Smoking remains the leading cause of preventable morbidity and mortality worldwide, contributing to more than 8 million deaths each year.1 Globally, in 2019, 1.14 billion individuals were current smokers who consumed 7.41 trillion cigarette-equivalents of tobacco. Tobacco use accounted for 200 million disability-adjusted life-years and was the leading risk factor for death among men.2 In 2012, the total global costs of tobacco smoking were estimated to be approximately USD 1432 billion, which represents 1.8% of the global annual GDP.3 Smoking is widely recognized to be associated with various health risks, including cancer, cardiovascular disease, and chronic obstructive pulmonary disease.4 Despite this knowledge, smoking rates remain high in certain nations.5

Metabolic diseases belong to a class of diseases caused by abnormal metabolic processes in the body, including—but not limited to—diabetes, obesity, and hyperlipidemia. Previous studies suggest that smoking itself may trigger metabolic syndrome (MetS).6,7 Substances contained in tobacco have adverse effects on the sugar metabolism and lipid metabolism of smokers, as they activate the human sympathetic nervous system, increase the level of glucocorticoids in the blood, and increase the level of growth hormone. Increased plasma cortisol concentration and insulin resistance can, in turn, lead to visceral fat accumulation and an increase in waist circumference.8,9 Smoking is also an independent risk factor for low bone mineral density.10 In addition, the harmful components contained in smoking products cause direct toxicity to organs, glands, and cells, further affecting systemic metabolism.

Previous studies have suggested that smoking has different effects on different aspects of the human metabolism. In this study, we combined the metabolic features of glucose metabolism, lipid metabolism, bone metabolism, and nucleic acid metabolism to classify the overall metabolic state. Using this variable, we aimed to study the effect of smoking on the overall metabolic networks and pathways in the human body.

Substances contained in tobacco have adverse effects on nodes of metabolism by activating the sympathetic nervous system. In addition to hormone disorders, the harmful components in smoke are directly toxic to organs, glands, and vascular endothelium, further affecting systemic metabolism.8,9 There is a close relationship between smoking and MetS. Smoking can not only increase blood pressure, triglyceride level, and abdominal obesity but can also increase the risk of type 2 diabetes through direct toxicity to islet and insulin resistance.11,12 Considering obesity, smoking may have a two-way effect. After individuals start smoking, appetite is reduced in the short term, and body metabolic rate is increased; however, long-term smoking may increase appetite and reduce body metabolic rate. Moreover, smoking may also promote the formation of abdominal obesity by increasing insulin resistance and promoting glucocorticoid secretion.13

Smoking has been associated with increased serum total cholesterol, triglyceride, and low-density lipoprotein cholesterol. High-density lipoprotein cholesterol has also been associated with smoking.14 Nicotine stimulates the adrenal cortex to release epinephrine, leading to the increase in serum free fatty acid concentrations observed in smokers.15 Free fatty acids are very low-density lipoproteins that promote hepatic secretion to produce triglycerides. The concentration of HDL in serum changes inversely with the concentration of very low-density lipoprotein. Free fatty acids also stimulate the synthesis and secretion of cholesterol by the liver. Smoking can seriously impair several steps of reverse cholesterol transport (RCT)16 and further aggravate low-density dyslipidemia. There was a dose-response relationship between smoking and the risk of type 2 diabetes.17 Insulin resistance develops in long-term smokers and people who use nicotine gum.12 Chemical components of tobacco smoke have direct toxic effects on pancreatic and β-cell function in both fetuses and adults.18 Some studies have found that uric acid levels are lower in smokers than in non-smokers, which may be due to the inhibition of xanthine oxidase by cyanide in cigarette smoke.19 In contrast, the elevated uric acid and gout-triggering effects of smoking may be the result of smoke-induced inflammation, or it may be the result of cellular uric acid leakage due to smoke-induced cellular damage. In the end, the overall health benefits of quitting smoking outweigh any possible protection against gout by continuing to smoke.20

