Environmental Planning and Non-Communicable Diseases: A Systematic Review on the Role of the Metabolomic Profile

Non-communicable diseases (NCDs) are the major cause of death worldwide and have economic, psychological, and social impacts. Air pollution is the second, contributing to NCDs-related deaths. Metabolomics are a useful diagnostic and prognostic tool for NCDs, as they allow the identification of biomarkers linked to emerging pathologic processes. The aim of the present study was to review the scientific literature on the application of metabolomics profiling in NCDs and to discuss environmental planning actions to assist healthcare systems and public managers based on early metabolic diagnosis. The search was conducted following PRISMA guidelines using Web of Science, Scopus, and PubMed databases with the following MeSH terms: “metabolomics” AND “noncommunicable diseases” AND “air pollution”. Twenty-nine studies were eligible. Eleven involved NCDs prevention, eight addressed diabetes mellitus, insulin resistance, systemic arterial hypertension, or metabolic syndrome. Six studies focused on obesity, two evaluated nonalcoholic fatty liver disease, two studied cancer, and none addressed chronic respiratory diseases. The studies provided insights into the biological pathways associated with NCDs. Understanding the cost of delivering care where there will be a critical increase in NCDs prevalence is crucial to achieving universal health coverage and improving population health by allocating environmental planning and treatment resources.


Introduction
Metabolomics is a rapidly evolving field that deals with the assessment of metabolite changes in response to endogenous and/or exogenous perturbations [1] as well as assists in the identification of biomarkers linked to emerging pathologic processes [2]. Metabolites are important intermediates and end products of metabolism, indicators of processes underlying a given disease, and can predict the responses of such conditions to therapeutic interventions [3]. In this sense, metabolomics is a useful tool, but it is still little applied in the diagnostic and prognostic processes relating to NCDs [3].
The increasing burden of NCDs is a relevant threat for the population worldwide. In addition to being the most major cause of death, they have economic, psychological, and social impacts [4]. The indirect costs of these diseases (loss of productivity, cost of caregivers, etc.) in developed countries exceed by five times their direct costs (treatment and hospitalization) [4]. In resource-limited settings, as in the case of low-and middleincome countries, treatment of NCDs is even more challenging due to their overlap with the burden of infectious diseases [5].r Numerous cases of NCDs have historically been associated with environmental variables. Among them, air pollution in medium and large urban centers has played a leading role in triggering inflammatory processes. Air pollution is the second most major cause contributing to NCDs-related deaths globally [6]. Human exposure to fine particles (PM2.5) is one of the main health concerns linked to mortality [7]. Reductions in air pollution, especially PM2.5, contribute to longevity gains and numerous other health benefits [8].
Developing countries experienced rapid urban development throughout the 20th century, and, consequently, severe challenges are faced by current populations. Among them, serious public health problems have arisen from a lack of strategic urban planning and health services that are essentially focused on the treatment of chronic diseases instead of promotion of early diagnosis, which would enable environmental planning for the territory to build a health promotion scenario. Thus, intersectoral and cross-disciplinary research combining health and environmental planning may contribute to increasing awareness and implementing solutions to this problem [9].
Efforts to improve the prevention and management of NCDs are part of the current international agenda [10]. Target 3.4 of the Sustainable Development Goals (SDGs, from the United Nations 2030 Agenda) aims to reduce premature mortality from NCDs by one third through prevention and treatment. A shared feature of all NCDs is chronic low-grade inflammation, promoted by unhealthy diets, environmental pollutants, microbial exposure, and psychological and biological stress [11]. Hence, acknowledging tools that provide early diagnosis of the biomarkers associated with this condition is imperative to effectively reaching this target.
Despite the publication of a mini review focusing on five conditions in which metabolomic tools have been employed for diagnosis, treatment, or prognosis prediction [3], no comprehensive systematic reviews were found covering this topic in the Cochrane Library database (May 2023). Furthermore, the redesign of urban areas to promote an active lifestyle and a healthier environment [12] and territory ordering that considers public health issues are key elements to promote equity in the achievement of SDGs.
Thus, the aim of the present study was to review the scientific literature on the application of metabolomics in NCDs and to discuss environmental planning actions to assist healthcare systems and public managers based on early diagnosis.

