zinc.
Subjects with high dietary zinc intake were more likely to be
younger, be male, and have higher BMI and educational
level and were less likely to be current smokers
compared to subjects with low dietary zinc intake. Zinc
supplement users were more likely to be female, older, and
current smokers and to have lower educational level
compared to non-users (Table 1). Baseline characteristics
of the study population by T2D status and SLC30A8
rs13266634 genotype are shown in Additional file 1:
Tables S1 and S2, respectively.
Zinc intake and risk of T2D
In multivariable analyses, we observed that total zinc
intake was non-linearly associated with risk of T2D
(Fig. 1; Table 2). The increased risk associated with total
zinc intake appeared to be driven by a linear positive
association between dietary zinc intake and T2D (Table 2).
This association was attenuated, but nominally
significant, after adjusting for both dietary and non-dietary
confounders, and excluding potential energy
misreporters and subjects with unstable food habits (Q5 versus
Q1: HR = 1.27, 95% CI 1.06–1.51; Ptrend = 0.001). With
additional adjustment for total protein intake, total iron
intake, and intake of red meat, the association was
further attenuated and no longer significant (Q5 versus Q1:
HR = 1.07, 95% CI 0.88–1.30; Ptrend = 0.42; data not
tabulated). Zinc supplement use was found to associate
with a lower risk of T2D (HR = 0.79, 95% CI 0.70–0.89),
and the inverse association remained significant in
multivariable analyses (Table 2). We further observed that a
high zinc to iron ratio was associated with a marginally
lower risk of T2D (Q5 versus Q1: HR = 0.91 (95% CI
0.81–1.02; Ptrend = 0.02), although the association was
attenuated in multivariable and sensitivity analyses
(Table 2). Additional adjustment for total iron intake did
not affect the observed association.
SLC30A8 rs13266634, zinc intake, and risk of type 2 diabetes
The CC genotype of SLC30A8 rs13266634, compared to
CT/TT, was associated with a higher risk of T2D
(HR = 1.16, 95% CI 1.07–1.24, P = 9.6 × 10−5)
(Additional file 1: Figure S2). We observed no significant
multiplicative interactions between quintiles of total zinc
intake (Pinteraction = 0.83), dietary zinc intake (Pinteraction = 0.53),
total zinc to iron ratio (Pinteraction = 0.32), or zinc supplement
use (yes/no; Pinteraction = 0.44) and the SLC30A8 genotype in
the fully adjusted model including both dietary and
nondietary factors. However, the lowest risk was observed
among the group of zinc supplement users with the TT
genotype (HR = 0.59, 95% CI 0.39–0.88) compared to
nonsupplement users with the CC genotype (Fig. 2). Further, a
high zinc to iron ratio was non-significantly associated with
lower risk of T2D among CC/CT genotypes while a clear
protective association was observed among the smaller
group of TT carriers (Ptrend = 0.009; data not tabulated).
SLC30A8 rs13266634, BMI, zinc intake, and risk of type 2 diabetes
In a non-stratified Cox model including BMI as a
continuous covariate and adjusting for dietary and non-dietary
factors (same as those presented in model 3, Table 2), we
observed no interaction between BMI and total zinc intake
(Pinteraction = 0.08), dietary zinc intake (Pinteraction = 0.32), or
zinc supplement use (Pinteraction = 0.15) on the risk of T2D.
However, we did observe a significant interaction between
BMI and the zinc to iron ratio (Pinteraction = 0.005; Fig. 3).
