Diabetologia (2022) 65:263–274

https://doi.org/10.1007/s00125-021-05618-w

REVIEW

Environmental risk factors of type 2 diabetes—

an exposome approach
Joline W. J. Beulens 1,2 & Maria G. M. Pinho 1 & Taymara C. Abreu 1 & Nicole R. den Braver 1 & Thao M. Lam 1 &
Anke Huss 3 & Jelle Vlaanderen 3 & Tabea Sonnenschein 2,4 & Noreen Z. Siddiqui 1 & Zhendong Yuan 3 &
Jules Kerckhoffs 3 & Alexandra Zhernakova 5 & Milla F. Brandao Gois 5 & Roel C. H. Vermeulen 2,3

Received: 16 June 2021 / Accepted: 7 October 2021 / Published online: 18 November 2021
# The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2021

Abstract
Type 2 diabetes is one of the major chronic diseases accounting for a substantial proportion of disease burden in Western
countries. The majority of the burden of type 2 diabetes is attributed to environmental risks and modifiable risk factors such
as lifestyle. The environment we live in, and changes to it, can thus contribute substantially to the prevention of type 2 diabetes at
a population level. The ‘exposome’ represents the (measurable) totality of environmental, i.e. nongenetic, drivers of health and
disease. The external exposome comprises aspects of the built environment, the social environment, the physico-chemical
environment and the lifestyle/food environment. The internal exposome comprises measurements at the epigenetic, transcript,
proteome, microbiome or metabolome level to study either the exposures directly, the imprints these exposures leave in the
biological system, the potential of the body to combat environmental insults and/or the biology itself. In this review, we describe
the evidence for environmental risk factors of type 2 diabetes, focusing on both the general external exposome and imprints of
this on the internal exposome. Studies provided established associations of air pollution, residential noise and area-level
socioeconomic deprivation with an increased risk of type 2 diabetes, while neighbourhood walkability and green space are
consistently associated with a reduced risk of type 2 diabetes. There is little or inconsistent evidence on the contribution of the
food environment, other aspects of the social environment and outdoor temperature. These environmental factors are thought to
affect type 2 diabetes risk mainly through mechanisms incorporating lifestyle factors such as physical activity or diet, the
microbiome, inflammation or chronic stress. To further assess causality of these associations, future studies should focus on
investigating the longitudinal effects of our environment (and changes to it) in relation to type 2 diabetes risk and whether these
associations are explained by these proposed mechanisms.

Keywords Built environment . Exposome . Food environment . Lifestyle . Metabolomics . Microbiome . Physico-chemical
environment . Review . Social environment . Type 2 diabetes

Abbreviation
HRMS High-resolution mass spectrometry

* Joline W. J. Beulens 3
Institute for Risk Assessment Sciences, Utrecht University,
j.beulens@amsterdamumc.nl Utrecht, the Netherlands
4
Department of Human Geography and Spatial Planning, Utrecht
1
Department of Epidemiology & Data Science, Amsterdam Public University, Utrecht, the Netherlands
Health, Amsterdam Cardiovascular Sciences, Amsterdam UMC, 5
Department of Genetics, University Medical Center Groningen,
location VUmc, Amsterdam, the Netherlands Groningen, the Netherlands
2
Julius Center for Health Sciences and Primary Care, University
Medical Center Utrecht, Utrecht, the Netherlands
264 Diabetologia (2022) 65:263–274

Introduction
Glossary
Type 2 diabetes is a major chronic disease burden in Western
countries, which is estimated to affect 642 million people Exposome The totality of environmental drivers of
worldwide by 2040 [1]. Its increasing prevalence can be disease that an individual experiences over the
explained by non-modifiable factors, such as the ageing of course of their life
the population, and modifiable factors like overweight/
General external exposome The external contri-
obesity and unhealthy lifestyle habits [1, 2]. Large prevention butors to the exposome, comprised of:
trials show that the risk of type 2 diabetes is reduced by
• The built environment The physical space we
approximately 50% by lifestyle modification in high-risk
live and work in
populations [3]. However, translation of such interventions
into real-life and less-controlled settings is challenging [4], • The social environment The social relationships
and social context in which groups of people live
with the overall risk reduction falling to just 15% after 6 years
and interact e.g. socioeconomic position or social
[5]. One reason for this might be that many interventions do interactions
not sufficiently account for the context or living environment
• The physico-chemical environment The chem-
of the individual and rely primarily on the cognitive capacities
ical or physical agents present in our local area
and intrinsic motivation of those targeted [6]. Therefore,
interventions may be more impactful if an individual’s • The lifestyle/food environment The accessibil-
ity, availability, affordability and promotion of
environment is accounted for.
foods and food retailers in our neighbourhood
The human genome project revolutionised our
understanding of the genetic origins of disease. Genome- Specific external exposome Exposures at an
wide association studies estimate that genetic variation individual level, such as health behaviours (e.g. diet,
solely explains 15–20% of the burden of type 2 diabetes exercise, smoking or alcohol intake), income/financial
[7], although other studies show an estimated genetic status and exposure to pollutants
contribution of around 45%, albeit with a wide range [8].
Internal exposome Internal biological processes,
Nevertheless, a large part of the burden of type 2 diabetes is such as metabolism and the microbiome, which may
attributed to modifiable and/or environmental risk factors be impacted by external exposures (exposure
and their interactions with our genetic make-up. Indeed, imprints) or allow the body to combat environmental
population attributable fractions range from 10% for insults
smoking to 48% for obesity [9]. However, many of the
environmental drivers of type 2 diabetes remain unknown,
hampering the development of effective prevention
programmes. One reason for the lack of understanding of area) and the lifestyle/food environment (the accessibility,
these environmental drivers is the lack of good measures availability, affordability and promotion of food and food
of our ‘environment’, both in coverage (the number of retailers in our neighbourhood) (Fig. 1) [11]. The specific
environmental factors that can be quantified) and resolution external exposome refers to individual exposures such as health
(the quality with which the proxies capture true exposure). behaviours. The internal exposome includes measurements at the
To address the imbalance of our abilities to measure epigenetic, transcriptome, proteome, microbiome or metabolome
environmental factors compared with genetic factors, the term level to study either the exposures directly (e.g. non-targeted
‘exposome’ was coined in 2005 [10]. The ‘exposome’ chemical screening [12]), the exposure imprints (e.g. biological
represents the (measurable) totality of environmental, i.e. imprints of smoking in the epigenome [13]), the potential of the
nongenetic, drivers of disease [10]. Analyses of biological body to combat environmental insults (i.e. allostatic load [14])
perturbations at different molecular levels, together with and/or the biology itself (Fig. 1) [15].
environmental measurements, should provide insights on the In this review, we describe the evidence for environmental
internal and external exposome contributors [11] (see the text risk factors of type 2 diabetes focusing on both the general
box for a glossary of terms). external exposome and imprints of this on the internal
The external contributors to the exposome (i.e. the general exposome. Because aspects of the specific external exposome
external exposome) comprises aspects of the built environment at the individual level, such as dietary patterns, physical
(the physical space where we live and work), the social activity and lifestyle modification programmes, have been
environment (the social relationships and social context in which reviewed elsewhere [3, 16], these factors will not be included.
groups of people live and interact, e.g. our socioeconomic We provide a relative grading of the evidence indicated as
position or social interactions), the physico-chemical stronger vs weaker evidence for a specific relationship. A
environment (the chemical or physical agents present in our local stronger grade indicates an established association and is only
Diabetologia (2022) 65:263–274 265