Materials and Methods
The present study was conducted following the "Preferred Reporting Items for Systematic Reviews and Meta-Analyses" guidelines (PRISMA-https://prisma-statement.org/, accessed on 15 June 2023). The criteria employed in the search and presentation of results were based on the PICO approach: population, intervention, comparison, and observation or result [13].
The study protocol awaits PROSPERO registry number (receipt 446279), and Prisma checklist is available (see Supplementary File S1).
Databases consulted were "Web of Science", "Scopus", and "PubMed". No publication date limit was set, and only studies in English involving humans of all ages were included. The search strategy involved the following MeSH terms: "metabolomics" AND "noncommunicable diseases" AND "air pollution"; "metabolomics" AND "noncommunicable diseases".

Results and Discussion
The first search ("metabolomics" AND "noncommunicable diseases" AND "air lution) resulted in zero publications after application of the exclusion criteria. We t proceeded to the second search ("metabolomics" AND "noncommunicable disease which resulted in 80 publications (PubMed n = 8, Scopus n = 25, and Web of Science 47). After checking for duplicate records and applying the exclusion criteria, 12 stu were dismissed, and 39 were excluded for other reasons after abstract or full text read ( Figure 1). Thereby, a total of 29 studies were eligible for the present review (PubMed 5, Scopus n = 7, Web of Science n = 17). Despite the absence of a publication date limit in the search, the oldest paper inclu in the present review was published in 2013, highlighting that this issue represents a q novel field of scientific investigation. Most studies were conducted in Iran (n = 7) an the UK (n = 6), but South Africa (n = 4) and China (n = 3) also had a number of stu covering this subject, as did Finland and Australia (two studies each) and Denmark, Netherlands, Slovenia, Estonia, Spain, the USA, and Ghana, with one publication e ( Figure 2). Despite the absence of a publication date limit in the search, the oldest paper included in the present review was published in 2013, highlighting that this issue represents a quite novel field of scientific investigation. Most studies were conducted in Iran (n = 7) and in the UK (n = 6), but South Africa (n = 4) and China (n = 3) also had a number of studies covering this subject, as did Finland and Australia (two studies each) and Denmark, the Netherlands, Slovenia, Estonia, Spain, the USA, and Ghana, with one publication each ( Figure 2).  Eleven studies involved NCDs prevention (Table 1), and, among them, five studied children or adolescents [14][15][16][17][18]. The remaining investigations focused on the comparison of lifelong athletes and sedentary subjects [19], trained and untrained individuals [20], night shift and day shift workers [21], a dietary intervention aimed at preventing NCDs [22], a retrospective study with patients who were hospitalized or died from NCDs [23], and a follow-up study on usual alcohol consumption [24]. The effects of APOC3(rs138326449) loss of function mutation in lipoprotein metabolism and its potential to affect triglyceride levels were characterized. The association Eleven studies involved NCDs prevention (Table 1), and, among them, five studied children or adolescents [14][15][16][17][18]. The remaining investigations focused on the comparison of lifelong athletes and sedentary subjects [19], trained and untrained individuals [20], night shift and day shift workers [21], a dietary intervention aimed at preventing NCDs [22], a retrospective study with patients who were hospitalized or died from NCDs [23], and a follow-up study on usual alcohol consumption [24]. Children and adults, males and females Aduts presented higher metabolite concentrations than children, and sex differences were also larger in adults when compared to children. There was also considerable evidence of association between parent and child measurements. Identification of risk biomarkers for several non-communicable diseases early in life can provide a target for interventions and follow up with efficacy Breastfeeding duration seems not to influence metabolomic profile after 20 years; rather, current lifestyle and environmental factors have a stronger potential influence [14]. Nonetheless, there is evidence of an association between parent and child metabolite concentration in plasma [15]. Genetics, eating habits, and physical activity may explain the association of parents' body mass index (BMI) with the higher BMI and waist circumference of their offspring at 20 years [14].
Overeating behavior in childhood seems to produce a metabolic profile characterized by chronic inflammation and higher lipid concentrations, a risk factor for NCDs [17]. When comparing children with a normal weight with their overweight or obese counterparts, some fecal metabolites allow differentiation between groups [18]. This may indicate that the intestinal microbiota and its metabolites play an important role in the development of obesity.
