Abstract
Background
The long-term respiratory health effects of artificial sweetener consumption remain unclear. This study aimed to investigate the associations of artificial sweetener intake with the risk of chronic obstructive pulmonary disease (COPD) and lung function, and to explore the potential mediating role of circulating metabolites.
Methods
Habitual intake of artificial sweeteners (teaspoon/day) was assessed via multiple 24-h dietary recalls. The multivariate Cox proportional hazards regression was used to analyze the association between AS and risk of COPD, and linear regression was used to analyze associations with forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1). A mediation analysis was conducted to assess the contribution of plasma metabolic biomarkers.
Results
A total of 164,656 participants were included. Over a median follow-up of 14.88 years, 4132 incident COPD cases were identified. Compared with non-consumption, higher consumption of artificial sweeteners was associated with a 27% increased risk of COPD (HR = 1.27, 95% CI: 1.14, 1.40) and decreased lung function, including FVC (β: –39.48 mL, 95% CI: -56.53, -22.43, p < 0.001) and FEV1 (β: –37.08 mL, 95% CI:-50.91, -23.25, p < 0.001). A significant positive dose–response relationship was observed between artificial sweetener intake and COPD risk, and a negative dose–response relationship between artificial sweetener intake and lung function parameters. Mediation analysis identified several key mediators, including lipids within intermediate-density and low-density lipoprotein particles, apolipoprotein B, and polyunsaturated fatty acids. Among these, phospholipids in large LDL showed the highest mediating effect (PM: 8.72%; 95% CI: 4.53%, 19.63%), followed by cholesteryl esters in IDL (PM: 8.68%; 95% CI: 4.78%, 20.58%). Beyond lipoprotein sub-fractions, apolipoprotein B served as a significant mediator (PM: 6.70%; 95% CI: 3.67%, 16.35%). Furthermore, key fatty acids were identified as significant mediators.
Conclusion
Higher intake of artificial sweeteners is associated with an elevated risk of COPD and impaired lung function. These associations are partially mediated through pathways related to lipoprotein metabolism and fatty acid profiles.
Graphical abstract
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12986-026-01142-z.
Keywords: UK Biobank, Chronic obstructive pulmonary disease, Artificial sweeteners, Metabolic biomarkers
Background
Chronic obstructive pulmonary disease (COPD), characterized by persistent respiratory symptoms and airflow limitation, remains a major global public health challenge [1]. It is currently the third leading cause of death worldwide, resulting in substantial social and economic burdens [2]. Data from the Global Burden of Disease Study 2019 estimated a global prevalence of 212.3 million cases of Chronic Obstructive Pulmonary Disease (COPD). Over the period from 1990 to 2019, the incidence of COPD increased by 86%, while deaths attributable to COPD rose by 30% [3]. The pathophysiology is driven by chronic inflammation that injures the lung parenchyma and airways, sustaining these symptoms and functional deficits [4]. While smoking is established as its primary risk factor [5], a considerable number of COPD cases occur in non-smokers [6], suggesting that other modifiable environmental and lifestyle factors play a crucial role [7–9].
Among various lifestyle factors, the relationship between dietary patterns and respiratory health is increasingly gaining attention. As a typical representative of modern industrial food systems, ultra-processed foods (UPFs) are closely associated with an increased risk of various chronic non-communicable diseases, such as obesity, type 2 diabetes, cardiovascular diseases and COPD [10, 11]. Notably, emerging epidemiological evidence has begun to reveal adverse associations between UPFs consumption and respiratory health and has indicated that higher UPF intake is significantly positively correlated with an increased incidence of COPD [12]. Artificially sweetened beverages (ASBs), a prominent category of UPFs, have been independently implicated in impaired respiratory health, such as an elevated risk of asthma and worsening lung function [13]. Moreover, a critical limitation of the existing epidemiological evidence is that this method fails to account for potential confounding by other beverage constituents and ignores tabletop sweeteners—an important exposure source with simpler ingredient profiles. Artificial sweeteners, widely used as sugar substitutes in "sugar-free" beverages and UPFs, primarily include aspartame and saccharin [14]. Although regulatory agencies like the U.S. FDA approve these sweeteners as safe based on short-term studies, accumulating evidence suggests that long-term intake is associated with adverse metabolic outcomes, including increased visceral adiposity and metabolic syndrome [15, 16]. COPD is characterized by disruptions in multiple metabolic pathways, including aberrations in amino acid metabolism, lipid processing, and increased oxidative stress [17–19]. Similarly, experimental studies, particularly on aspartame, have documented its capacity to induce remarkably similar disturbances, including impaired glucose regulation, perturbed lipid metabolism, promoted lipid peroxidation, and dysregulated adipokine signaling [20–22]. However, the translation of these experimental findings to human health outcomes, particularly in the context of COPD, is not fully established and needs further investigation in large human cohorts. Performing mediation analysis on factors linked to both COPD and its risk factors, such as artificial sweeteners, could elucidate potential mechanistic pathways. The UK Biobank, with its extensive data on plasma metabolites and inflammatory markers, presents a valuable resource to facilitate such analyses.
Utilizing data from the UK Biobank, this study assessed the relationships between artificial sweetener consumption, lung function decline, and COPD incidence. We propose the hypothesis that the adverse effects of artificial sweeteners on respiratory health are mediated through their influence on biomarkers. A mediation analysis was therefore conducted to investigate the role of these biomarkers.
Methods
Study population
This investigation utilized data obtained from the UK Biobank, a large-scale prospective study that enrolled more than half a million individuals aged 40 to 69 from 2006 to 2010. Electronic informed consent was secured from every participant. Further details of the study design and methods are described in the Supplementary Methods.
Ascertainment of COPD and lung function
Prevalent COPD cases at baseline were determined through a combination of self-reported physician diagnoses and prior hospital records. Incident cases of COPD-related hospitalizations and mortality during follow-up were identified from national health and death registries. Diagnoses were defined according to the International Classification of Diseases, Tenth Revision (ICD-10), using code J44. The duration of follow-up was calculated from the date of enrollment until the first occurrence of COPD, death, or July 1, 2022, whichever came first. At the baseline assessment, pre-bronchodilator spirometry was conducted on eligible participants to obtain measurements. Measurements included forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC). These were obtained from two to three maneuvers conducted over approximately six minutes, using a Vitalograph Pneumotrac 6800 spirometer. A third exhalation was omitted if the difference between the first two acceptable maneuvers for FEV₁ and FVC was within 5%.
Assessment of artificial sweetener intake
Dietary intake data were collected using the Oxford WebQ (https://www.ceu.ox.ac.uk/research/oxford-webq), a web-based 24-h dietary recall tool designed to capture consumption of common foods and beverages [23]. Participants completed the questionnaire on multiple occasions over a five-year period, and mean values were calculated from available responses to represent habitual intake. Artificial sweetener consumption was specifically assessed through participants’ reports of the number of teaspoons of non-caloric sweeteners added to coffee, tea, and cereal, corresponding to UK Biobank data-field IDs 100,380, 100,500, and 100,910, respectively. To compute total daily artificial sweetener intake, we first quantified consumption frequencies of the relevant food categories. Total coffee consumption was derived by summing daily cups across six mutually exclusive coffee types (instant, filtered, cappuccino, latte, espresso, and other coffee). Similarly, total tea intake was calculated from five tea categories (standard, rooibos, green, herbal, and other tea), while cereal consumption was aggregated from eight cereal varieties (porridge, muesli, oat crunch, sweetened cereal, plain cereal, bran cereal, whole-wheat cereal, and other cereal). The amount of artificial sweetener added to each food category was then multiplied by its respective consumption frequency. Finally, total artificial sweetener intake was determined by summing the sweetener quantities across all three categories, expressed as teaspoons per day, providing a comprehensive measure of tabletop artificial sweetener consumption.