Fig. 1 An overview of the three xternal exposo

different domains of the
eneral e me
exposome, adapted from Wild G
(2012) [112]. The general
external exposome encompasses
wider external influences of the
living environment, whereas the
specific external exposome
reflects individual exposures. The Urban/rural environment
internal exposome reflects how an
individual’s biological processes
and metabolism are impacted by Social capital
these external exposures. This Climate
figure is available as part of a
downloadable slideset Internal exposome
Biological measures
Nature/green spaces reflecting internal biological Traffic
processes and metabolism
and the biological impact
of exposures
Chemicals/pollution/
Tobacco radiation

Nutrition
Physical
activity

Alcohol Medicine
status
Spe me
c ifi c e o
x t er n a l e x p o s

provided when a systematic review or meta-analysis reported Food environment research most frequently defines the
consistent results in at least three studies. A weaker grade food environment as geographical availability and
indicates a suspected association and is provided when only accessibility to food retailers in the home neighbourhood,
incidental studies reported on the specific relationship or when mainly operationalised as measures of the density of food
a systematic review or meta-analysis did not yield consistent retailers or distance to a specific food outlet, derived from
evidence. We conclude by providing methodological underlying spatial data [19–21]. Exposure to unhealthy food
considerations and future perspectives for exposome research retailers such as fast-food outlets has been associated with less
in the field of type 2 diabetes. healthy diets, obesity, increased insulin resistance, increased
triacylglycerol concentration and type 2 diabetes, mainly in
studies performed in the USA [22–24]. Despite these findings
Environmental risk factors of type 2 diabetes of a link between the food environment and type 2 diabetes,
systematic reviews including over 15 studies report
The food environment The food environment encompasses inconsistent findings or null results (Fig. 2) [20, 25, 26].
the accessibility, availability, affordability and promotion of Subjective measures of exposure to food retailers, such as
foods and food retailers [17]. The investigation of the food perceived availability or use of food retailers, are more
environment in relation to diet quality, obesity and chronic consistently associated with type 2 diabetes risk than objective
disease risk gained interest due to evidence of ‘food deserts’ measures [27, 28]. Finally, interventions that change the
in North America. Living in a ‘food desert’ with virtually no exposure to food retailers are scarce, but a natural experiment
geographical access to food retailers has been associated with in residential relocation after the 2011 earthquake and tsunami
lower diet quality or disease outcomes [18]. Similarly, ‘food in Japan suggested that shortened distances to food outlets
swamps’ represent environments where unhealthy food increased the risk of obesity and other cardiometabolic risk
options outweigh healthy options. factors [29, 30].
266 Diabetologia (2022) 65:263–274

The main reasons explaining these inconsistencies are: the distribution of food outlets over time might also play a role, as
over-simplistic definition of ‘exposure’, ignoring other several studies showed substantial changes in the food
environmental attributes like the social environment; the focus environment, mostly with increases of unhealthier food outlets,
on the residential neighbourhood only, not accounting for particularly in neighbourhoods with low socioeconomic position
exposure to the food environment at work or in transit; and the [33–36]. Whether such changes also affect the observed
lack of insight into individuals’ behavioural interactions with the associations of the food environment with the risk of type 2
food environment [19–21, 26, 31, 32]. Changes in the spatial diabetes has been scarcely investigated. A study from Mexico

General external Specific external Internal

exposome exposome exposome

PM2.5 Altered endothelial

Air and NO2
High chemical function,
urbanisation pollutants inflammation
and insulin resistance
Increased
Stronger evidence

Chronic stress,
T2D risk
Area-level Sedentary
socioeconomic behaviours inflammatory
deprivation responses

Green spaces Lower insulin

and higher Increased
resistance,
Decreased
physical activity
walkability lower BMI T2D risk

Increased (residential) Suppressed insulin

High road traffic noise secretion
urbanisation and peripheral
insulin sensitivity

Obesogenic Poor dietary Higher BMI,

food behaviours altered gut
microbiota
Weaker evidence

environment
Increased
Worse glucose
T2D risk
High metabolism,
Heat islands
urbanisation lower brown
adipose tissue

Light Increased
High exposure glucose
urbanisation at night levels

Fig. 2 An overview of the evidence of how elements of the food systematic reviews or meta-analyses of at least three studies. A weaker
environment, built environment, physico-chemical environment and grade indicates a suspected association, where only incidental studies
social environment are related to risk of type 2 diabetes, highlighting have reported the relationship, or where a systematic review or meta-
the potential pathways in which the three aspects of the exposome analysis has not yielded consistent evidence. PM2.5, fine particulate
(general external, specific external and internal) interact. Evidence is matter 2.5 μm or less in diameter; T2D, type 2 diabetes. This figure is
indicated as stronger or weaker for each pathway. A stronger grade available as part of a downloadable slideset
indicates an established association, based on consistent results in
Diabetologia (2022) 65:263–274 267

that prospectively analysed changes in the food environment in particulate matter smaller than 10 μm (PM10) and 1.05–1.08 for
relation to diabetes found that individuals living in nitrogen dioxide (NO2) (Fig. 2) [20, 45]. Studies also suggested
neighbourhoods that experienced a decrease in the density of women may be more susceptible to exposure to pollution
fruit and vegetable stores, and an increase in the density of because it was posited that they may spend more time in and
convenience stores, had higher odds of diabetes compared with around the home than men [20]. Data on other pollutants are very
individuals living in neighbourhoods where these stores did not limited, although a recent study showed sulphur dioxide as a risk
change [37]. factor [46]. There is inconsistency as to what pollutant is most
correlated to type 2 diabetes, mainly because of the limited
The built environment The built environment is defined as number and mixed results of multi-pollutant models [20].
man-made characteristics of the physical environment in Although only a few studies investigated the effect of light
which people live, work and recreate, including buildings, exposure on metabolic diseases [47], a relatively large study in
streets, open spaces and infrastructure [38, 39]. The built Japanese care settings indicated that exposure to light at night
environment is hypothesised to be associated with type 2 could be associated with type 2 diabetes [48]. This association
diabetes incidence primarily through physical activity-related could potentially be explained by the consequent effects on
pathways [40]. Indeed, meta-analyses consistently showed an lifestyle, particularly sleep disruption, which could result in
established association of living in neighbourhoods with high elevated glucose levels [49].
walkability and green space with a 10–20% lower risk of type Two meta-analyses showed an established association
2 diabetes [20, 25, 26] (Fig. 2), although mainly in studies between higher residential, but not occupational, noise
performed in North America and Australia, while evidence exposure and increased risk of type 2 diabetes (Fig. 2) [26,
for European countries is limited. Walkability of a 50]. Noise can act as an environmental stressor that leads to
neighbourhood is characterised by population density, land- insulin secretion and peripheral insulin sensitivity [50], but
use mix (i.e. heterogeneity of land uses in an area such as other lifestyle factors such as sleep may also be involved
residential, industrial and natural) and connectivity (i.e. [50]. Finally, the associations of noise and light with type 2
intersections). Other elements of the built environment in diabetes could be confounded by other factors associated with
relation to risk of type 2 diabetes have not been investigated urbanisation, such as air pollution.
intensively. Six studies investigated availability of sports Higher body temperature could negatively affect
facilities in relation to risk of type 2 diabetes, but reported glucose metabolism by decreasing the brown adipose
inconsistent results ranging from no association to a reduced tissue mass and activity [51]. Experimental studies report
risk with a higher availability [26]. A systematic review high efficiency of cold exposure as a potential therapy for
showed that interventions to increase physical activity by type 2 diabetes [52, 53]. Nevertheless, with the exception
changes in the built environment generally improved levels of one study reporting higher diabetes incidence with
of physical activity [41], but effects on type 2 diabetes have increasing annual outdoor temperature [54], evidence is
not been evaluated. Furthermore, a large Finnish cohort study limited for an association between ambient temperature
of over 100,000 individuals showed that changes in residential and type 2 diabetes.
greenness were associated with a 12% reduced risk of type 2 Finally, several meta-analyses established associations of
diabetes [42]. Finally, a meta-analysis showed an established chemical pollutants, such as persistent organic pollutants [55],
association that urban dwellers, particularly in low- and pesticides [56] and heavy metals [57], with an increased risk
middle-income countries, have a 40% increased risk of type of type 2 diabetes. These pollutants may originate from the
2 diabetes [26], but the underlying drivers of this association complex chemical environment including occupational
are not entirely clear. Apart from mainly lifestyle-related hazards, air or water pollution, and food contaminants.
pathways, the built environment is characterised by Many results suggest a stronger association between chemical
interrelated factors [26, 43] that not only influence human pollutants and type 2 diabetes in women and individuals with
behaviour, but can also directly affect disease risk (e.g. air overweight or obesity [57].
pollution).
Social environment The social environment is generally
Physico-chemical environment Air pollution has been understood as the social relationships and social context in
documented to change endothelial function, trigger inflammation which groups of people live and interact [58]. Examples of
and insulin resistance, alter the gut microbiota and be associated social environment components include area-level
with an elevated risk of hypertension [20, 44]. Recent meta- deprivation, social capital, ethnic segregation and perceived
analyses showed an established association of an increased risk safety. While individual-level social factors (e.g. education
of type 2 diabetes with increased exposure to air pollution, with and income) are consistently associated with type 2 diabetes
odds ratios (per 10 μg/m3 increase) ranging from 1.08–1.10 for risk [59, 60], environmental social factors have recently
particulate matter smaller than 2.5 μm (PM2.5), 1.10–1.12 for received increased attention.
268 Diabetologia (2022) 65:263–274