The zinc to iron ratio was only significantly associated with
lower risk of T2D among subjects with high BMI. The HR
for the highest tertile of zinc to iron ratio compared to the
lowest tertile among obese subjects (BMI >30 kg/m2) was
0.70 (95% CI 0.60–0.82; Ptrend = 1.4 × 10−5). We further
found that the T2D risk-increasing CC genotype was not
associated with an increased risk among normal-weight
subjects (HR = 1.09, 95% CI 0.94–1.26). Among overweight
(HR = 1.15, 95% CI 1.04–1.28) and obese (HR = 1.31, 95%
CI 1.14–1.50) subjects, the CC genotype was associated
with a higher risk of T2D. There was a significant
interaction between the SLC30A8 genotype and BMI after
adjustment for potential confounders (Pinteraction = 0.007;
Fig. 4). The interaction between BMI and SLC30A8
ce en 7 7 )3 .3 .12 9 )1 8 )7
an em seY 414 .12 .(8 26 (1 .01 .(3 .01 .(7
n su 5 )
t c ,7 2 )4 .4 .16 2 )4
ie inZ seu oN 12 .11 .(3 16 (1 .11 .(3 –
Q 52 .15 .(6 19 (1
5 )1 .0 .07 8 )8 5 )1
is ) in
r 6 c
te 9 in
c 9 z
a 1 l
ra –1 ta
frseuuobm il(tzcaongm il(rtaonogm ii(rtzcyaenm )ya lisceuZnppm )(/yaagdmm rsseu ryaehopgDm )(rsyaeegA l(sxaee%M 2I(/kBgMm (52<% –.99252 ≥(03% isyhhpgH ii(ttycva% rrteunC (rskeo%m ieuhgd l)(ve%
N T T D d
.1 –.00 .)4
.5 –.99 .)2 .4 –.31 .)8 ie
4 3 9 2 2 1 .5 d
.3 –.46 .)1 .6 –.58 .)6 b
.2 –.08 .)9 9 9 4 4 9 3 .0 ira
95 (9 99 29 (2 03 40 (3 14 68 a
.5 –.65 .)5 .5 –.93 .)9 o
.7 –.75 .)2 g
1 5 6 8 0 7 .4 te
60 (5 64 36 (3 63 41 (4 24 87 a
.9 –.10 .)9 .9 –.07 .)3
.8 –.89 .)7 .4 –.61 .)9
c e s .5 –.26 .)4 .3 –.97 .)2
69 (6 72 33 (3 33 41 (4 24 68 ic
sc o k
it (C ta
c 9 iz
ra 1 l
a –1 ta
ssed )y -sew saeg nd lsaeb (/gd
e a r r a t e
rco /d aug vee irtu eeg ffeo
P (g S b F v C
Total zinc intake (mg/day)
genotype also remained after further adjustment for
additional dietary confounders (Pinteraction = 0.008; data not
tabulated). We observed a significant three-way interaction
between the SLC30A8 genotype, BMI, and the zinc to iron
ratio (Pinteraction = 0.007; Fig. 5). A high zinc to iron ratio
was strongly associated with a lower risk of T2D among
obese subjects with the CT/TT genotype compared to the
CC genotype (HR = 0.56, 95% CI 0.42–0.73; Ptrend = 1.5 × 10
−5; Fig. 5). A non-significant three-way interaction was
similarly observed with zinc supplement use, BMI, and
SLC30A8 genotype (Pinteraction = 0.11), in that the combined
protective association of the T-allele and zinc supplement
use was stronger among obese subjects. No evidence for
effect modification by BMI on the SLC30A8 interaction
with total zinc intake or dietary zinc intake on risk of T2D
In this prospective cohort study, we found that zinc
supplement use may confer a protective effect on risk of
T2D after adjustment for known and putative
confounders. We observed that BMI modified the
association between the SLC30A8 genotype and T2D risk. In
addition, a higher zinc to iron ratio was associated with
a lower risk of T2D, particularly among overweight and
obese subjects with the CT/TT genotypes of SLC30A8
rs13266634. Surprisingly, a high dietary zinc intake is
associated with higher risk of T2D in our study
population. After additional adjustment for protein intake, iron
intake, and red meat, there was a null association
between dietary zinc intake and T2D risk in this study
population. This finding may suggest that dietary zinc
intake within the normal range does not influence risk
of T2D without fully accounting for other nutrients
which affect the bioavailability of zinc (e.g., iron).