Alterations in a common component of triglyceride-rich lipoproteins, APOC3, a small 99-amino-acid peptide, are associated with a substantial decrease in the risk of coronary artery disease, reduction in triglyceride, and changes in VLDL and HDL levels. A follow-up study with more than 13,000 subjects, including children, mothers, and elderly women, was able to confirm and characterize these associations, useful for assessing drug targets in dyslipidemia [16].
Changes in metabolic profile were also verified among night shift workers [21]. They support the association between night shift working and many common NCDs. Opposite to this finding, Orrú et al. [19] evaluated muscle biopsies of lifelong athletes and age-matched untrained subjects. The authors found that physical activity significantly influences the expression of proteins and metabolites involved in the reduction of the onset of NCDs. In fact, only 10 days of aerobic exercise already allows distinction between trained and untrained individuals, assessed by urine metabolomics [20], reinforcing the role of regular exercise in the prevention of NCDs.
Diet also seems to influence urinary metabolomics. Biomarkers from healthy foods are significantly higher after the consumption of a diet following WHO guidelines for prevention of NCDs [22]. Low-income countries suffer from problems of child malnutrition and lack of nutritional quality in school meals [25]. In this sense, the metabolomic profile can play an important role as an effective monitoring and diagnostic tool for understanding nutritional problems and proposing effective public policies to attend to this population.
A cohort study with patients who either died or were hospitalized with NCDs revealed that more than 400 metabolites are common to at least two NCDs [23]. Risk factors such as smoking and low-grade inflammation were identified as antecedents of NCD multimorbidity and have the potential for early prevention.
Another risk factor for NCDs is increased alcohol intake. This behavior was studied by Würtz et al. [24], who observed that it is associated with cardiometabolic risk markers across multiple metabolic pathways. Table 2 presents eight studies involving diabetes [26][27][28][29], insulin resistance [30], systemic arterial hypertension [31,32], and metabolic syndrome [33]. Food neophobia was significantly associated with many health-related biomarkers, most strongly with ω-3 fatty acids. It was also associated with poor overall dietary quality in Finnish individuals. A high level of food neophobia was associated with increased incidence of T2D in the Finnish cohort, whereas an increased incidence of CHD was detected in the Estonian Biobank cohort Arjmand et al., [30] 403 non-diabetic adults (aged 18-75 years) from a population-based study (Iran) Cross-sectional study involving plasma metabolite profiling

Patients with and without IR
A specific metabolomic profile perturbation is associated with IR.
There was a strong positive association between serum BCAAs (valine and leucine), AAAs (tyrosine, tryptophan, and phenylalanine), alanine, and C0 (free carnitine) and IR, while C18:1 (oleoyl L-carnitine) was negatively correlated with IR There seems to be a causal link between red meat intake and type 2 diabetes (T2D) as red meat metabolite score is associated with T2D incidence and potentially with other cardiometabolic diseases [26]. Another eating habit associated with T2D incidence is food neophobia (a behavioral characteristic in which a person withdraws from tasting unfamiliar or new foods). Food neophobia may also by associated with coronary heart disease [29], as this habit is related to health biomarkers.
Some amino acids and acylcarnitines are considered potential risk markers for diabetes as they reflect disturbances in several metabolic pathways among diabetic individuals [28]. Metabolic patterns such as higher levels of leucine and its catabolic intermediates are useful for identifying and monitoring T2D risk prior to disease onset [27].
Insulin resistance (IR) also presents a specific metabolic signature, identified by Arjmand et al. [30]. When compared to healthy individuals, persons with IR present an increase in branched-chain amino acids (valine and leucine), aromatic amino acids (tyrosine, tryptophan, and phenylalanine), alanine, and free carnitine.