Assessment of covariates
Based on a review of the existing literature, we identified and adjusted for a comprehensive set of potential confounding factors [12, 24]. These included demographic background, socioeconomic status, lifestyle factors, body measurements, and health history. Demographic data comprised age, gender (male or female), and ethnic background (classified as White, Mixed, South Asian, Black, Chinese, or other). Socioeconomic position was evaluated using both educational attainment (grouped into no relevant qualifications, college or university degree, A levels/AS levels or equivalent, O levels/GCSEs or equivalent, CSEs or equivalent, or others) and the Townsend deprivation index, a geographic indicator of material disadvantage. Lifestyle factors included smoking behavior (current, former, never) and alcohol drinking (never, previous, current). Healthy diet score was constructed based on recommended intake levels of various food groups linked to cardiovascular wellness, emphasizing fruits, vegetables, whole grains, fish, dairy products and vegetable oils, with reduced or no consumption of refined grains, processed or unprocessed meats and sugar-sweetened beverages [25]. Physical activity during leisure time was captured by asking participants about their engagement in various types of activities over the previous four weeks, such as walking for pleasure, light do-it-yourself (DIY) projects, heavy DIY work, strenuous sports, other exercises, or none of the above. For respondents reporting any activity, we recorded how often and for how long they performed it. Activity intensity was assigned standard metabolic equivalent of task (MET) values, defined as the ratio of activity-related energy expenditure to the resting metabolic rate. Following this, overall weekly volume (MET-minutes/week) was determined by multiplying the frequency, duration, and MET value for each activity, then summing the results across all activities. The assigned MET values were as follows: 3.5 for walking for pleasure, 2.5 for light DIY, 5.5 for heavy DIY, 8.0 for strenuous sports, and 4.0 for other physical activities [26]. Furthermore, body mass index (BMI) was derived from measured height and weight using the standard formula: weight (kg)/height (m)2. Finally, environmental exposures to air pollutants were adjusted for, including nitrogen dioxide (NO₂), nitrogen oxides (NOX), particulate matter with a diameter ≤ 2.5 μm (PM2.5), particulate matter with a diameter ≤ 10 μm (PM10), and particulate matter with a diameter between 2.5 and 10 μm (PM2.5–10), with each pollutant variable incorporated as quintiles or as missing.
Assessment of the mediating biomarkers
Plasma metabolomic profiling was initiated during the first assessment visit (2006–2010), following informed consent. The first repeat visit took place in 2012–2013, while metabolomic quantification was conducted in two phases: Phase 1 from June 2019 to April 2020 and Phase 2 from April 2020 to June 2022. Metabolomic assays using nuclear magnetic resonance (NMR) spectroscopy were performed by Nightingale Health Ltd across both phases. Detailed procedures for sample collection and metabolite quantification have been described previously [27]. Over 280,000 participants had metabolomic data available, with an additional 16,000 undergoing repeat assessment. The present analysis focused on 143 directly quantified metabolites that cannot be derived from other measures [28], including 104 lipoprotein and lipid measures, 3 glycolysis-related metabolites, 15 fatty acids, 9 amino acids, 5 ketone bodies, 4 choline-containing compounds, 1 inflammatory marker, and 2 fluid balance-related metabolites, to identify potential mediators of the association between artificial sweetener intake and COPD.
Statistical analysis
Baseline characteristics were stratified according to the median intake of artificial sweeteners. Continuous variables are presented as means and standard deviations (SDs), while categorical variables are expressed as numbers and percentages.
The Cox proportional hazards models were used to examine the association between artificial sweeteners and risk of COPD. Subsequently, linear regression was used to analyze the relationship between artificial sweeteners and lung function (FVC/FEV1). Effect estimates were presented as hazard ratios (HRs) or β coefficients with their 95% confidence intervals (CIs). Person-years of follow-up were calculated from baseline until the date of death, diagnosis, loss to follow-up, or the end of the follow-up period, whichever occurred first. Four models were constructed in the primary analysis: a crude model without any adjustments; Model 1 adjusted for age, sex, race (White, Mixed, South Asian, Black, Chinese, Other, or missing), educational attainment (no qualifications, college or university degree, A-level/AS-level or equivalent, O-level/GCSEs or equivalent, CSEs or equivalent, other qualifications, or missing), and Townsend deprivation index (quintiles or missing); Model 2 further adjusted for BMI (< 18.5, 18.5 to < 25, 25 to < 30, ≥ 30 kg/m2, or missing), smoking status (never, former, current, or missing), alcohol drinking (never, former, current, or missing), healthy diet (yes, no, or missing), leisure-time physical activity (< 500, 500 to < 1000, ≥ 1000 MET-minutes/week, or missing), family history of lung disease (yes, no, or missing), and diabetes (yes, no, or missing); Model 3 additionally adjusted for air pollutants, including NO2 (quintiles or missing), NOx (quintiles or missing), PM2.5 (quintiles or missing), PM10 (quintiles or missing), and PM2.5–10 (quintiles or missing). Furthermore, restricted cubic spline (RCS) models were applied to examine potential dose–response relationships.
We performed stratified analyses by age, sex, educational attainment, Townsend deprivation index, BMI, smoking status, alcohol drinking, healthy diet, leisure time physical activity, diabetes, family history of chronic lung disease, NO2, NOX, PM2.5, PM10, and PM2.5–10 to assess the association between artificial sweeteners and COPD risk and lung function parameters across different socioeconomic, familial, and environmental exposure subgroups. Furthermore, to examine the robustness of the findings, a sensitivity analysis of the association between artificial sweetener intake and incident COPD risk was conducted, excluding participants with follow-up duration shorter than 3 or 5 years.
Moreover, we conducted a mediation analysis to investigate whether the association between artificial sweetener intake (exposure) and COPD (outcome) could be explained by metabolic biomarkers (mediators). Based on data availability in the UK Biobank, 143 biomarkers were included in the mediation analysis. The mediation model was estimated using a bootstrap procedure with 1,000 repetitions, which comprised two components: the mediator model and the outcome model. The mediator fit was specified as “Mediator ~ Covariates + Exposure”, while the outcome fit was defined as “Outcome ~ Covariates + Exposure + Mediator”. The indirect effect was calculated as the product of the effect of the exposure on the mediator and the effect of the mediator on the outcome. The proportion mediated was computed as the indirect effect divided by the total effect, and its 95%CI was estimated using the bootstrapping method [29, 30].
A two-tailed P-value < 0.05 was used for primary analyses. For mediation analysis of 143 biomarkers, we applied the Benjamini–Hochberg false discovery rate (FDR) correction to control for multiple comparisons. Statistical significance was defined as FDR controlled P < 0.01. All analyses used R version 4.1.1 (https://www.R-project.org/).