Hyperhomocysteinemia is associated with occlusive arterial disease—especially in the brain, heart, and kidneys—as well as Alzheimer's disease, venous thrombosis, cancer, chronic renal failure, depression, miscarriage, Parkinson's disease, stroke, and pre-eclampsia. Although there is a clear association between smoking and hyperhomocysteinemia, the mechanism is unclear.21 Smoking is an independent risk factor for bone mineral density loss, and severe bone density loss increases the risk of fractures in smokers. Smoking may affect bone metabolism through indirect pathways such as influencing body weight and hormone levels—including reducing blood 25-hydroxyvitamin D and estrogen levels and increasing glucocorticoid levels. It can also induce osteoporosis by direct action on the skeletal system.22,23

In many patients, the adverse metabolic conditions occur together, further increasing morbidity and mortality risks for the individual.24 The metabolic pathways of glucose metabolism, lipid metabolism, nucleic acid metabolism, and bone metabolism that we have mentioned are more like a huge network that provides material support for the functioning of the human body. They reach a certain balance in their inextricable interconnectedness. In this study, the clustering results of the metabolic statuses of individuals with different smoking statuses were very interesting. We clustered the participants into three groups based on their overall metabolic profile. In the excellent metabolism group, the manifestations of glucose metabolism, lipid metabolism, body weight, and bone metabolism were in a synergetic good state. In this group, abdominal circumference (an indicator of central obesity), fasting insulin (an indicator of insulin resistance), and blood calcium (an indicator of osteoclast activity) were the lowest among the three groups. Serum uric acid was also much lower than the Intermediate metabolism group. In the adverse metabolism group, all metabolic measures except uric acid and 25OHD had the worst performance. All metabolic measures except uric acid and 25OHD in the intermediate metabolism group were in the intermediate state between excellent and adverse metabolism. This indicates the possible existence of a synergistic network relationship between the metabolism of different substances. Among the different metabolites, only uric acid differed from the others. The uric acid-lowering effect of smoking may be involved in this phenomenon.

In this study, non-smokers had the highest proportion of excellent metabolizers, and current smokers had the highest proportion of adverse metabolizers. The proportion of adverse metabolizers in the ex-smoker group was clinically relevantly lower than that of smokers. The difference in the distribution of smokers in different metabolic clusters indicated that smoking had an upstream intervention on the whole metabolic network. This may be due to its effects on the levels of multiple hormones; moreover, mitochondrial dysfunction, oxidative stress, and inflammation contribute to the direct toxic effects caused by smoking.8,9,18

This study had several limitations. First, selection bias may have occurred because the data were collected from patients who visited the clinic within a certain period of time and only male participants older than 50 years were included. Second, it is important to consider that a substantial proportion of women's metabolism could be affected by passive smoking, but this study was not able to observe this aspect. The main strength of this study is the exploratory approach of integrating multiple metabolic markers by cluster methodology, which organically reflects whole-body metabolic characteristics. We look forward to further validation and exploration of the metabolic network through this method.

Although the prevalence of smoking has decreased significantly since 1990, among both men and women, population growth has led to a significant increase in the total number of smokers.2 Cardiovascular diseases, cancers, and chronic respiratory diseases were the three leading causes of smoking-attributable age-standardized deaths and disability-adjusted life years (DALYs).25 Metabolic diseases such as obesity and type 2 diabetes among older adults pose a major challenge to global public health.26 From 2011 to 2016, the weighted MetS prevalence was up to 34.7% in the United States.27,28 If the population with poor nucleic acid metabolism and poor bone metabolism were added, the total proportion would be even greater. The huge smoking population increases the risk of metabolic disorders, which will inevitably increase the burden of metabolic diseases worldwide. Smoking-related metabolic diseases would also be another notable cause of smoking-attributable disease burden globally.

This study was approved by the ethics committee of Peking Union Medical College Hospital (I-23PJ234). The study was conducted in accordance with the Declaration of Helsinki.

The institutional review board waived the requirement for informed consent as the data were anonymous and because of the retrospective nature of the study.

The data presented in this study are available on request from the corresponding author.

This research was supported by the National High Level Hospital Clinical Research Funding of China [2022-PUMCH-B-133].

The authors declare no conflict of interest.