Systematic reviews consistently show that area-level Nevertheless, an analysis of over 40 observational studies has
socioeconomic deprivation is associated with an up to twofold identified specific members of the gut microbiome that appear
increased incidence of type 2 diabetes (Fig. 2) [25, 61, 62]. In to be consistently associated with type 2 diabetes [81]. Five
line with these studies, and suggesting causal relationship, genera showed a recurrent protective role in relation to type 2
natural experiments and studies of residential relocation show diabetes (Bacteroides, Bifidobacterium, Akkermansia, Roseburia
that moving to low deprivation neighbourhoods reduces and Faecalibacterium). These genera are associated with several
HbA1c and risk of type 2 diabetes [42, 63, 64]. Despite metabolic mechanisms that affect host physiology, such as
growing evidence for the relationship between social capital reduction of endotoxemia, increase of energy yield, drug
and health [65], the few studies investigating social capital and metabolism, decrease of tissue inflammation and production of
type 2 diabetes incidence present mixed findings [61, 65–68]. bacterial metabolites [81, 82, 84–86]. Gou et al recently applied a
Although the importance of good social networks in machine learning framework to correlate gut microbiome
decreasing the risk of type 2 diabetes has been shown in large features to type 2 diabetes in large scale cohorts [87]. They
longitudinal studies, experimental studies are lacking [69]. demonstrated that a microbial risk score of 14 microbial features
Less is known about the impact of ethnic segregation on type (MRS) yielded a superior disease prediction accuracy than other
2 diabetes incidence. Nonetheless, a review by Kershaw and environmental aspects (i.e. lifestyle, diet and host genetics). The
Pender (2016) concluded that ethnic segregation can influence combination of all factors, however, showed the highest
the severity of type 2 diabetes but not its development [70]. predictive accuracy [87]. Nonetheless, this study highlights the
Regarding discrimination, studies mostly investigated the role of the microbiome in type 2 diabetes and the complex
impact of individually experienced, but not neighbourhood- interaction of the gut microbiome with environmental factors,
level, discrimination on type 2 diabetes incidence [71]. and their link to the onset and progression of type 2 diabetes
Furthermore, only a few studies have investigated the impact requires further research.
of perceived crime within the neighbourhood on type 2
diabetes incidence, with some finding a significant association Metabolome and exposome scans T e c h n o l o g i c a l
[72], and others indicating perceived crime as a weak developments, in which untargeted liquid- (LC-) and gas
moderator of amenities density and type 2 diabetes incidence (GC-) chromatography are combined with high-resolution
[73]. mass spectrometry (HRMS), have made it possible to
Even though the evidence for an association between comprehensively, and in a high-throughput fashion, measure
certain components of the social environment and risk of type the patterns of thousands of metabolites that are present in
2 diabetes may be sparse or inconclusive, stronger links have biological fluids, known as the metabolome [88, 89]. The
been observed between social environment factors and metabolome provides a picture of the functional status of the
lifestyle behaviours and obesity [70, 74–78], suggesting a biological system and as such can provide insights into the
plausible link with type 2 diabetes. Moreover, social pathophysiology of type 2 diabetes and enable identification
environment factors are also hypothesised to influence type of type 2 diabetes biomarkers [88, 89]. Several endogenous
2 diabetes incidence through chronic stress and inflammatory metabolites were identified as early biomarkers of type 2
responses [79, 80]. diabetes, including branched-chain amino acids, aromatic
amino acids, 2-aminoadipic acid, sphingomyelin, glycine,
acyl-alkyl-phosphatidylcholines, lysophosphatidylcholine,
Internal exposome hexose, β-hydroxybutyrate, linoleoylglycerophosphocholine,
and glyoxylate [89–94].
Although our internal exposome consists of proteins, lipids, In light of the possibilities for future (public health)
metabolites and so on, the role of the metabolome and interventions, it is of interest to assess which environmental
microbiome appear to be of particular importance in the factors impact type 2 diabetes directly or through perturbations
aetiology of type 2 diabetes. of the metabolome. In the USA, Patel et al measured 266
environmental factors in urine or blood as part of the National
The microbiome The gut microbiota is known for its ability to Health and Nutrition Examination Survey [95]. The pesticide
modulate inflammation, metabolise xenobiotics, maintain derivative heptachlor epoxide, vitamin γ-tocopherol, and specific
intestinal integrity and its interactions with dietary components. polychlorinated biphenyls were identified as risk factors for type
Changes in the abundance and diversity of the gut microbiota 2 diabetes. β-Carotenes (among others) were identified as
have been linked to the progression of many metabolic diseases, protective factors for type 2 diabetes.
including type 2 diabetes [81–83]. The specific associations Recent optimisation of LC-HRMS and GC-HRMS platforms
between the gut microbiome and type 2 diabetes in humans have combined with innovative data extraction approaches now
conveyed conflicting results, which can be partly explained by enable the detection of an even wider range of exposure-related
large methodological differences in microbiome studies [81–83]. chemicals present at very low concentrations in biological fluids
Diabetologia (2022) 65:263–274 269