Results
Characteristics of the study participants
A total of 164,656 participants were included in this study, with mean age of 56.46 years, and 45.7% were male (Supplementary Figure S1). Table 1 presents the baseline characteristics of the study participants stratified by artificial sweetener consumption categories. The total study population categorized as non-consumers (n = 147,639), lower consumers (n = 8,711), and higher consumers (n = 8,306). Significant differences were observed across groups for most demographic, lifestyle, and clinical characteristics. Participants with higher artificial sweetener consumption were older, more likely to be male, and had a higher prevalence of obesity (BMI ≥ 30 kg/m2), previous or current smoking, and diabetes. They also exhibited lower educational attainment, higher socioeconomic deprivation, and a lower proportion of healthy diet adherence. In contrast, no significant differences were observed in air pollutant exposures (NO₂, NOₓ, PM₂.₅, PM₁₀, and PM₂.₅-₁₀) across consumption categories.
Table 1.
Baseline characteristics of the study population
| Total population (n = 164,656) |
Non-consumers (n = 147,639) |
Lower consumers (n = 8711) |
Higher consumers (n = 8306) |
P-value | |
|---|---|---|---|---|---|
| Age, mean (SD), years | 56.46 ± 7.88 | 56.29 ± 7.88 | 57.62 ± 7.82 | 58.21 ± 7.48 | < 0.001 |
| Male, n (%) | 75,246 (45.70) | 67,367 (45.63) | 3676 (42.20) | 4203 (50.60) | < 0.001 |
| Ethnicity, n (%) | < 0.001 | ||||
| White | 156,887 (95.28) | 140,730 (95.32) | 8168 (93.77) | 7989 (96.18) | |
| Mixed background | 995 (0.60) | 907 (0.61) | 56 (0.64) | 32 (0.39) | |
| South Asian | 2368 (1.44) | 2034 (1.38) | 222 (2.55) | 112 (1.35) | |
| Black | 2048 (1.24) | 1836 (1.24) | 138 (1.58) | 74 (0.89) | |
| Chinese | 522 (0.32) | 496 (0.34) | 17 (0.20) | 9 (0.11) | |
| Other | 1204 (0.73) | 1060 (0.72) | 87 (1.00) | 57 (0.69) | |
| Missing | 632 (0.38) | 576 (0.39) | 23 (0.26) | 33 (0.40) | |
| Educational attainment, n (%) | < 0.001 | ||||
| College or University degree | 72,004 (43.73) | 66,887 (45.30) | 2783 (31.95) | 2334 (28.10) | |
| A levels/AS levels or equivalent | 21,221 (12.89) | 19,050 (12.90) | 1122 (12.88) | 1049 (12.63) | |
| O levels/GCSEs or equivalent | 33,201 (20.16) | 28,902 (19.58) | 2158 (24.77) | 2141 (25.78) | |
| CSEs or equivalent | 6602 (4.01) | 5693 (3.86) | 451 (5.18) | 458 (5.51) | |
| Others | 16,918 (10.27) | 14,853 (10.06) | 1058 (12.15) | 1007 (12.12) | |
| No relevant qualifications | 13,902 (8.44) | 11,543 (7.82) | 1081 (12.41) | 1278 (15.39) | |
| Missing | 808 (0.49) | 711 (0.48) | 58 (0.67) | 39 (0.47) | |
| Socioeconomic deprivation index, mean (SD) | 75,246 (45.70) | 67,367 (45.63) | 3676 (42.20) | 4203 (50.60) | < 0.001 |
| Missing | 184 (0.11) | 156 (0.11) | 19 (0.22) | 9 (0.11) | |
| Body mass index, n (%) | < 0.001 | ||||
| < 18.5 kg/m2 | 999 (0.61) | 960 (0.65) | 28 (0.32) | 11 (0.13) | |
| 18.5 to < 25 kg/m2 | 65,215 (39.61) | 60,885 (41.24) | 2354 (27.02) | 1976 (23.79) | |
| 25 to < 30 kg/m2 | 67,921 (41.25) | 60,136 (40.73) | 3965 (45.52) | 3820 (45.99) | |
| ≥ 30 kg/m2 | 30,055 (18.25) | 25,252 (17.10) | 2336 (26.82) | 2467 (29.70) | |
| Missing | 466 (0.28) | 406 (0.27) | 28 (0.32) | 32 (0.39) | |
| Smoking status, n (%) | < 0.001 | ||||
| Never | 93,956 (57.06) | 86,105 (58.32) | 4445 (51.03) | 3406 (41.01) | |
| Previous | 57,545 (34.95) | 50,183 (33.99) | 3581 (41.11) | 3781 (45.52) | |
| Current | 12,716 (7.72) | 10,964 (7.43) | 657 (7.54) | 1095 (13.18) | |
| Missing | 439 (0.27) | 387 (0.26) | 28 (0.32) | 24 (0.29) | |
| Alcohol drinking, n (%) | < 0.001 | ||||
| Never | 5261 (3.20) | 4550 (3.08) | 362 (4.16) | 349 (4.20) | |
| Previous | 4681 (2.84) | 4072 (2.76) | 249 (2.86) | 360 (4.33) | |
| Current | 154,550 (93.86) | 138,871 (94.06) | 8085 (92.81) | 7594 (91.43) | |
| Missing | 164 (0.10) | 146 (0.10) | 15 (0.17) | 3 (0.04) | |
| Healthy diet, n (%) | 29,083 (17.66) | 26,580 (18.00) | 1366 (15.68) | 1137 (13.69) | < 0.001 |
| Missing | 3414 (2.07) | 3031 (2.05) | 170 (1.95) | 213 (2.56) | |
| Leisure time physical activity, n (%) | < 0.001 | ||||
| < 500 MET mins/week | 52,679 (31.99) | 46,639 (31.59) | 3128 (35.91) | 2912 (35.06) | |
| 500 to < 1000 MET mins/week | 37,604 (22.84) | 33,878 (22.95) | 1930 (22.16) | 1796 (21.62) | |
| ≥ 1000 MET mins/week | 66,381 (40.31) | 60,306 (40.85) | 3111 (35.71) | 2964 (35.69) | |
| Missing | 7992 (4.85) | 6816 (4.62) | 542 (6.22) | 634 (7.63) | |
| Diabetes, n (%) | 5550 (3.37) | 4103 (2.78) | 703 (8.07) | 744 (8.96) | < 0.001 |
| Missing | 344 (0.21) | 288 (0.20) | 24 (0.28) | 32 (0.39) | |
| Family history of chronic lung disease, n (%) | 22,569 (13.71) | 19,888 (13.47) | 1330 (15.27) | 1351 (16.27) | < 0.001 |
| Missing | 2271 (1.38) | 1977 (1.34) | 144 (1.65) | 150 (1.81) | |
| NO2, mean (SD) | 26.45 ± 7.72 | 26.47 ± 7.76 | 26.34 ± 7.42 | 26.33 ± 7.31 | 0.112 |
| Missing | 2588 (1.57) | 2318 (1.57) | 139 (1.60) | 131 (1.58) | |
| NOX, mean (SD) | 43.27 ± 15.43 | 43.26 ± 15.44 | 43.28 ± 15.35 | 43.45 ± 15.20 | 0.587 |
| Missing | 2588 (1.57) | 2318 (1.57) | 139 (1.60) | 131 (1.58) | |
| PM2.5, mean (SD) | 9.91 ± 1.03 | 9.91 ± 1.03 | 9.92 ± 0.99 | 9.92 ± 1.00 | 0.293 |
| Missing | 10,021 (6.09) | 9056 (6.13) | 487 (5.59) | 478 (5.75) | |
| PM10, mean (SD) | 16.20 ± 1.89 | 16.20 ± 1.90 | 16.21 ± 1.88 | 16.24 ± 1.87 | 0.229 |
| Missing | 10,021 (6.09) | 9056 (6.13) | 487 (5.59) | 478 (5.75) | |
| PM2.5–10, mean (SD) | 6.42 ± 0.88 | 6.42 ± 0.88 | 6.42 ± 0.89 | 6.42 ± 0.88 | 0.865 |
| Missing | 10,021 (6.09) | 9056 (6.13) | 487 (5.59) | 478 (5.75) |
NO2: nitrogen dioxide, NOX: nitrogen oxides, PM2.5: particulate matter with a diameter ≤ 2.5 μm, PM10: particulate matter with a diameter ≤ 10 μm
Association of artificial sweetener intake with risk of incident COPD and lung function
After a median follow-up of 14.88 years (interquartile range [IQR]: 12.51 to 17.15), 4132 individuals developed COPD. Table 2 summarizes the associations between artificial sweetener intake and the risk of incident COPD and lung function decline. In the fully adjusted model, each one-teaspoon increase in artificial sweetener intake was associated with a significantly increased risk of COPD (HR: 1.02, 95% CI: 1.02, 1.03, p < 0.001). Compared with non-consumers, higher consumers had a significantly elevated risk of COPD (HR: 1.27, 95% CI: 1.14, 1.40, p < 0.001), while lower consumers did not show a significant association (HR: 1.08, 95% CI: 0.95, 1.22, p = 0.225). Similarly, higher sweetener intake was associated with reduced lung function, as measured by FVC and FEV₁. In Model 3, higher consumers showed significant declines in FVC (β: –39.48 mL, 95% CI: -56.53, -22.43, p < 0.001) and FEV₁ (β: –37.08 mL, 95% CI:-50.91, -23.25, p < 0.001) compared to non-consumers. The RCS results demonstrate the significant positive dose–response relationship between artificial sweetener consumption and risk of COPD and the negative dose–response relationship between artificial sweetener consumption and lung function level (Fig. 1).