Two prospective studies in women have previously
reported a lower risk of T2D among subjects with a high
zinc to iron ratio [
]. In contrast to these studies, we
observed no protective effect of dietary zinc intake, but a
lower risk of T2D only among zinc supplement users.
Zinc supplementation has in randomized trials been
shown to lower fasting glucose and HbA1c , yet no
strong evidence exists showing that this would translate
to a lower risk of T2D [
]. In a case-control study,
plasma zinc concentration was observed to interact with
the SLC30A8 genotype (rs13266634) on T2D risk, such
that the protective effect of higher plasma zinc was
stronger among subjects with the TT genotype [
Further, in a meta-analysis of 14 cohorts, including the
cardiovascular sub-cohort of the MDCS (N = 4867),
higher total zinc associated with lower fasting glucose
levels and this association was stronger among A-allele
carriers of rs11558471 SLC30A8 variant (in strong
linkage disequilibrium with rs13266634) [
]. In a
genotype-based clinical trial in a healthy Amish
population, a 14-day zinc supplementation (50 mg two times
per day) resulted in improved early insulin response
after an intravenous glucose load among subjects with
the CT/TT genotype of rs13266634 [
]. The results
from these studies along with the results from our study
suggest that zinc interventions may benefit from
considering SLC30A8 rs13266634 genotype. To our knowledge,
an interaction between BMI and SLC30A8 genotype on
prospective risk of T2D has not been shown earlier. Our
finding that the effect of the risk-increasing CC genotype
on T2D risk was stronger among obese subjects may be
considered to be in contrast to two previous reports.
Timpson et al. reported nominal evidence for a
genomewide effect size heterogeneity for the SLC30A8 locus,
with the rs13266634 SNP showing evidence for an
association between the risk allele and reduced BMI when
restricted to subjects with T2D [
]. In a study by
Cauchi et al., the SLC30A8 variant was only associated with
T2D among non-obese subjects, although there was no
significant heterogeneity [
]. It is impossible to deduce
from this observational setting whether obesity per se
influences the function of the SLC30A8 gene variant.
However, it is plausible to speculate that there may be a
synergistic effect of obesity on T2D risk in relation to
the SLC30A8 variant, potentially driven by the combined
impact of lower insulin secretion and obesity-associated
elevations in glucose levels and insulin resistance. The
interaction between BMI, SLC30A8 genotype, and zinc
to iron ratio does suggest that zinc supplementation
may have potential to lower risk of T2D among obese
subjects and that this effect may be particularly
Number of subjects (T2D cases)
Number of subjects (T2D cases)
Dietary zinc intake
Zinc to iron ratio
Model 1 Cox proportional hazards model with attained age as the time metric and adjusted for sex; Model 2 BMI-stratified Cox proportional hazards model with
attained age as time metric and adjusted for sex, calendar year of study entry, educational level, smoking status, leisure-time physical activity, season of dietary
data collection, and total energy intake; Model 3 as model 2 with additional adjustment for alcohol consumption, dietary fiber, fruit and vegetables, processed
meat, sugar-sweetened beverages, coffee intake and diet supplement use (any); Model 4 as model 3 with exclusion of past food habit changers and potential
energy misreporters (n excluded = 9204
prominent among subjects with the T-allele of
rs13266634. Notably, regardless of genotype, zinc
supplementation was associated with lower risk of T2D in
our population of middle-aged Swedish men and
women. The contrasting findings regarding dietary zinc
intake in this study compared to a previous study within
the Nurses’ Health Study (NHS) cohort [
] could be due
to differences in food sources of zinc as well as food
sources of other nutrients that may impact the
bioavailability of zinc. While the bioavailability of zinc from
supplements is higher than from foods, no additional
benefit of zinc supplements was found in the NHS as
compared to dietary zinc [
]. Similar to differences in
food sources of zinc between these cohorts, the type,
amount, and consumption pattern of diet supplements
may also be very different between populations.