(e.g. blood and urine) [15]. Application of these technologies to outlets in certain regions, may offer additional evidence by
human populations allows detection and characterisation of a creating a sudden and larger change in the environment.
large range of exogenously derived small chemicals, including Furthermore, most studies have focused on the residential living
pharmaceuticals, pesticides, preservatives, dietary compounds environment, while activity space (routes, destinations and work
and microbial metabolites, in small quantities of biological environments) is likely to be relevant, as well as an individual’s
materials [96]. Collectively, these measurements provide a interaction with their environments [109]. Finally, although
snapshot of the internal exposome, which can be used to urban residence is associated with an increased risk of type 2
elucidate some of the underlying biological mechanisms of diabetes compared with rural residence, particularly in low- and
known or suspected risk factors (such as heavy metals middle-income countries [26], for other environmental risk
[97–100], other trace elements [101], persistent organic factors most evidence comes from high-income countries [20,
pollutants [102, 103], drug use [104] and air pollution [105]) 25, 26]. With the exception of a few studies addressing effect
on the development of type 2 diabetes. modification by sex, ethnicity, income or other characteristics
[28, 110], little is known about differential effects of
environmental factors across subpopulations based on sex or
Methodological considerations ethnicity, which should be further investigated.
A more thorough understanding of the underlying
Our living environment is mostly an indirect determinant of type biological pathways linking the environment to risk of type
2 diabetes and changes in the living environment are often 2 diabetes may also help in assessing causality. In addition to
difficult to investigate in controlled studies. Therefore, the providing deeper biological insights into previously identified
majority of evidence on environmental risk factors of type 2 risk factors, assessment of the internal exposome will
diabetes stems from observational studies with relatively limited contribute to the identification of currently unknown risk
evidence from natural experiments. Causal inference based on factors of type 2 diabetes, and will provide insight into how
observational studies is challenging, since they can be susceptible these exogenous chemicals collectively interact with type 2
to reverse causation and confounding, as residence preference diabetes risk profiles [11]. Such insights require the
and selection is not a random process [106]. Although most application of advanced methods for annotation [96], and
studies adjust for relevant confounding factors like statistical and biological interpretation of the generated data
socioeconomic position, most studies cannot control for [11]. Large studies are needed to make solid inferences using
processes underlying choice of residence. Moreover, complex and high-dimensional data.
environmental characteristics often correlate with a certain Even though high-quality methods are available today,
location. For example, urbanisation is often associated with progress can still be made in terms of the quality of the
neighbourhood socioeconomic position, and certain food outlets assessment of the living environment. This pertains to both
also cluster in highly urbanised neighbourhoods with lower the external and the internal exposome. The type of tools and
socioeconomic position. Therefore, conclusions on causality of methods that are needed varies from domain to domain but
the observed associations can only be drawn for environmental should typically incorporate the ability to assess a wide range
factors where natural experiments or well-controlled of factors with high sensitivity. A recent review provided a
observational studies of residential relocation confirm critical overview of the tools and methods that are currently
associations found in observational studies. Based on the current available for exposome studies and indicated where progress
evidence, this mainly holds true for neighbourhood deprivation, can still be made [15].
green space and walkability. For other environmental factors,
longitudinal studies, and particularly natural experiments, are
needed to confirm observed associations. Such studies should Future perspectives
account for changes in the living environment in relation to
incidence of type 2 diabetes either by selecting a cohort of people Future studies should investigate longitudinal changes in our
moving to a different location to study the changes in living environment in relation to risk of type 2 diabetes, including
environment or by selecting a cohort of people residing in the pathway analyses of health behaviours. Effects of our
same location for a longer time period to study environmental environment on the microbiome or endogenous metabolites
changes in this neighbourhood in relation to type 2 diabetes. should also be incorporated to investigate underlying
However, such studies thus far are scarce. Furthermore, for pathways in the association of our environment with type 2
certain environmental factors, with the exception of the food diabetes through effects on stress responses or inflammation.
environment [35], changes over time will be small, and By analysing not only the gut microbiome in relation to the
development of type 2 diabetes is slow, warranting long disease, but also considering the influence of other
follow-up durations [107, 108]. Natural experiments, for instance environmental exposures, such as diet, pollution and
implementing a new car-free cycling zone or limiting fast-food medication, as well as host metabolism, we expect to be able
270 Diabetologia (2022) 65:263–274

to better understand the role of the gut microbiome in type 2 inflammation or chronic stress. Future studies should focus
diabetes. For example, for light at night, more research is on investigating the longitudinal association of changes in
necessary to understand the underlying mechanism driving our environment in relation to risk of type 2 diabetes and
the association with increased risk of type 2 diabetes, as it is whether these associations are explained by these proposed
suggested that hormone levels, circadian rhythms or sleep mechanisms. When robust evidence for the association of
quality, may play a role in these associations [48]. environmental factors with risk of type 2 diabetes is available,
For the food environment, exposure should be natural experiments or health impact modelling can help to
operationalised more accurately. More insight is needed into evaluate the impact of environmental interventions on disease
mobility patterns combined with behavioural insights on use burden. This will contribute to urban planning and help shape
of food retailers and food delivery services, and perceptions of a healthier living environment.
the food environment. This can be done by detailed
assessment of these environmental exposures over time in Supplementary Information The online version contains a slideset of the
smaller longitudinal observational studies that account for figures for download, which is available to authorised users at https://doi.
exposures other than the residential environment and activity org/10.1007/s00125-021-05618-w.
space of individuals. These studies can contribute to a more
accurate exposure assessment in cohort studies or registries to Funding Work in the authors’ laboratories is supported by EXPOSOME-
NL and EXPANSE. EXPOSOME-NL is funded through the Gravitation
investigate the association with incidence of type 2 diabetes. program of the Dutch Ministry of Education, Culture, and Science and the
The exposure assessment in cohorts could be improved using Netherlands Organization for Scientific Research (NWO grant number
regression calibration techniques when the necessary data are 024.004.017). EXPANSE has received funding from the European
available. Otherwise, assessment should be improved by Union’s Horizon 2020 research and innovation programme under grant
agreement No 874627.
incorporating relevant exposure measures in follow-up
questionnaires. For the social environment, future research Authors’ relationships and activities The authors declare that there are
should account for methodological challenges in investigating no relationships or activities that might bias, or be perceived to bias, their
the link between social environment and type 2 diabetes risk, work. JWJB declares to be a member of the editorial board of
such as inconsistent conceptualisations of the social Diabetologia.
environment, measurements unable to tease apart different
Authors’ contribution All authors were responsible for drafting the article
social phenomena operating at multiple levels of influence, and revising it critically for important intellectual content. All authors
long time lag between exposure and impact on type 2 diabetes approved the version to be published.
development, and risk accumulation over the life course [58,
72, 111].
For green space, walkability, air pollution and neighbourhood
References
socioeconomic position, consistent and robust associations with
type 2 diabetes have been documented. For these aspects, 1. Standl E, Khunti K, Hansen TB, Schnell O (2019) The global
structural interventions should be evaluated for their effect on epidemics of diabetes in the 21st century: current situation and
risk of type 2 diabetes and other chronic diseases using health perspectives. Eur J Prev Cardiol 26(2_suppl):7–14. https://doi.
impact modelling or natural experiments. These studies will org/10.1177/2047487319881021
2. Dal Canto E, Ceriello A, Ryden L et al (2019) Diabetes as a
contribute to urban planning and help shape a healthier living cardiovascular risk factor: an overview of global trends of macro
environment. and micro vascular complications. Eur J Prev Cardiol
26(2_suppl):25–32. https://doi.org/10.1177/2047487319878371
3. Uusitupa M, Khan TA, Viguiliouk E et al (2019) Prevention of
Conclusion type 2 diabetes by lifestyle changes: a systematic review and
Meta-analysis. Nutrients 11(11). https://doi.org/10.3390/
nu11112611
In conclusion, our environment accounts for a substantial 4. Duijzer G, Bukman AJ, Meints-Groenveld A et al (2019) Cost-
proportion of disease burden due to type 2 diabetes. Studies effectiveness of the SLIMMER diabetes prevention intervention
have provided consistent evidence that air pollution, in Dutch primary health care: economic evaluation from a
randomised controlled trial. BMC Health Serv Res 19(1):824.
residential noise, neighbourhood walkability, green space https://doi.org/10.1186/s12913-019-4529-8
and area-level socioeconomic deprivation are associated with 5. Yoon U, Kwok LL, Magkidis A (2013) Efficacy of lifestyle
risk of type 2 diabetes. Current evidence for the contribution interventions in reducing diabetes incidence in patients with
of the food environment, other aspects of the social impaired glucose tolerance: a systematic review of randomized
controlled trials. Metabolism 62(2):303–314. https://doi.org/10.
environment and temperature is low or inconsistent. Our
1016/j.metabol.2012.07.009
environment is thought to affect the risk of type 2 diabetes 6. Backholer K, Beauchamp A, Ball K et al (2014) A framework for
mainly through mechanisms incorporating lifestyle factors evaluating the impact of obesity prevention strategies on
such as physical activity or diet, the microbiome,
Diabetologia (2022) 65:263–274 271