Table 2.
Association of artificial sweeteners intake with risk of incident COPD and lung function
| Per 1-teaspoon increase in artificial sweeteners |
Non-consumers | Lower consumers | Higher consumers | ||||
|---|---|---|---|---|---|---|---|
| HR/β (95% CI) | P-value | HR/β (95% CI) | HR/β (95% CI) | P-value | HR/β (95% CI) | P-value | |
| COPD | |||||||
| Events/total | 4132/164,656 | 3418/147,639 | 287/8711 | 427/8306 | |||
| Crude | 1.08 (1.07, 1.09) | < 0.001 | 1.0 (Ref.) | 1.44 (1.28, 1.62) | < 0.001 | 2.28 (2.06, 2.52) | < 0.001 |
| Model 1 | 1.05 (1.05, 1.06) | < 0.001 | 1.0 (Ref.) | 1.19 (1.05, 1.34) | 0.005 | 1.68 (1.52, 1.86) | < 0.001 |
| Model 2 | 1.02 (1.02, 1.03) | < 0.001 | 1.0 (Ref.) | 1.08 (0.96, 1.22) | 0.211 | 1.26 (1.14, 1.40) | < 0.001 |
| Model 3 | 1.02 (1.02, 1.03) | < 0.001 | 1.0 (Ref.) | 1.08 (0.95, 1.22) | 0.225 | 1.27 (1.14, 1.40) | < 0.001 |
| Lung function parameters | |||||||
| FVC | |||||||
| Events/total | 2935/122,425 | 2434/110,033 | 204/6296 | 297/6096 | |||
| Crude | -13.80 (-16.41, -11.19) | < 0.001 | 0.0 (Ref.) | -188.07 (-212.73, -163.41) | < 0.001 | -136.27 (-161.31, -111.23) | < 0.001 |
| Model 1 | -10.30 (-12.10, -8.50) | < 0.001 | 0.0 (Ref.) | -72.60 (-89.58, -55.62) | < 0.001 | -98.57 (-115.85, -81.30) | < 0.001 |
| Model 2 | -3.70 (-5.48, -1.93) | < 0.001 | 0.0 (Ref.) | -27.88 (-44.57, -11.18) | 0.001 | -39.66 (-56.71, -22.61) | < 0.001 |
| Model 3 | -3.69 (-5.47, -1.91) | < 0.001 | 0.0 (Ref.) | -27.83 (-44.52, -11.13) | 0.001 | -39.48 (-56.53, -22.43) | < 0.001 |
| FEV1 | |||||||
| Events/total | 2935/122,425 | 2434/110,033 | 204/6296 | 297/6096 | |||
| Crude | -12.29 (-14.33, -10.25) | < 0.001 | 0.0 (Ref.) | -138.03 (-157.36, -118.71) | < 0.001 | -119.74 (-139.36, -100.11) | < 0.001 |
| Model 1 | -8.33 (-9.78, -6.88) | < 0.001 | 0.0 (Ref.) | -43.63 (-57.31, -29.96) | < 0.001 | -78.47 (-92.39, -64.56) | < 0.001 |
| Model 2 | -3.68 (-5.12, -2.24) | < 0.001 | 0.0 (Ref.) | -18.08 (-31.62, -4.54) | 0.009 | -37.12 (-50.95, -23.29) | < 0.001 |
| Model 3 | -3.68 (-5.12, -2.24) | < 0.001 | 0.0 (Ref.) | -17.99 (-31.53, -4.45) | 0.009 | -37.08 (-50.91, -23.25) | < 0.001 |
Model 1 was adjusted age, sex, ethnicity (White, Mixed, South Asian, Black, Chinese, other races, or missing), educational attainment (no relevant qualifications, college or university degree, A levels/AS levels or equivalent, O levels/GCSEs or equivalent, CSEs or equivalent, others, or missing), and socioeconomic deprivation index (quintiles, or missing); Model 2 was further adjusted for body mass index (< 18.5, 18.5 to < 25, 25 to < 30, ≥ 30 kg/m2, or missing), smoking status (never, previous, current, or missing), alcohol drinking (never, previous, current, or missing), healthy diet (yes, no, or missing), leisure time physical activity (< 500, 500 to < 1000, ≥ 1000 MET mins/week, or missing), family history of lung diseases (yes, no, or missing), diabetes (yes, no, or missing); Model 3 was further adjusted for air pollutants (NO2 [quintiles, or missing], NOX [quintiles, or missing], PM2.5 [quintiles, or missing], PM10 [quintiles, or missing], PM2.5–10 [quintiles, or missing])
Fig. 1.
Dose–response associations of artificial sweetener intake with COPD risk and lung function. (A) Hazard ratio for COPD risk by daily intake of artificial sweeteners (teaspoon/day). P overall < 0.001, P non-linearity < 0.001. (B) Association between artificial sweetener intake and baseline FVC (mL). P overall < 0.001, P non-linearity = 0.024. (C) Association between artificial sweetener intake and baseline FEV1 (mL). P overall < 0.001, P non-linearity < 0.001
Stratified analysis and sensitivity analysis
Based on the stratified analyses presented in Supplementary Tables S1–S3, the association between artificial sweetener intake and incident COPD risk was generally consistent across most subgroups, though some effect modifications were observed. Significant multiplicative interactions were identified for smoking status (P for interaction = 0.010) and leisure-time physical activity (P for interaction = 0.018), suggesting that the association was more pronounced among former or current smokers and those with lower physical activity levels.