As well as in diabetes, high plasma levels of carnitines and acylcarnitines seem to play a crucial role in the development and progression of systemic arterial hypertension, together with low plasma levels of glycine [32]. Hypertensive patients present a distinctive metabolomic profile when compared to normotensive subjects, mainly characterized by alterations in branched-chain amino acid (BCAA) metabolism. Strauss-Kruger et al. [31] conducted a study on masked hypertension (normotensive at the doctor's office but hypertensive out of it) and observed that these changes in BCAA metabolism may be modulated by central adiposity, indicating that metabolic dysfunction may be an underlying contributor to the etiology of hypertension.
A group of interrelated cardiometabolic abnormalities, including hypertension, hyperglycemia, dyslipidemia, and central adiposity, and metabolic syndrome (MetS) have also been studied with a metabolomic approach. The alteration in circulating levels of amino acids and acylcarnitines is related to the increase in the number of MetS components [33].
Six studies focused on obesity [2,[34][35][36][37][38][39] (Table 3). Intake of sugar-sweetened beverages was positively associated with obesity-related markers [36], and a set of 9 amino acids and 10 polar lipids may also be a potential biomarker of adult obesity [38]. Some metabolites are able to discriminate metabolically healthy obesity from metabolically unhealthily obese subjects [2]. Regarding this latter issue, a particular pattern of amino acids and choline-containing phospholipids may assist in the identification of metabolic health among obese patients [2]. Table 3. Studies involving obesity (n = 6).

Reference Population Intervention Comparison Observation
Tareen et al., (2020) [34] 57 Caucasian participants of a weight loss study conducted in the Netherlands aged 32-67 years Low-or very-low-calorie diets for 12 and 5 weeks, respectively, plus follow up Weight loss period and maintenance period Obese individuals who changed their metabolic profiles in response to caloric restriction had a significant retention of lost weight compared to individuals who did not change it. Cellular metabolism was downregulated during weight loss, with gene expression of all major cellular metabolic processes being lowered during weight loss and maintenance. In parallel, gene expression of immune-system-related processes involving interferons and interleukins increased Mendham et al.,

Exercise group and sedentary controls
Exercise training altered specific intramuscular lipid intermediates which were associated with content-driven increases in mitochondrial respiration and not whole-body insulin sensitivity or GLUT-4 protein content. Exercise increases lipid utilization in the more bioenergetically active organelles and membranes, which may prevent future skeletal muscle lipotoxicity Yan et al., (2023) [36] 86 healthy Chinese young adults, 31 men and 55 women Cross-sectional study involving anthropometry measures, food and drink frequency and lifestyle habits questionnaires, metabolomic and lipid profile analysis of plasma samples Association between intake of sugar-sweetened beverages and obesity-related, gut-microbiota-related metabolic markers and blood lipids SSBs were positively associated with obesity-related markers and blood lipids. In contrast, presweetened coffee was negatively associated with the obesity-related traits. A total of 79 metabolites were associated with both SSBs and metabolic markers, particularly obesity markers. Branched-chain amino acid catabolism and aminoacyl-tRNA biosynthesis were identified as links between SSB intake and metabolic health outcomes, relevant to non-communicable diseases [37] Black women from USA (n = 69), South Africa (n = 97), and Ghana (n = 82) Cross-sectional study involving BMI, physical activity, dietary intake, and metabolomic analysis of plasma samples Normal weight and obese black women A common amino acid metabolite profile was associated with obesity in black women from the USA and South Africa but was not detected in Ghanian women. Site-specific, obesity-related metabolites associated with obesity included intermediates in lipid and carbohydrate metabolism, reflecting the populations' habitual diet. Local environment (and consequently lifestyle factors) seems to be a key moderator of obesity Bagheri et al., [38] 200 Iranian obese adults and 100 healthy controls Cross-sectional study involving plasma metabolomic profiling, arterial blood pressure, anthropometric data, dietary intake, and physical activity Obese and healthy controls A metabolomic profile containing 9 amino acids and 10 polar lipids may serve as a potential biomarker of adult obesity. Obese patients had significantly higher blood pressure than controls. No significant associations between metabolites and energy intake or physical activity were observed Also noteworthy is the influence of site-specific, obesity-related metabolites, reflecting the populations' habitual diet and lifestyle habits [37]. This was observed in a study comparing South African and Ghanaian women, in which authors concluded that lifestyle may be a key moderator of obesity [37]. These habits are modifiable risk factors, and individuals who change their metabolic profile in response to caloric restriction have significantly better retention of weight loss compared to obese individuals who have not changed it [34]. In parallel to this finding, exercise training also has the potential to alter specific intramuscular lipid intermediates, indicating that an increase in lipid utilization may prevent skeletal muscle lipotoxicity [35].