There are several limitations of this study that should
be discussed. Although we adjusted for several known
and putative lifestyle and dietary confounders, it is not
possible to exclude the possibility of residual
confounding explaining the observed associations,
particularly since there are several dietary factors that have
been proposed to influence β-cell function [
addition, isolation of single-nutrient effects in
observational settings is difficult due to collinearity between
dietary intakes and correlated measurement errors, both
which may affect multivariable modeling. A major
limitation of this study is that assessment of dietary and
supplement zinc use was only assessed at the baseline
examinations and relied on self-report. The current
study has several notable strengths. To the best of our
knowledge, this is the first prospective study examining
the potential interaction between SLC30A8 rs13266634
and zinc intake, from both diet and supplements, on
T2D risk. The prospective nature of the study allowed
zinc intake assessment prior to development of T2D
suggesting that any exposure misclassification is
nondifferential. The large sample size with virtually no loss
to follow-up (< 0.5%) assured that we had sufficient
power to detect potential interactions between zinc
intakes, measured using a high-validity dietary
assessment method, and the SLC30A8 genetic variant. In
addition, we were able to access several national and
regional registries in order to identify incident T2D
cases, minimizing potential outcome misclassification.
Any residual misclassification of T2D status would
plausibly attenuate the observed associations and
interactions. Since the MDCS is a population-based cohort,
the results are also likely to be fairly generalizable to
other similar populations [
While there is accumulating evidence to suggest that
rs13266634 impacts diabetes risk and insulin secretion
traits and is affected by total zinc intake and circulating
zinc levels, the functional implications of the rs13266634
variant remain unclear [
]. Indeed, a recent study
found that rare loss-of-function variants in SLC30A8 are
protective against diabetes [
], which may propose that
the rs13266634 C-allele could have some kind of
gainof-function effect. Further, since zinc is an essential
component of numerous proteins and exert independent
actions, zinc supplementation may have far-reaching
]. As noted previously  and based on
our results, other nutrients that affect absorption (e.g.,
iron) or actions of zinc may need to be considered if
aiming for individualized prevention or treatment based
on the rs13266634 genotype.
This study provides novel evidence that zinc supplement
use and/or a high zinc to iron ratio may be associated
with a lower risk of T2D, particularly among subjects
with high BMI, and that the effect may be modified by
the SLC30A8 rs13266634 genotype. The findings
implicate that when using zinc supplementation for T2D
prevention, both obesity status and SLC30A8 genotype may
need to be considered. However, due to the
observational nature of our study, the findings should be
interpreted with caution and are in need of further
Additional file 1: Figure S1. Flow chart of analytical study population.
Figure S2. Nelson-Aalen cumulative hazard estimates for type 2 diabetes
by SLC30A8 rs13266634 genotype among 20,929 participants in the
Malmö Diet and Cancer Study (Plogrank = 0.0005). Table S1. Baseline
characteristics by type 2 diabetes (T2D) status in the Malmö Diet and
Cancer Study at baseline (1991–1996). Table S2. Baseline characteristics by
SLC30A8 genotype (rs13266634) in the Malmö Diet and Cancer Study at
baseline (1991–1996). (DOCX 10736 kb)
BMI: Body mass index; CI: Confidence interval; HR: Hazard ratio;
MDCS: Malmö Diet and Cancer Study; SLC30A8: Solute carrier family 30 member 8; T2D: Type 2 diabetes mellitus
The authors would like to express their sincere gratitude to all the
participants of the Malmö Diet and Cancer Study. We also thank Malin
Svensson for excellent technical assistance and Anders Dahlin for data management.