A series of sensitivity analyses further supported the robustness of the primary findings. Excluding participants with follow-up duration shorter than 3 or 5 years did not materially alter the results (Supplementary Table S4). The association remained significant after adjusting for multiple covariates, indicating that the observed relationship is not substantially confounded by these variables. We conducted additional sensitivity analyses excluding participants with COPD events occurring within 3 years of follow-up, stratified by obesity status. These analyses showed that the association remained significant in both non-obese and obese subgroups after excluding early events (Supplementary Table S5).
Mediation analysis
We performed a comprehensive mediation analysis to assess the role of 143 circulating metabolites in the association between artificial sweetener consumption and COPD. After FDR correction for testing 143 metabolites, 63 showed significant mediating effects. Lipoproteins constituted the majority of findings (n = 50; 79.4% of selected metabolites), involved key components such as cholesterol, cholesteryl esters, and phospholipids. Additionally, we identified significant fatty acids (n = 8), as well as other lipids (n = 4) and other metabolites (n = 1) (Fig. 2, Supplementary Table S6). The most prominent mediators were components of low-density lipoprotein (LDL) and related particles (Fig. 3). Phospholipids in large LDL showed the highest mediating effect (PM: 8.72%, 95% CI: 4.53% to 19.63%). Similarly, the cholesteryl esters content in intermediate-density lipoprotein (IDL) demonstrated significant mediation (PM: 8.68%, 95% CI: 4.78% to 20.58%), as did cholesterol in large, medium, and small LDL particles (PM: 4.20%, 5.86%, and 5.80%, respectively). Beyond lipoprotein subfractions, apolipoprotein B (ApoB) served as a significant mediator (PM: 6.70%, 95% CI: 3.67% to 16.35%), while the ApoB to ApoA1 ratio exhibited a weaker effect (PM: 1.87%, 95% CI: 0.53% to 4.99%). Furthermore, key fatty acids were identified as significant mediators, including linoleic acid (PM: 7.16%, 95% CI: 3.92% to 16.47%) and the total pool of polyunsaturated fatty acids (PM: 6.55%, 95% CI: 3.35% to 18.14%). In conclusion, the relationship between artificial sweetener intake and COPD is partially mediated through multiple pathways related to LDL metabolism, ApoB, and specific fatty acids, with each explaining a small portion of the total effect.
Fig. 2.
The estimated mediation proportions of metabolic biomarkers on the associations of artificial sweetener intake with COPD risk. Models were adjusted for age, sex, ethnicity, educational attainmen, socioeconomic deprivation index, body mass index, smoking status, alcohol drinking, healthy diet, leisure time physical activity, family history of lung diseases, diabetes, and air pollutants (NO2, NOX, PM2.5, PM10, PM2.5-10). Asterisks indicate statistical significance (*P < 0.05, **P < 0.01, ***P < 0.001). Apo, apolipoprotein; DHA, docosahexaenoic acid; FA, fatty acids; acid; HDL, high-density lipoproteins; IDL, intermediate-density lipoproteins; L, large; LA, linoleic acid; LDL, low-density lipoproteins; LP, lipoprotein; M, medium; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; S, small; SFA, saturated fatty acids; VLDL, very low-density lipoproteins; XL, very large; XS, very small; XXL, extremely large
Fig. 3.
Mediating effects of key mediating metabolic biomarkers on the associations between artificial sweetener intake and COPD risk: cholesterol in IDL, total lipids in large LDL, free cholesterol in IDL, cholesteryl esters in IDL, phospholipids in large LDL. Models were adjusted for age, sex, ethnicity, educational attainmen, socioeconomic deprivation index, body mass index, smoking status, alcohol drinking, healthy diet, leisure time physical activity, family history of lung diseases, diabetes, and air pollutants (NO2, NOX, PM2.5, PM10, PM2.5–10). Prop., proportion mediated; ***P < 0.001
Discussion
This large-scale prospective cohort study within the UK Biobank demonstrated that participants with higher consumption of artificial sweeteners had an elevated risk of COPD and a decreased lung function level compared to those without consumption of artificial sweeteners. These associations remained robust after extensive adjustment for potential confounders and were consistent across all sensitivity analyses. Our mediation analysis indicates that the association between artificial sweetener consumption and COPD is partially explained by the concerted action of specific metabolites. These metabolites are primarily involved in LDL metabolism, apolipoprotein B, and key polyunsaturated fatty acids. However, the proportion of the total effect mediated by each individual pathway was modest.
The potential relationship between artificial sweetener intake and COPD risk remains an unexamined area in nutritional epidemiology. While previous research has explored the role of diet in chronic respiratory diseases, investigations into UPFs and COPD have yielded conflicting conclusions [12, 31]. This inconsistency may stem from the substantial heterogeneity inherent in the UPF category, which encompasses a wide array of products with divergent nutritional compositions, thereby obscuring the specific effects of constituents like artificial sweeteners. More direct evidence comes from studies on sugary beverages. A recent large-scale prospective cohort study from the UK Biobank reported that higher consumption of ASBs was associated with an increased risk of incident COPD and asthma [13]. Similarly, several cross-sectional studies have identified a positive association between ASBs consumption and the prevalence of both COPD and asthma [32–36]. These observations direct attention to the common feature of beverage sweeteners. However, much of the existing literature has focused on the detrimental role of sugars in sugar-sweetened beverages or has used ASBs as a proxy for artificial sweeteners. This approach fails to isolate the independent effect of artificial sweeteners themselves, as the observed associations could be confounded by other compounds present in these beverages. Furthermore, a significant limitation of prior studies is the general neglect of tabletop artificial sweeteners (e.g., those added to coffee, tea, or food), which constitute a major dietary source of artificial sweeteners and have simpler ingredient lists, potentially providing a clearer perspective on their health effects. Our study was specifically designed to address this pivotal question.
To generate hypotheses on potential biological pathways, we conducted an exploratory mediation analysis to identify metabolites that may mediate the association between artificial sweetener intake and COPD risk. After FDR correction for testing 143 metabolites, 63 showed significant mediating effects. The significant mediators were predominantly clustered within a few highly correlated metabolic classes, notably pro-atherogenic lipoprotein profiles (e.g., medium-density and all subclasses of LDL particles and apolipoprotein B) and polyunsaturated fatty acids (e.g., linoleic acid). However, it is crucial to emphasize that the mediating proportions were modest (< 10%) and that significant col-linearity exists among these metabolites. Therefore, these results should be interpreted with caution and are primarily hypothesis-generating rather than conclusive proof of mechanism. These findings are consistent with previous research linking lipid metabolism to respiratory health. A recent large-scale study using UK Biobank data investigated the association between air pollution exposure and COPD risk [24]. The study found that air pollution was associated with both lung function decline and increased COPD incidence. Importantly, mediation analysis revealed that lipid-related metabolites significantly mediated this relationship [24]. Previous studies have suggested that the lungs represent a target organ of systemic damage in cardiovascular disease (CVD) [37]. Existing CVD conditions, such as heart failure and ischemic heart disease, have been shown to significantly increase the risk of disease exacerbation and mortality in COPD patients. Arterial stiffness, a modifiable risk factor for CVD, is markedly elevated in individuals with COPD, indicating its potential utility in risk stratification and as a target for intervention [38, 39]. Nonetheless, these exploratory findings offer novel, data-driven clues for generating hypotheses about a potential link between artificial sweetener intake and respiratory health. Future research is needed to specifically test in more controlled settings (e.g., interventional trials or animal models) whether and how these specific metabolic pathways mediate the effects of artificial sweeteners on lung function.