The study by Chashmniam et al. [39] evaluated the serum metabolomic profile of patients with nonalcoholic fatty liver disease (NAFLD) and healthy controls and observed alteration in 19 metabolites in NAFLD patients, particularly a reduction in precursors of some amino acids. The authors suggested that supplementation of these amino acids might be useful in the treatment of NAFLD. Supplementation with curcumin in adults with NAFLD for 8 weeks produced a decrease in inflammatory mediators and an effect in some amino acids, tricarboxylic acid cycle metabolites, gut-microbiota-derived metabolites, and bile acids [40], stressing the use of polyphenols as anti-inflammatory compounds.
Both studies with cancer patients compared patients with healthy controls. Amiri-Dashatan et al. [41] studied females with invasive breast cancer and noticed 20 significantly altered metabolites in patients, mostly amino acids and lipids. These results indicate a dysregulation in metabolic pathways, and this information can be useful for identifying diagnostic and prognostic biomarkers. Chinese researchers investigated the metabolomic profile of colorectal cancer patients and observed a clear differentiation of biomarker panel with lipid changes as the disease progressed [42].
Lifestyle and environmental factors have strong potential to positively impact the metabolomic profile of individuals, thus preventing NCDs [14]. There is evidence that longterm physical activity, but also short-term practice, plays an important role in influencing the expression of metabolites involved in NCDs development [19,20]. Furthermore, adhering to the recommended WHO diet [22], reducing red meat intake [26], and developing a welcoming palate [29] are desirable habits to prevent NCDs. Most of these healthy habits are built during childhood and adolescence, and, in this scenario, families, communities, and schools are important actors [43].
In this context, planned cities offer more opportunities for healthy practices and health promotion since green spaces make up the urban fabric. The scientific literature has exhaustively shown the benefits that urban forests bring to the physical and mental health of the population [44][45][46]. There are many authors evidencing that well-planned green spaces can prevent diseases, especially NCDs [47][48][49], with the caveat that, in low-income countries, studies involving urban green areas and health are still limited and need to be redesigned, especially to include longitudinal studies that use more robust methods [50]. In this sense, the assessment of health indicators through the lens of metabolomics can be an alternative to fill this existing gap. This study has some limitations. The studies included reflect the publications indexed under the respective descriptors and may not represent the full picture of the publications in this area. Moreover, among the studies included in the present review, there was none addressing chronic respiratory diseases. Air pollution is under-examined and inadequately addressed by existing approaches to NCD prevention [9]. In addition to affecting the ecological systems, pollutants also affect human health, contributing to the burden of NCDs and their economic and social associated costs [51].
Efforts to equitably allocate NCD resources must include a balance of both prevention and treatment of existing cases [5]. Understanding the cost of delivering care in regions where there will be a critical increase in NCD prevalence is crucial to achieving universal health coverage and improving overall population health [5].
The use of a noninvasive or minimally invasive mode of sample collection for metabolomics analysis is a positive aspect that provides acceptability to potential participants. While studies conducted so far have provided insights on the characterization of biological pathways associated with various NCDs, other factors remain unknown and deserve future investigations to enable novel clinical, industrial, and political applications to benefit populations worldwide.

Conclusions
This systematic review analyzed several studies ranging from those on breastfeeding, muscle biopsies, and the health of night workers to numerous cases of metabolomic profiling in people with diabetes, validating metabolomics markers as predictors of inflammatory processes and NCDs.
There is still a knowledge gap about the metabolomic profile associated with NCDs in low-income countries. However, there is a need to expand the use of the metabolomics tool to other fields of knowledge, promoting new studies that associate the metabolomic profile with aspects of urban planning, since unhealthy or poorly planned environments can trigger NCDs. These novel findings may help resolve open questions about the use of urban green spaces and the benefits to physical and mental health in addition to contributing to the prevention and treatment of future cases of NCDs.