This study was funded by the Swedish Research Council, the European Research
Council (Consolidator grant no. 649021, Orho-Melander), the Swedish Heart and
Lung Foundation, the Novo Nordic Foundation, the Swedish Diabetes Foundation,
the Påhlsson Foundation, the Region Skåne, Skåne University Hospital, and the
Linnéus Foundation for the Lund University Diabetes Centre. The Malmö Diet and
Cancer Study was initially funded by the Swedish Research Council, the Swedish
Heart and Lung Foundation, and the Swedish Cancer Society. The funding bodies had no role in the design of the study, analysis or interpretation of data.
Availability of data and materials
The datasets used for analysis during the current study are not publicly available but available from the corresponding author on reasonable request.
ID performed all analyses and drafted the first manuscript. ID, GH, UE, and
MOM interpreted the data, and GH, UE, and MOM contributed to the review of the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
In agreement with the Declaration of Helsinki, this study was approved by the Ethical committee at Lund University (LU 51-90) and informed written consent was obtained from all participants.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Submit your next manuscript to BioMed Central
and we will help you at every step:
1. Danaei G , Finucane MM , Lu Y , Singh GM , Cowan MJ , Paciorek CJ , et al. National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants . Lancet . 2011 ; 378 ( 9785 ): 31 - 40 .
2. Qi L , FB H , Hu G . Genes, environment, and interactions in prevention of type 2 diabetes: a focus on physical activity and lifestyle changes . Curr Mol Med . 2008 ; 8 ( 6 ): 519 - 32 .
3. Fukada T , Yamasaki S , Nishida K , Murakami M , Hirano T. Zinc homeostasis and signaling in health and diseases: zinc signaling . J Biol Inorg Chem . 2011 ; 16 ( 7 ): 1123 - 34 .
4. Rutter GA , Chabosseau P , Bellomo EA , Maret W , Mitchell RK , Hodson DJ , et al. Intracellular zinc in insulin secretion and action: a determinant of diabetes risk? Proc Nutr Soc . 2016 ; 75 ( 1 ): 61 - 72 .
5. Taylor CG . Zinc, the pancreas, and diabetes: insights from rodent studies and future directions . Biometals . 2005 ; 18 ( 4 ): 305 - 12 .
6. Sun Q , van Dam RM , Willett WC , FB H. Prospective study of zinc intake and risk of type 2 diabetes in women . Diabetes Care . 2009 ; 32 ( 4 ): 629 - 34 .
7. Vashum KP , McEvoy M , Shi Z , Milton AH , Islam MR , Sibbritt D , et al. Is dietary zinc protective for type 2 diabetes? Results from the Australian longitudinal study on women's health . BMC Endocr Disord . 2013 ; 13 : 40 .
8. Capdor J , Foster M , Petocz P , Samman S. Zinc and glycemic control: a metaanalysis of randomised placebo controlled supplementation trials in humans . J Trace Elem Med Biol . 2013 ; 27 ( 2 ): 137 - 42 .
9. El Dib R , Gameiro OL , Ogata MS , Modolo NS , Braz LG , Jorge EC , et al. Zinc supplementation for the prevention of type 2 diabetes mellitus in adults with insulin resistance . Cochrane Database Syst Rev . 2015 ; 5 : CD005525 .
10. Chimienti F , Favier A , Seve M. ZnT-8, a pancreatic beta-cell-specific zinc transporter . Biometals . 2005 ; 18 ( 4 ): 313 - 7 .
11. Tamaki M , Fujitani Y , Hara A , Uchida T , Tamura Y , Takeno K , et al. The diabetes-susceptible gene SLC30A8/ZnT8 regulates hepatic insulin clearance . J Clin Invest . 2013 ; 123 ( 10 ): 4513 - 24 .
12. Rutter GA , Chimienti F. SLC30A8 mutations in type 2 diabetes . Diabetologia . 2015 ; 58 ( 1 ): 31 - 6 .
13. Cauchi S , Del Guerra S , Choquet H , D'Aleo V , Groves CJ , Lupi R , et al. Metaanalysis and functional effects of the SLC30A8 rs13266634 polymorphism on isolated human pancreatic islets . Mol Genet Metab . 2010 ; 100 ( 1 ): 77 - 82 .