Our metabolomic analysis points to systemic inflammatory and oxidative stress pathways, which are established hallmarks of COPD. While observational data cannot establish direct biological mechanisms, findings from experimental models offer a plausible framework to interpret these associations. The biological plausibility of this link is supported by two convergent lines of evidence. First, a hallmark of COPD itself is a state of systemic inflammation, characterized by elevated circulating levels of cytokines such as CRP, IL-6, and TNF-α [40, 41]. Second, and independently, experimental studies suggest that artificial sweeteners may promote a similar pro-inflammatory and pro-oxidative state. For instance, studies in animals and in vitro systems suggest that commonly used artificial sweeteners like aspartame and sucralose can induce the production of key pro-inflammatory cytokines (e.g., TNF-α, IL-6) and oxidative stress [42–45]. At the same time, chronic consumption of artificial sweeteners alters the gut microbiota by reducing its diversity and enriching pro-inflammatory genes [46, 47]. These changes together create a self-reinforcing cycle of systemic inflammation. These alterations extend beyond the gut, correlating with histopathological damage and elevated inflammatory markers in the liver and other organs [48, 49], thereby establishing a pervasive pro-inflammatory milieu. Evidence for acesulfame-K and saccharin is less consistent, but studies in animals show that Ace-K can trigger gut inflammation and increase intestinal permeability [50, 51]. It is crucial to emphasize that these mechanistic insights are derived primarily from controlled, often high-dose, experimental settings. The extent to which these pathways are activated and contribute to disease in humans following chronic, low-level consumption of artificial sweeteners remains to be fully elucidated and represents a key area for future research. While acknowledging the limitations of cross-study comparisons, the convergence of our epidemiological findings with these independent experimental observations may strengthen the biological plausibility of the association we report.
In addition to analyzing the overall association between artificial sweetener consumption and COPD, we also conducted subgroup analyses based on socioeconomic status, familial factors, and environmental exposures, as identifying susceptible sub-populations is crucial for developing targeted strategies to mitigate the disease burden. Our subgroup analysis revealed that smoking status significantly modified the association between artificial sweetener consumption and COPD risk, suggesting a potential biological interaction between these factors. Cigarette smoke imposes a substantial oxidant burden and provokes a persistent inflammatory state in the lungs, potentially rendering them more vulnerable to secondary insults [52, 53]. Smokers tend to exhibit other unhealthy lifestyle behaviors, such as higher intake of sugars and processed foods. Therefore, the elevated risk observed in the smoking population may partially result from the synergistic effects of “unhealthy dietary patterns” and “smoking.” In disadvantaged populations, smoking often coexists with other adverse factors including alcohol drinking, poor dietary habits and others, forming clustered behavioral risk factors that collectively increase the cumulative risk of comorbidities [54]. The consumption of artificial sweeteners may merely serve as a group of such unhealthy dietary patterns. Despite statistical adjustments for known confounders, residual confounding remains a concern. The influence of the overall dietary pattern is difficult to capture precisely, and this effect may be particularly relevant among smokers.
This study has several strengths. First, the prospective design and the large sample size from the UK Biobank provide substantial statistical power for our analyses. The cohort’s detailed baseline characterization, encompassing socioeconomic status, dietary factors, lifestyle, environmental exposures, and medical history, allowed for comprehensive adjustment for a wide array of potential confounders. Second, our study is the first to provide an in-depth examination of the relationships between artificial sweetener intake and COPD incidence and lung function, including a comprehensive analysis of variability across key subgroups. Third, our work breaks new ground by moving beyond established risk factors to interrogate the role of the plasma metabolome. The formal mediation analysis represents a major analytical step, generating preliminary hypotheses regarding specific metabolic shifts that might link sweetener intake to respiratory outcomes. Finally, the consistency of our primary findings across multiple sensitivity checks reinforces the reliability of the observed relationships.
Several limitations must be acknowledged. First, due to the observational nature of this study, a causal relationship cannot be established. Artificial sweetener use clusters with obesity, smoking, diabetes, and lower socioeconomic status. Although we adjusted for these factors, it remains difficult to distinguish a direct causal effect from behavioral marking of an underlying high-risk profile, a concern underscored by stronger associations in smokers and inactive participants. Moreover, reverse causation is possible, as individuals with obesity or metabolic disturbances may selectively consume artificial sweeteners as a sugar substitute, potentially introducing indication bias, which may partly explain the attenuated association in the obese subgroup. To mitigate this concern, sensitivity analyses excluding short follow-up periods (3 and 5 years) yielded consistent results. Nonetheless, residual reverse causality cannot be entirely ruled out. Second, our COPD ascertainment, based on hospital and mortality codes, likely captures more severe disease and may lack specificity without post-bronchodilator spirometry. This limits inferences regarding early disease processes. Third, the mediation analysis is exploratory. The simultaneous testing of numerous, highly correlated metabolites raises concerns regarding multiple testing and collinearity, and the small mediated proportions preclude strong causal inference. These results are hypothesis-generating and require experimental validation. Fourth, measurement bias may exist. Artificial sweetener intake was assessed by self-reported dietary data, which may not be fully accurate. Additionally, selection bias may arise from the exclusion of participants without complete dietary data, potentially limiting generalizability. Compared with the general UK population, UK Biobank participants are more likely to be older, female, and live in more socio-economically advantaged areas, and to be healthier [55]. Therefore, our findings may represent conservative estimates, and generalization to the total UK population and other populations cannot be made. Additional validation through clinical trials or intervention studies is warranted before the results can be applied practically. Finally, COPD diagnosis in the UK Biobank was based on self-report and hospital records, without spirometry confirmation. This may introduce outcome misclassification, particularly for mild or early-stage cases. We were also unable to examine individual sweetener subtypes.
Conclusions
In this population-based large prospective cohort, the results demonstrated that higher consumption of artificial sweeteners is associated with an increased risk of incident COPD and a decline lung function level, with a significant dose–response relationship observed. This association is partially mediated by specific circulating metabolites, particularly those low-and intermediate-density lipoproteins metabolism. The observed relationships were robust across multiple sensitivity analyses. These findings highlight the potential adverse respiratory effects of artificial sweeteners and suggest that reducing their consumption may contribute to the prevention of COPD. Further experimental studies are warranted to elucidate the underlying biological mechanisms and confirm the causal nature of these associations.
Supplementary Information
Acknowledgements
We want to thank all UK Biobank participants and the management team for their participation and assistance.
Author contributions
MF, YC, and XY contributed equally to this work. All authors participated in data collection. MF, YC, and XY conceptualized and designed the research. SZ, JF, QZ, and CJ assisted in data acquisition and interpretation. MF, YC and CJ performed data analysis. MF, YC, and XY drafted the initial manuscript. SZ, JF, QZ, ZL, HB, and LC critically reviewed and revised the manuscript. HB and LC provided study supervision and important intellectual input. All authors read and approved the final manuscript.
Funding
This work was supported by by grants from the National Key Research and Development Program of China (NO. 2022YFF1100604).
Data availability
The datasets analyzed during the current study are available in a public, open-access repository (https://www.ukbiobank.ac.uk). All data used in this study were obtained from the UK Biobank under Application Number 1126547.