socioeconomic inequalities in weight. Am J Public Health aged adults: the CARDIA study. Arch Intern Med 171(13):
104(10):e43–e50. https://doi.org/10.2105/AJPH.2014.302066 1162–1170. https://doi.org/10.1001/archinternmed.2011.283
7. Mahajan A, Taliun D, Thurner M et al (2018) Fine-mapping type 23. Duffey KJ, Gordon-Larsen P, Steffen LM, Jacobs DR Jr, Popkin
2 diabetes loci to single-variant resolution using high-density BM (2009) Regular consumption from fast food establishments
imputation and islet-specific epigenome maps. Nat Genet relative to other restaurants is differentially associated with
50(11):1505–1513. https://doi.org/10.1038/s41588-018-0241-6 metabolic outcomes in young adults. J Nutr 139(11):2113–2118.
8. Fuchsberger C, Flannick J, Teslovich TM et al (2016) The genetic https://doi.org/10.3945/jn.109.109520
architecture of type 2 diabetes. Nature 536(7614):41–47. https:// 24. Pereira MA, Kartashov AI, Ebbeling CB et al (2005) Fast-food
doi.org/10.1038/nature18642 habits, weight gain, and insulin resistance (the CARDIA study):
9. Li Y, Wang DD, Ley SH et al (2017) Time trends of dietary and 15-year prospective analysis. Lancet 365(9453):36–42. https://
lifestyle factors and their potential impact on diabetes burden in doi.org/10.1016/S0140-6736(04)17663-0
China. Diabetes Care 40(12):1685–1694. https://doi.org/10.2337/ 25. Bilal U, Auchincloss AH, Diez-Roux AV (2018) Neighborhood
dc17-0571 environments and diabetes risk and control. Curr Diab Rep 18(9):
10. Wild CP (2005) Complementing the genome with an "exposome": 62. https://doi.org/10.1007/s11892-018-1032-2
the outstanding challenge of environmental exposure 26. den Braver NR, Lakerveld J, Rutters F, Schoonmade LJ, Brug J,
measurement in molecular epidemiology. Cancer Epidemiol Beulens JWJ (2018) Built environmental characteristics and
Biomark Prev 14(8):1847–1850. https://doi.org/10.1158/1055- diabetes: a systematic review and meta-analysis. BMC Med
9965.EPI-05-0456 16(1):12. https://doi.org/10.1186/s12916-017-0997-z
11. Vermeulen R, Schymanski EL, Barabasi AL, Miller GW (2020) 27. Auchincloss AH, Diez Roux AV, Mujahid MS, Shen M, Bertoni
The exposome and health: where chemistry meets biology. AG, Carnethon MR (2009) Neighborhood resources for physical
Science 367(6476):392–396. https://doi.org/10.1126/science. activity and healthy foods and incidence of type 2 diabetes
aay3164 mellitus: the multi-ethnic study of atherosclerosis. Arch Intern
12. Pourchet M, Debrauwer L, Klanova J et al (2020) Suspect and Med 169(18):1698–1704. https://doi.org/10.1001/archinternmed.
non-targeted screening of chemicals of emerging concern for 2009.302
human biomonitoring, environmental health studies and support 28. Christine PJ, Auchincloss AH, Bertoni AG et al (2015)
to risk assessment: from promises to challenges and harmonisation Longitudinal associations between neighborhood physical and
issues. Environ Int 139:105545. https://doi.org/10.1016/j.envint. social environments and incident type 2 diabetes mellitus: the
2020.105545 multi-ethnic study of atherosclerosis (MESA). JAMA Intern
13. Guida F, Sandanger TM, Castagne R et al (2015) Dynamics of Med 175(8):1311–1320. https://doi.org/10.1001/jamainternmed.
smoking-induced genome-wide methylation changes with time 2015.2691
since smoking cessation. Hum Mol Genet 24(8):2349–2359. 29. Hikichi H, Aida J, Kondo K, Tsuboya T, Kawachi I (2019)
https://doi.org/10.1093/hmg/ddu751 Residential relocation and obesity after a natural disaster: a natural
14. Chadeau-Hyam M, Bodinier B, Vermeulen R et al (2020) experiment from the 2011 Japan earthquake and tsunami. Sci Rep
Education, biological ageing, all-cause and cause-specific 9(1):374. https://doi.org/10.1038/s41598-018-36906-y
mortality and morbidity: UK biobank cohort study. 30. Shiba K, Hanazato M, Aida J et al (2020) Cardiometabolic profiles
EClinicalMedicine 29-30:100658. https://doi.org/10.1016/j. and change in neighborhood food and built environment among
eclinm.2020.100658 older adults: a natural experiment. Epidemiology 31(6):758–767.
15. Niedzwiecki MM, Walker DI, Vermeulen R, Chadeau-Hyam M, https://doi.org/10.1097/EDE.0000000000001243
Jones DP, Miller GW (2019) The Exposome: molecules to 31. Frankenfeld CL, Leslie TF, Makara MA (2015) Diabetes, obesity,
populations. Annu Rev Pharmacol Toxicol 59:107–127. https:// and recommended fruit and vegetable consumption in relation to
doi.org/10.1146/annurev-pharmtox-010818-021315 food environment sub-types: a cross-sectional analysis of
16. Bellou V, Belbasis L, Tzoulaki I, Evangelou E (2018) Risk factors behavioral risk factor surveillance system, United States census,
for type 2 diabetes mellitus: an exposure-wide umbrella review of and food establishment data. BMC Public Health 15:491. https://
meta-analyses. PLoS One 13(3):e0194127. https://doi.org/10. doi.org/10.1186/s12889-015-1819-x
1371/journal.pone.0194127 32. Mezuk B, Li X, Cederin K, Rice K, Sundquist J, Sundquist K
17. Glanz K, Sallis JF, Saelens BE, Frank LD (2005) Healthy nutrition (2016) Beyond access: characteristics of the food environment
environments: concepts and measures. Am J Health Promot 19(5): and risk of diabetes. Am J Epidemiol 183(12):1129–1137.
330–333, ii. https://doi.org/10.4278/0890-1171-19.5.330 https://doi.org/10.1093/aje/kwv318
18. Walker RE, Keane CR, Burke JG (2010) Disparities and access to 33. James P, Seward MW, James O'Malley A, Subramanian SV,
healthy food in the United States: a review of food deserts Block JP (2017) Changes in the food environment over time:
literature. Health Place 16(5):876–884. https://doi.org/10.1016/j. examining 40 years of data in the Framingham heart study. Int J
healthplace.2010.04.013 Behav Nutr Phys Act 14(1):84. https://doi.org/10.1186/s12966-
19. Caspi CE, Sorensen G, Subramanian SV, Kawachi I (2012) The 017-0537-4
local food environment and diet: a systematic review. Health Place 34. Maguire ER, Burgoine T, Monsivais P (2015) Area deprivation
18(5):1172–1187. https://doi.org/10.1016/j.healthplace.2012.05. and the food environment over time: a repeated cross-sectional
006 study on takeaway outlet density and supermarket presence in
20. Dendup T, Feng X, Clingan S, Astell-Burt T (2018) Norfolk, UK, 1990-2008. Health Place 33:142–147. https://doi.
Environmental risk factors for developing type 2 diabetes org/10.1016/j.healthplace.2015.02.012
mellitus: a systematic review. Int J Environ Res Public Health 35. Pinho MGM, Mackenbach JD, den Braver NR, Beulens JJW,
15(1). https://doi.org/10.3390/ijerph15010078 Brug J, Lakerveld J (2020) Recent changes in the Dutch
21. Lytle LA, Sokol RL (2017) Measures of the food environment: a foodscape: socioeconomic and urban-rural differences. Int J
systematic review of the field, 2007-2015. Health Place 44:18–34. Behav Nutr Phys Act 17(1):43. https://doi.org/10.1186/s12966-
https://doi.org/10.1016/j.healthplace.2016.12.007 020-00944-5
22. Boone-Heinonen J, Gordon-Larsen P, Kiefe CI, Shikany JM, 36. Richardson AS, Meyer KA, Howard AG et al (2014)
Lewis CE, Popkin BM (2011) Fast food restaurants and food Neighborhood socioeconomic status and food environment: a
stores: longitudinal associations with diet in young to middle- 20-year longitudinal latent class analysis among CARDIA
272 Diabetologia (2022) 65:263–274