14. Sladek R , Rocheleau G , Rung J , Dina C , Shen L , Serre D , et al. A genomewide association study identifies novel risk loci for type 2 diabetes . Nature . 2007 ; 445 ( 7130 ): 881 - 5 .
15. Shan Z , Bao W , Zhang Y , Rong Y , Wang X , Jin Y , et al. Interactions between zinc transporter-8 gene (SLC30A8) and plasma zinc concentrations for impaired glucose regulation and type 2 diabetes . Diabetes . 2014 ; 63 ( 5 ): 1796 - 803 .
16. Maret W , Sandstead HH . Zinc requirements and the risks and benefits of zinc supplementation . J Trace Elem Med Biol . 2006 ; 20 ( 1 ): 3 - 18 .
17. Lonnerdal B . Dietary factors influencing zinc absorption . J Nutr . 2000 ; 130 ( 5S Suppl) : 1378S - 83S .
18. Solomons NW . Competitive interaction of iron and zinc in the diet: consequences for human nutrition . J Nutr . 1986 ; 116 ( 6 ): 927 - 35 .
19. Berglund G , Elmstahl S , Janzon L , Larsson SA . The Malmo Diet and Cancer Study . Design and feasibility . J Intern Med . 1993 ; 233 ( 1 ): 45 - 51 .
20. Manjer J , Carlsson S , Elmstahl S , Gullberg B , Janzon L , Lindstrom M , et al. The Malmo Diet and Cancer Study: representativity, cancer incidence and mortality in participants and non-participants . Eur J Cancer Prev . 2001 ; 10 ( 6 ): 489 - 99 .
21. Berglund G , Nilsson P , Eriksson KF , Nilsson JA , Hedblad B , Kristenson H , et al. Long-term outcome of the Malmo preventive project: mortality and cardiovascular morbidity . J Intern Med . 2000 ; 247 ( 1 ): 19 - 29 .
22. Cederholm J , Eeg-Olofsson K , Eliasson B , Zethelius B , Nilsson PM , Gudbjornsdottir S , et al. Risk prediction of cardiovascular disease in type 2 diabetes: a risk equation from the Swedish National Diabetes Register . Diabetes Care . 2008 ; 31 ( 10 ): 2038 - 43 .
23. Lindholm E , Agardh E , Tuomi T , Groop L , Agardh CD . Classifying diabetes according to the new WHO clinical stages . Eur J Epidemiol . 2001 ; 17 ( 11 ): 983 - 9 .
24. Hanas R , John G , International HBAcCC. 2010 consensus statement on the worldwide standardization of the hemoglobin A1C measurement . Diabetes Care . 2010 ; 33 ( 8 ): 1903 - 4 .
25. Hoelzel W , Weykamp C , Jeppsson JO , Miedema K , Barr JR , Goodall I , et al. IFCC reference system for measurement of hemoglobin A1c in human blood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study . Clin Chem . 2004 ; 50 ( 1 ): 166 - 74 .
26. Elmstahl S , Riboli E , Lindgarde F , Gullberg B , Saracci R. The Malmo Food Study: the relative validity of a modified diet history method and an extensive food frequency questionnaire for measuring food intake . Eur J Clin Nutr . 1996 ; 50 ( 3 ): 143 - 51 .
27. Riboli E , Elmstahl S , Saracci R , Gullberg B , Lindgarde F. The Malmo Food Study: validity of two dietary assessment methods for measuring nutrient intake . Int J Epidemiol . 1997 ; 26 ( Suppl 1 ): S161 - 73 .
28. Elmstahl S , Gullberg B , Riboli E , Saracci R , Lindgarde F. The Malmo Food Study: the reproducibility of a novel diet history method and an extensive food frequency questionnaire . Eur J Clin Nutr . 1996 ; 50 ( 3 ): 134 - 42 .