Declarations
Ethics approval and consent to participate
All participants provided written informed consent, and the study protocol was approved by the North West Multi-Center Research Ethics Committee in the United Kingdom (11/NW/0382). The study was conducted according to principles of the Declaration of Helsinki. All authors approved the final version of the article, including the authorship list.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Miaomiao Fan, Yan Chen, Xiaolin Yin and Shasha Zhao contributed equally to this work.
Contributor Information
Hao Bai, baihao1991@email.sdu.edu.cn.
Liyong Chen, chenle73@email.sdu.edu.cn.
References
- 1.Vogelmeier CF, Criner GJ, Martínez FJ, Anzueto A, Barnes PJ, Bourbeau J, et al. Global Strategy for the Diagnosis, Management, and Prevention of Chronic Obstructive Lung Disease 2017 report: GOLD executive summary. Arch Bronconeumol. 2017;53:128–49. [DOI] [PubMed] [Google Scholar]
- 2.Adeloye D, Song P, Zhu Y, Campbell H, Sheikh A, Rudan I. Global, regional, and national prevalence of, and risk factors for, chronic obstructive pulmonary disease (COPD) in 2019: a systematic review and modelling analysis. Lancet Respir Med. 2022;10:447–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Safiri S, Carson-Chahhoud K, Noori M, Nejadghaderi SA, Sullman MJM, Ahmadian Heris J, et al. Burden of chronic obstructive pulmonary disease and its attributable risk factors in 204 countries and territories, 1990–2019: results from the Global Burden of Disease Study 2019. BMJ. 2022;378:e069679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Brightling C, Greening N. Airway inflammation in COPD: progress to precision medicine. Eur Respir J. 2019. 10.1183/13993003.00651-2019. [DOI] [PubMed] [Google Scholar]
- 5.Global burden of chronic respiratory diseases and risk factors, 1990–2019: an update from the Global Burden of Disease Study 2019. E Clinical Med 2023, 59:101936. [DOI] [PMC free article] [PubMed]
- 6.Salvi SS, Barnes PJ. Chronic obstructive pulmonary disease in non-smokers. Lancet. 2009;374:733–43. [DOI] [PubMed] [Google Scholar]
- 7.Trends and risk factors of mortality and disability adjusted life years for chronic respiratory diseases from 1990 to 2017: systematic analysis for the Global Burden of Disease Study 2017. BMJ 2020, 370:m3150. [DOI] [PMC free article] [PubMed]
- 8.Schulze MB, Martínez-González MA, Fung TT, Lichtenstein AH, Forouhi NG. Food based dietary patterns and chronic disease prevention. BMJ. 2018;361:k2396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Scoditti E, Massaro M, Garbarino S, Toraldo DM. Role of diet in chronic obstructive pulmonary disease prevention and treatment. Nutrients. 2019. 10.3390/nu11061357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Beslay M, Srour B, Méjean C, Allès B, Fiolet T, Debras C, et al. Ultra-processed food intake in association with BMI change and risk of overweight and obesity: a prospective analysis of the French NutriNet-Santé cohort. PLoS Med. 2020;17:e1003256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Crimarco A, Landry MJ, Gardner CD. Ultra-processed foods, weight gain, and co-morbidity risk. Curr Obes Rep. 2022;11:80–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.He Q, Sun M, Zhao H, Sun N, Han Q, Feng Z, et al. Ultra-processed food consumption, mediating biomarkers, and risk of chronic obstructive pulmonary disease: a prospective cohort study in the UK Biobank. Food Funct. 2023;14:8785–96. [DOI] [PubMed] [Google Scholar]
- 13.Li S, Xiang Y, Yang X, Chen J, Xian W, Wang Y. Associations of sugary beverage consumption with chronic obstructive pulmonary disease, asthma, and asthma-chronic obstructive pulmonary disease overlap syndrome: a prospective cohort study. Am J Clin Nutr. 2024;120:707–18. [DOI] [PubMed] [Google Scholar]
- 14.Riess LE, Huynh BQ, Nachman KE. Aspartame exposures in the US population: demonstration of a novel approach for exposure estimates to food additives using NHANES data. J Expo Sci Environ Epidemiol. 2025;35:351–61. [DOI] [PubMed] [Google Scholar]
- 15.Azad MB, Abou-Setta AM, Chauhan BF, Rabbani R, Lys J, Copstein L, et al. Nonnutritive sweeteners and cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials and prospective cohort studies. CMAJ. 2017;189:E929-e939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Santos NC, de Araujo LM, De Luca Canto G, Guerra ENS, Coelho MS, Borin MF. Metabolic effects of aspartame in adulthood: a systematic review and meta-analysis of randomized clinical trials. Crit Rev Food Sci Nutr. 2018;58:2068–81. [DOI] [PubMed] [Google Scholar]
- 17.Engelen MP, De Castro CL, Rutten EP, Wouters EF, Schols AM, Deutz NE. Enhanced anabolic response to milk protein sip feeding in elderly subjects with COPD is associated with a reduced splanchnic extraction of multiple amino acids. Clin Nutr. 2012;31:616–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rahman I. The role of oxidative stress in the pathogenesis of COPD: implications for therapy. Treat Respir Med. 2005;4:175–200. [DOI] [PubMed] [Google Scholar]
- 19.Navarrete A, Rupérez FJ, Mendes TO, Pérez-Rial S, Girón-Martínez A, Terrón-Expósito R, et al. A metabolomic approach shows sphingosine 1-phosphate and lysophospholipids as mediators of the therapeutic effect of liver growth factor in emphysema. J Pharm Biomed Anal. 2017;139:238–46. [DOI] [PubMed] [Google Scholar]
- 20.Gul SS, Hamilton AR, Munoz AR, Phupitakphol T, Liu W, Hyoju SK, et al. Inhibition of the gut enzyme intestinal alkaline phosphatase may explain how aspartame promotes glucose intolerance and obesity in mice. Appl Physiol Nutr Metab. 2017;42:77–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Griebsch LV, Theiss EL, Janitschke D, Erhardt VKJ, Erhardt T, Haas EC, et al. Aspartame and its metabolites cause oxidative stress and mitochondrial and lipid alterations in SH-SY5Y cells. Nutrients. 2023. 10.3390/nu15061467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lebda MA, Tohamy HG, El-Sayed YS. Long-term soft drink and aspartame intake induces hepatic damage via dysregulation of adipocytokines and alteration of the lipid profile and antioxidant status. Nutr Res. 2017;41:47–55. [DOI] [PubMed] [Google Scholar]
- 23.Liu B, Young H, Crowe FL, Benson VS, Spencer EA, Key TJ, et al. Development and evaluation of the Oxford WebQ, a low-cost, web-based method for assessment of previous 24 h dietary intakes in large-scale prospective studies. Public Health Nutr. 2011;14:1998–2005. [DOI] [PubMed] [Google Scholar]
- 24.Wang T, Liang M, Ye X, Huang X, Chen H, He X, et al. Ambient air pollution exposure, mediating biomarkers and risk of COPD: a cohort study and meta-analysis. Eur Respir Rev. 2025. 10.1183/16000617.0055-2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Said MA, Verweij N, van der Harst P. Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK Biobank study. JAMA Cardiol. 2018;3:693–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett DR Jr., Tudor-Locke C, et al. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43:1575–81. [DOI] [PubMed] [Google Scholar]