participants. Health Place 30:145–153. https://doi.org/10.1016/j. 52. Hanssen MJ, Hoeks J, Brans B et al (2015) Short-term cold
healthplace.2014.08.011 acclimation improves insulin sensitivity in patients with type 2
37. Perez-Ferrer C, Auchincloss AH, Barrientos-Gutierrez T et al diabetes mellitus. Nat Med 21(8):863–865. https://doi.org/10.
(2020) Longitudinal changes in the retail food environment in 1038/nm.3891
Mexico and their association with diabetes. Health Place 66: 53. Ouellet V, Labbe SM, Blondin DP et al (2012) Brown adipose
102461. https://doi.org/10.1016/j.healthplace.2020.102461 tissue oxidative metabolism contributes to energy expenditure
38. Amuda AT, Berkowitz SA (2019) Diabetes and the built during acute cold exposure in humans. J Clin Invest 122(2):545–
environment: evidence and policies. Curr Diab Rep 19(7):35. 552. https://doi.org/10.1172/JCI60433
https://doi.org/10.1007/s11892-019-1162-1 54. Blauw LL, Aziz NA, Tannemaat MR et al (2017) Diabetes
39. Sallis JF, Floyd MF, Rodriguez DA, Saelens BE (2012) Role of incidence and glucose intolerance prevalence increase with higher
built environments in physical activity, obesity, and outdoor temperature. BMJ Open Diabetes Res Care 5(1):e000317.
cardiovascular disease. Circulation 125(5):729–737. https://doi. https://doi.org/10.1136/bmjdrc-2016-000317
org/10.1161/CIRCULATIONAHA.110.969022 55. Wu H, Bertrand KA, Choi AL et al (2013) Persistent organic
40. Gorman S, Larcombe AN, Christian HE (2021) Exposomes and pollutants and type 2 diabetes: a prospective analysis in the nurses'
metabolic health through a physical activity lens: a narrative health study and meta-analysis. Environ Health Perspect 121(2):
review. J Endocrinol 249(1):R25–R41. https://doi.org/10.1530/ 153–161. https://doi.org/10.1289/ehp.1205248
JOE-20-0487 56. Evangelou E, Ntritsos G, Chondrogiorgi M et al (2016) Exposure
41. Smith M, Hosking J, Woodward A et al (2017) Systematic to pesticides and diabetes: a systematic review and meta-analysis.
literature review of built environment effects on physical activity Environ Int 91:60–68. https://doi.org/10.1016/j.envint.2016.02.
and active transport - an update and new findings on health equity. 013
Int J Behav Nutr Phys Act 14(1):158. https://doi.org/10.1186/ 57. Misra BB, Misra A (2020) The chemical exposome of type 2
s12966-017-0613-9 diabetes mellitus: opportunities and challenges in the omics era.
42. Kivimaki M, Batty GD, Pentti J et al (2021) Modifications to Diabetes Metab Syndr 14(1):23–38. https://doi.org/10.1016/j.dsx.
residential neighbourhood characteristics and risk of 79 common 2019.12.001
health conditions: a prospective cohort study. Lancet Public 58. Barnett E, Casper M (2001) A definition of "social environment".
Health 6(6):e396–e407. https://doi.org/10.1016/S2468-2667(21) Am J Public Health 91(3):465–465. https://doi.org/10.2105/
00066-9 AJPH.91.3.465a
59. Kivimäki M, Virtanen M, Kawachi I et al (2015) Long working
43. den Braver NR, Kok JG, Mackenbach JD et al (2020)
hours, socioeconomic status, and the risk of incident type 2
Neighbourhood drivability: environmental and individual
diabetes: a meta-analysis of published and unpublished data from
characteristics associated with car use across Europe. Int J
222120 individuals. Lancet Diabetes Endocrinol 3(1):27–34.
Behav Nutr Phys Act 17(1):8. https://doi.org/10.1186/s12966-
https://doi.org/10.1016/S2213-8587(14)70178-0
019-0906-2
60. Seiglie JA, Marcus ME, Ebert C et al (2020) Diabetes prevalence
44. Fouladi F, Bailey MJ, Patterson WB et al (2020) Air pollution
and its relationship with education, wealth, and BMI in 29 low-
exposure is associated with the gut microbiome as revealed by
and middle-income countries. Diabetes Care 43(4):767–775.
shotgun metagenomic sequencing. Environ Int 138:105604.
https://doi.org/10.2337/dc19-1782
https://doi.org/10.1016/j.envint.2020.105604
61. Hill-Briggs F, Adler NE, Berkowitz SA et al (2020) Social
45. Yang BY, Fan S, Thiering E et al (2020) Ambient air pollution and determinants of health and Diabetes: A Scientific Review.
diabetes: a systematic review and meta-analysis. Environ Res 180: Diabetes Care. https://doi.org/10.2337/dci20-0053
108817. https://doi.org/10.1016/j.envres.2019.108817 62. Müller G, Kluttig A, Greiser KH et al (2013) Regional and
46. Shan A, Zhang Y, Zhang LW et al (2020) Associations between neighborhood disparities in the odds of type 2 diabetes: results
the incidence and mortality rates of type 2 diabetes mellitus and from 5 population-based studies in Germany (DIAB-CORE
long-term exposure to ambient air pollution: a 12-year cohort consortium). Am J Epidemiol 178(2):221–230. https://doi.org/
study in northern China. Environ Res 186:109551. https://doi. 10.1093/aje/kws466
org/10.1016/j.envres.2020.109551 63. Ludwig J, Sanbonmatsu L, Gennetian L et al (2011)
47. Afroz-Hossain A, Dawkins M, Myers AK (2019) Sleep and Neighborhoods, obesity, and diabetes–a randomized social
environmental factors affecting glycemic control in people with experiment. N Engl J Med 365(16):1509–1519. https://doi.org/
type 2 diabetes mellitus. Curr Diab Rep 19(7):40. https://doi.org/ 10.1056/NEJMsa1103216
10.1007/s11892-019-1159-9 64. White JS, Hamad R, Li X et al (2016) Long-term effects of
48. Obayashi K, Saeki K, Iwamoto J, Ikada Y, Kurumatani N (2014) neighbourhood deprivation on diabetes risk: quasi-experimental
Independent associations of exposure to evening light and evidence from a refugee dispersal policy in Sweden. Lancet
nocturnal urinary melatonin excretion with diabetes in the elderly. Diabetes Endocrinol 4(6):517–524. https://doi.org/10.1016/
Chronobiol Int 31(3):394–400. https://doi.org/10.3109/07420528. S2213-8587(16)30009-2
2013.864299 65. Rodgers J, Valuev AV, Hswen Y, Subramanian SV (2019) Social
49. Tan X, Chapman CD, Cedernaes J, Benedict C (2018) Association capital and physical health: an updated review of the literature for
between long sleep duration and increased risk of obesity and type 2007-2018. Soc Sci Med 236:112360. https://doi.org/10.1016/j.
2 diabetes: a review of possible mechanisms. Sleep Med Rev 40: socscimed.2019.112360
127–134. https://doi.org/10.1016/j.smrv.2017.11.001 66. Hu F, Hu B, Chen R et al (2014) A systematic review of social
50. Dzhambov AM (2015) Long-term noise exposure and the risk for capital and chronic non-communicable diseases. Biosci Trends
type 2 diabetes: a meta-analysis. Noise Health 17(74):23–33. 8(6):290–296. https://doi.org/10.5582/bst.2014.01138
https://doi.org/10.4103/1463-1741.149571 67. Riumallo-Herl CJ, Kawachi I, Avendano M (2014) Social capital,
51. Ouellet V, Routhier-Labadie A, Bellemare W et al (2011) Outdoor mental health and biomarkers in Chile: assessing the effects of
temperature, age, sex, body mass index, and diabetic status social capital in a middle-income country. Soc Sci Med 105:47–
determine the prevalence, mass, and glucose-uptake activity of 58. https://doi.org/10.1016/j.socscimed.2013.12.018
18F-FDG-detected BAT in humans. J Clin Endocrinol Metab 68. Holtgrave DR, Crosby R (2006) Is social capital a protective factor
96(1):192–199. https://doi.org/10.1210/jc.2010-0989 against obesity and diabetes? Findings from an exploratory study.
Diabetologia (2022) 65:263–274 273