29. Willett WC , Howe GR , Kushi LH . Adjustment for total energy intake in epidemiologic studies . Am J Clin Nutr . 1997 ; 65 ( 4 Suppl) : 1220S - 8S . discussion 9S-31S
30. Mattisson I , Wirfalt E , Aronsson CA , Wallstrom P , Sonestedt E , Gullberg B , et al. Misreporting of energy: prevalence, characteristics of misreporters and influence on observed risk estimates in the Malmo Diet and Cancer Cohort . Br J Nutr . 2005 ; 94 ( 5 ): 832 - 42 .
31. Sonestedt E , Wirfalt E , Gullberg B , Berglund G . Past food habit change is related to obesity, lifestyle and socio-economic factors in the Malmo Diet and Cancer Cohort . Public Health Nutr . 2005 ; 8 ( 7 ): 876 - 85 .
32. Wirfalt E , Mattisson I , Johansson U , Gullberg B , Wallstrom P , Berglund G. A methodological report from the Malmo Diet and Cancer study: development and evaluation of altered routines in dietary data processing . Nutr J . 2002 ; 1 : 3 .
33. Manjer J , Elmstahl S , Janzon L , Berglund G . Invitation to a population-based cohort study: differences between subjects recruited using various strategies . Scand J Public Health . 2002 ; 30 ( 2 ): 103 - 12 .
34. Richardson MT , Leon AS , Jacobs DR Jr, Ainsworth BE , Serfass R . Comprehensive evaluation of the Minnesota Leisure Time Physical Activity Questionnaire . J Clin Epidemiol . 1994 ; 47 ( 3 ): 271 - 81 .
35. Li C , Aronsson CA , Hedblad B , Gullberg B , Wirfalt E , Berglund G . Ability of physical activity measurements to assess health-related risks . Eur J Clin Nutr . 2009 ; 63 ( 12 ): 1448 - 51 .
36. Kanoni S , Nettleton JA , Hivert MF , Ye Z , van Rooij FJ , Shungin D , et al. Total zinc intake may modify the glucose-raising effect of a zinc transporter (SLC30A8) variant: a 14-cohort meta-analysis . Diabetes . 2011 ; 60 ( 9 ): 2407 - 16 .
37. Maruthur NM , Clark JM , Fu M , Linda Kao WH , Shuldiner AR . Effect of zinc supplementation on insulin secretion: interaction between zinc and SLC30A8 genotype in Old Order Amish . Diabetologia. 2015 ; 58 ( 2 ): 295 - 303 .
38. Timpson NJ , Lindgren CM , Weedon MN , Randall J , Ouwehand WH , Strachan DP , et al. Adiposity-related heterogeneity in patterns of type 2 diabetes susceptibility observed in genome-wide association data . Diabetes . 2009 ; 58 ( 2 ): 505 - 10 .
39. Cauchi S , Nead KT , Choquet H , Horber F , Potoczna N , Balkau B , et al. The genetic susceptibility to type 2 diabetes may be modulated by obesity status: implications for association studies . BMC Med Genet . 2008 ; 9 : 45 .
40. Moore WT , Bowser SM , Fausnacht DW , Staley LL , Suh KS , Liu D. Beta cell function and the nutritional state: dietary factors that influence insulin secretion . Curr Diab Rep . 2015 ; 15 ( 10 ): 76 .
41. Flannick J , Thorleifsson G , Beer NL , Jacobs SB , Grarup N , Burtt NP , et al. Loss-of-function mutations in SLC30A8 protect against type 2 diabetes . Nat Genet . 2014 ; 46 ( 4 ): 357 - 63 .
42. Fukada T , Kambe T. Molecular and genetic features of zinc transporters in physiology and pathogenesis . Metallomics . 2011 ; 3 ( 7 ): 662 - 74 .