- 27.Würtz P, Kangas AJ, Soininen P, Lawlor DA, Davey Smith G, Ala-Korpela M. Quantitative serum nuclear magnetic resonance metabolomics in large-scale epidemiology: a primer on -omic technologies. Am J Epidemiol. 2017;186:1084–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bragg F, Trichia E, Aguilar-Ramirez D, Bešević J, Lewington S, Emberson J. Predictive value of circulating NMR metabolic biomarkers for type 2 diabetes risk in the UK Biobank study. BMC Med. 2022;20:159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Huang YT, Yang HI. Causal mediation analysis of survival outcome with multiple mediators. Epidemiology. 2017;28:370–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Zhuo B, Ran S, Qian AM, Zhang J, Tabet M, Howard SW, et al. Air pollution metabolomic signatures and chronic respiratory diseases risk: a longitudinal study. Chest. 2024;166:975–86. [DOI] [PubMed] [Google Scholar]
- 31.Salehi Z, Malmir H, Ghosn B, Onvani S, Ardestani ME, Feizi A, et al. Exploring the association between ultra-processed foods and COPD: a case-control study. BMC Pulm Med. 2024;24:124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.DeChristopher LR, Uribarri J, Tucker KL. Intake of high fructose corn syrup sweetened soft drinks is associated with prevalent chronic bronchitis in U.S. adults, ages 20-55 y. Nutr J. 2015;14:107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Shi Z, Dal Grande E, Taylor AW, Gill TK, Adams R, Wittert GA. Association between soft drink consumption and asthma and chronic obstructive pulmonary disease among adults in Australia. Respirology. 2012;17:363–9. [DOI] [PubMed] [Google Scholar]
- 34.Berentzen NE, van Stokkom VL, Gehring U, Koppelman GH, Schaap LA, Smit HA, et al. Associations of sugar-containing beverages with asthma prevalence in 11-year-old children: the PIAMA birth cohort. Eur J Clin Nutr. 2015;69:303–8. [DOI] [PubMed] [Google Scholar]
- 35.Maslova E, Strøm M, Olsen SF, Halldorsson TI. Consumption of artificially-sweetened soft drinks in pregnancy and risk of child asthma and allergic rhinitis. PLoS ONE. 2013;8:e57261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ibrahim AA, Qamar B, Fituri S, Akbar ZA, Al-Abdi T, Shi Z. Association between soft drink consumption and asthma among Qatari adults. Nutrients. 2019. 10.3390/nu11030606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Roversi S, Fabbri LM, Sin DD, Hawkins NM, Agustí A. Chronic obstructive pulmonary disease and cardiac diseases. An urgent need for integrated care. Am J Respir Crit Care Med. 2016;194:1319–36. [DOI] [PubMed] [Google Scholar]
- 38.Vanfleteren LE, Spruit MA, Groenen MT, Bruijnzeel PL, Taib Z, Rutten EP, et al. Arterial stiffness in patients with COPD: the role of systemic inflammation and the effects of pulmonary rehabilitation. Eur Respir J. 2014;43:1306–15. [DOI] [PubMed] [Google Scholar]
- 39.Vivodtzev I, Tamisier R, Baguet JP, Borel JC, Levy P, Pépin JL. Arterial stiffness in COPD. Chest. 2014;145:861–75. [DOI] [PubMed] [Google Scholar]
- 40.Barnes PJ. Inflammatory mechanisms in patients with chronic obstructive pulmonary disease. J Allergy Clin Immunol. 2016;138:16–27. [DOI] [PubMed] [Google Scholar]
- 41.Thomsen M, Dahl M, Lange P, Vestbo J, Nordestgaard BG. Inflammatory biomarkers and comorbidities in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2012;186:982–8. [DOI] [PubMed] [Google Scholar]
- 42.Lawal AO, Agboola OO, Akinjiyan MO, Ijatuyi TT, Dahunsi DT, Okeowo OM, et al. The antioxidative, anti-inflammatory and anti-apoptotic effects of Tetrapleura tetraptera (Aidan) ethanol leaf extract in the brain of Wistar rats exposed to aspartame. Mol Neurobiol. 2025;62:9430–48. [DOI] [PubMed] [Google Scholar]
- 43.Guo M, Liu X, Tan Y, Kang F, Zhu X, Fan X, et al. Sucralose enhances the susceptibility to dextran sulfate sodium (DSS) induced colitis in mice with changes in gut microbiota. Food Funct. 2021;12:9380–90. [DOI] [PubMed] [Google Scholar]
- 44.Lin CH, Li HY, Wang SH, Chen YH, Chen YC, Wu HT. Consumption of non-nutritive sweetener, Acesulfame Potassium exacerbates atherosclerosis through dysregulation of lipid metabolism in ApoE(-/-) mice. Nutrients. 2021. 10.3390/nu13113984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Li X, Liu Y, Wang Y, Li X, Liu X, Guo M, et al. Sucralose promotes colitis-associated colorectal cancer risk in a murine model along with changes in microbiota. Front Oncol. 2020;10:710. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Bian X, Tu P, Chi L, Gao B, Ru H, Lu K. Saccharin induced liver inflammation in mice by altering the gut microbiota and its metabolic functions. Food Chem Toxicol. 2017;107:530–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Martínez-Carrillo BE, Rosales-Gómez CA, Ramírez-Durán N, Reséndiz-Albor AA, Escoto-Herrera JA, Mondragón-Velásquez T, et al. Effect of chronic consumption of sweeteners on microbiota and immunity in the small intestine of young mice. Int J Food Sci Technol. 2019;2019:9619020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Bian X, Chi L, Gao B, Tu P, Ru H, Lu K. Gut microbiome response to sucralose and its potential role in inducing liver inflammation in mice. Front Physiol. 2017;8:487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Dai X, Wang C, Guo Z, Li Y, Liu T, Jin G, et al. Maternal sucralose exposure induces Paneth cell defects and exacerbates gut dysbiosis of progeny mice. Food Funct. 2021;12:12634–46. [DOI] [PubMed] [Google Scholar]
- 50.Hanawa Y, Higashiyama M, Kurihara C, Tanemoto R, Ito S, Mizoguchi A, et al. Acesulfame potassium induces dysbiosis and intestinal injury with enhanced lymphocyte migration to intestinal mucosa. J Gastroenterol Hepatol. 2021;36:3140–8. [DOI] [PubMed] [Google Scholar]
- 51.Zhai Z, Zhang Y, Liang X, Li J, Chen Z, Zhang J, et al. Acesulfame potassium triggers inflammatory bowel disease via the inhibition of focal adhesion pathway. J Hazard Mater. 2024;476:134901. [DOI] [PubMed] [Google Scholar]
- 52.Johannsen A, Susin C, Gustafsson A. Smoking and inflammation: evidence for a synergistic role in chronic disease. Periodontol 2000. 2014;64:111–26. [DOI] [PubMed] [Google Scholar]
- 53.Caliri AW, Tommasi S, Besaratinia A. Relationships among smoking, oxidative stress, inflammation, macromolecular damage, and cancer. Mutat Res Rev Mutat Res. 2021;787:108365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Meader N, King K, Moe-Byrne T, Wright K, Graham H, Petticrew M, et al. A systematic review on the clustering and co-occurrence of multiple risk behaviours. BMC Public Health. 2016;16:657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Fry A, Littlejohns TJ, Sudlow C, Doherty N, Adamska L, Sprosen T, et al. Comparison of sociodemographic and health-related characteristics of UK Biobank participants with those of the general population. Am J Epidemiol. 2017;186:1026–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets analyzed during the current study are available in a public, open-access repository (https://www.ukbiobank.ac.uk). All data used in this study were obtained from the UK Biobank under Application Number 1126547.