Ann Epidemiol 16(5):406–408. https://doi.org/10.1016/j. 85. Li R, Andreu-Sanchez S, Kuipers F, Fu J (2021) Gut microbiome
annepidem.2005.04.017 and bile acids in obesity-related diseases. Best Pract Res Clin
69. Schram MT, Assendelft WJJ, van Tilburg TG, Dukers-Muijrers N Endocrinol Metab 101493. https://doi.org/10.1016/j.beem.2021.
(2021) Social networks and type 2 diabetes: a narrative review. 101493
Diabetologia 64(9):1905–1916. https://doi.org/10.1007/s00125- 86. Qin J, Li Y, Cai Z et al (2012) A metagenome-wide association
021-05496-2 study of gut microbiota in type 2 diabetes. Nature 490(7418):55–
70. Kershaw KN, Pender AE (2016) Racial/ethnic residential 60. https://doi.org/10.1038/nature11450
segregation, obesity, and diabetes mellitus. Curr Diab Rep 87. Gou W, Ling CW, He Y et al (2021) Interpretable machine
16(11):108. https://doi.org/10.1007/s11892-016-0800-0 learning framework reveals robust gut microbiome features
71. Whitaker KM, Everson-Rose SA, Pankow JS et al (2017) associated with type 2 diabetes. Diabetes Care 44(2):358–366.
Experiences of discrimination and incident type 2 diabetes https://doi.org/10.2337/dc20-1536
mellitus: the multi-ethnic study of atherosclerosis (MESA). Am 88. Gonzalez-Franquesa A, Burkart AM, Isganaitis E, Patti ME
J Epidemiol 186(4):445–455. https://doi.org/10.1093/aje/kwx047 (2016) What have metabolomics approaches taught us about type
72. Dendup T, Astell-Burt T, Feng XQ (2019) Residential self- 2 diabetes? Curr Diab Rep 16(8):74. https://doi.org/10.1007/
selection, perceived built environment and type 2 diabetes s11892-016-0763-1
incidence: a longitudinal analysis of 36,224 middle to older age 89. Wang TJ, Larson MG, Vasan RS et al (2011) Metabolite profiles
adults. Health Place 58. https://doi.org/10.1016/j.healthplace. and the risk of developing diabetes. Nat Med 17(4):448–453.
2019.102154 https://doi.org/10.1038/nm.2307
73. Dendup T, Feng X, O'Shaughnessy PY, Astell-Burt T (2021) Role 90. Ferrannini E, Natali A, Camastra S et al (2013) Early metabolic
of perceived neighbourhood crime in the longitudinal association markers of the development of dysglycemia and type 2 diabetes
between perceived built environment and type 2 diabetes mellitus: and their physiological significance. Diabetes 62(5):1730–1737.
a moderated mediation analysis. J Epidemiol Community Health https://doi.org/10.2337/db12-0707
75(2):120–127. https://doi.org/10.1136/jech-2020-214175 91. Floegel A, Stefan N, Yu Z et al (2013) Identification of serum
74. Christakis NA, Fowler JH (2007) The spread of obesity in a large metabolites associated with risk of type 2 diabetes using a targeted
social network over 32 years. N Engl J Med 357(4):370–379. metabolomic approach. Diabetes 62(2):639–648. https://doi.org/
https://doi.org/10.1056/NEJMsa066082 10.2337/db12-0495
75. Choi YJ, Ailshire JA, Crimmins EM (2020) Living alone, social 92. Padberg I, Peter E, Gonzalez-Maldonado S et al (2014) A new
networks in neighbourhoods, and daily fruit and vegetable metabolomic signature in type-2 diabetes mellitus and its
consumption among middle-aged and older adults in the USA. pathophysiology. PLoS One 9(1):e85082. https://doi.org/10.
Public Health Nutr 23(18):3315–3323. https://doi.org/10.1017/ 1371/journal.pone.0085082
S1368980020002475 93. Wang TJ, Ngo D, Psychogios N et al (2013) 2-Aminoadipic acid
76. Shareck M, Ellaway A (2011) Neighbourhood crime and is a biomarker for diabetes risk. J Clin Invest 123(10):4309–4317.
smoking: the role of objective and perceived crime measures. https://doi.org/10.1172/JCI64801
BMC Public Health 11:930. https://doi.org/10.1186/1471-2458- 94. Wang-Sattler R, Yu Z, Herder C et al (2012) Novel biomarkers for
11-930 pre-diabetes identified by metabolomics. Mol Syst Biol 8:615.
77. McNeill LH, Kreuter MW, Subramanian SV (2006) Social https://doi.org/10.1038/msb.2012.43
environment and physical activity: a review of concepts and 95. Patel CJ, Bhattacharya J, Butte AJ (2010) An environment-wide
evidence. Soc Sci Med 63(4):1011–1022. https://doi.org/10. association study (EWAS) on type 2 diabetes mellitus. PLoS One
1016/j.socscimed.2006.03.012 5(5):e10746. https://doi.org/10.1371/journal.pone.0010746
78. Ng Fat L, Scholes S, Jivraj S (2017) The relationship between 96. Meijer J, Lamoree M, Hamers T et al (2021) An annotation
drinking pattern, social capital, and area-deprivation: findings database for chemicals of emerging concern in exposome
from the health survey for England. J Stud Alcohol Drugs 78(1): research. Environ Int 152:106511. https://doi.org/10.1016/j.
20–29. https://doi.org/10.15288/jsad.2017.78.20 envint.2021.106511
79. Agardh E, Allebeck P, Hallqvist J, Moradi T, Sidorchuk A (2011) 97. Barregard L, Bergstrom G, Fagerberg B (2013) Cadmium
Type 2 diabetes incidence and socio-economic position: a exposure in relation to insulin production, insulin sensitivity and
systematic review and meta-analysis. Int J Epidemiol 40(3):804– type 2 diabetes: a cross-sectional and prospective study in women.
818. https://doi.org/10.1093/ije/dyr029 Environ Res 121:104–109. https://doi.org/10.1016/j.envres.2012.
80. Stringhini S, Batty GD, Bovet P et al (2013) Association of 11.005
lifecourse socioeconomic status with chronic inflammation and 98. He K, Xun P, Liu K, Morris S, Reis J, Guallar E (2013) Mercury
type 2 diabetes risk: the Whitehall II prospective cohort study. exposure in young adulthood and incidence of diabetes later in
PLoS Med 10(7):e1001479. https://doi.org/10.1371/journal. life: the CARDIA trace element study. Diabetes Care 36(6):
pmed.1001479 1584–1589. https://doi.org/10.2337/dc12-1842
81. Gurung M, Li Z, You H et al (2020) Role of gut microbiota in type 99. Hendryx M, Luo J, Chojenta C, Byles JE (2019) Exposure to
2 diabetes pathophysiology. EBioMedicine 51:102590. https:// heavy metals from point pollution sources and risk of incident
doi.org/10.1016/j.ebiom.2019.11.051 type 2 diabetes among women: a prospective cohort analysis. Int
82. Sharma S, Tripathi P (2019) Gut microbiome and type 2 diabetes: J Environ Health Res 1–12. https://doi.org/10.1080/09603123.
where we are and where to go? J Nutr Biochem 63:101–108. 2019.1668545
https://doi.org/10.1016/j.jnutbio.2018.10.003 100. Lin JL, Lin-Tan DT, Yu CC, Li YJ, Huang YY, Li KL (2006)
83. Xu WT, Nie YZ, Yang Z, Lu NH (2016) The crosstalk between Environmental exposure to lead and progressive diabetic
gut microbiota and obesity and related metabolic disorders. Future nephropathy in patients with type II diabetes. Kidney Int 69(11):
Microbiol 11:825–836. https://doi.org/10.2217/fmb-2015-0024 2049–2056. https://doi.org/10.1038/sj.ki.5001505
84. Canfora EE, Meex RCR, Venema K, Blaak EE (2019) Gut 101. Navas-Acien A, Silbergeld EK, Pastor-Barriuso R, Guallar E
microbial metabolites in obesity, NAFLD and T2DM. Nat Rev (2008) Arsenic exposure and prevalence of type 2 diabetes in
Endocrinol 15(5):261–273. https://doi.org/10.1038/s41574-019- US adults. JAMA 300(7):814–822. https://doi.org/10.1001/jama.
0156-z 300.7.814
274 Diabetologia (2022) 65:263–274

102. Airaksinen R, Rantakokko P, Eriksson JG, Blomstedt P, Kajantie 108. Drewnowski A, Buszkiewicz J, Aggarwal A, Rose C, Gupta S,
E, Kiviranta H (2011) Association between type 2 diabetes and Bradshaw A (2020) Obesity and the built environment: a
exposure to persistent organic pollutants. Diabetes Care 34(9): reappraisal. Obesity (Silver Spring) 28(1):22–30. https://doi.org/
1972–1979. https://doi.org/10.2337/dc10-2303 10.1002/oby.22672
103. Wolf K, Bongaerts BWC, Schneider A et al (2019) Persistent 109. Lin CY, Koohsari MJ, Liao Y et al (2020) Workplace
organic pollutants and the incidence of type 2 diabetes in the neighbourhood built environment and workers' physically-active
CARLA and KORA cohort studies. Environ Int 129:221–228. and sedentary behaviour: a systematic review of observational
https://doi.org/10.1016/j.envint.2019.05.030 studies. Int J Behav Nutr Phys Act 17(1):ARTN 148. https://
104. Boursi B, Mamtani R, Haynes K, Yang YX (2015) The effect of doi.org/10.1186/s12966-020-01055-x
past antibiotic exposure on diabetes risk. Eur J Endocrinol 172(6): 110. Cunningham-Myrie CA, Theall KP, Younger NO et al (2015)
639–648. https://doi.org/10.1530/EJE-14-1163 Associations between neighborhood effects and physical activity,
105. Hansen AB, Ravnskjaer L, Loft S et al (2016) Long-term exposure obesity, and diabetes: the Jamaica health and lifestyle survey
to fine particulate matter and incidence of diabetes in the Danish 2008. J Clin Epidemiol 68(9):970–978. https://doi.org/10.1016/j.
nurse cohort. Environ Int 91:243–250. https://doi.org/10.1016/j. jclinepi.2014.08.004
envint.2016.02.036 111. Diez Roux AV, Mair C (2010) Neighborhoods and health. Ann N
106. Cao XY, Mokhtarian PL, Handy SL (2009) Examining the Y Acad Sci 1186:125–145. https://doi.org/10.1111/j.1749-6632.
impacts of residential self-selection on travel behaviour: a focus 2009.05333.x
on empirical findings. Transp Rev 29(3):359–395. https://doi.org/ 112. Wild CP (2012) The exposome: from concept to utility. Int J
10.1080/01441640802539195 Epidemiol 41(1):24–32. https://doi.org/10.1093/ije/dyr236
107. McCormack GR, Shiell A (2011) In search of causality: a
systematic review of the relationship between the built
environment and physical activity among adults. Int J Behav Publisher’s note Springer Nature remains neutral with regard to
Nutr Phys Act 8:Artn 125. https://doi.org/10.1186/1479-5868-8- jurisdictional claims in published maps and institutional affiliations.
125