By Mark Gold, MD
Scientists and the general public have been aware of the harmful effects of tobacco for years now, recognizing its links to cardiovascular problems, cancer, and an array of other serious health problems. The Centers for Disease Control states that cigarette smoking accounts for over 480,000 deaths every year in the United States, and seven million deaths worldwide annually, which makes smoking the leading cause of preventable death in the world.1 Of those 480,000 deaths per year in the United States, the CDC includes more than 41,000 deaths resulting from secondhand smoke exposure. Secondhand smoke can cause or worsen a wide range of damaging health effects in children and adults, including lung cancer, respiratory infections and asthma. Scientists have concluded that the only real difference between first and secondhand smoke is consent. Children and others exposed to secondhand smoke suffer the same consequences but did not intend to smoke at all.2 In toto, cigarette smoking results in one in five deaths annually, or 1,300 deaths every day. On average, smokers die 10 years earlier than nonsmokers.
Unfortunately for individuals who are smoking or exposed to cigarettes, new research indicates that tobacco use contributes to yet another health problem: biological aging. Previously, studies on the effects of tobacco use on aging have been fairly limited in their conclusions, but in recent years researchers have focused more on the question, applying new scientific tools in their work. One recent study, by Mamoshina et. al., used artificial intelligence to analyze blood and cell counts of smokers and non-smokers and to measure how much tobacco use aged smokers.
What did this study find about tobacco use and aging?
This study found that tobacco use has a significant effect on the acceleration of biological age in smokers. Biological age refers to how old an individual is in terms of his or her health, distinct from chronological age, the number of years an individual has lived. In this study, the researchers ran data on blood biochemistry and cell count results through an algorithm trained to measure age. The administrative data in the study looked at the records of 149,000 individuals, 49,000 of whom were smokers. Their information was compiled as a representative sample of Alberta, Canada, reflecting proportionate portions of Alberta’s rural, urban, and ethnic populations. Each group in the study also had similar numbers of males and females and a median age of 55. The researchers created age-prediction models through supervised “deep neural networks,” a kind of artificial intelligence that applies complex math models.
The study found that smokers showed higher biological aging rates than nonsmokers, independent of other factors like fasting glucose levels or cholesterol. Its models also show that, compared to nonsmokers, female smokers were twice their chronological age. The models predicted that male smokers were one and a half times their chronological age. They also predicted that smokers in the study between the ages of 31-40 were biologically older: 44% were between the ages of 41-50, and 26% between 51-60.
The study indicated advanced aging in smokers through 55 years of age. After 55, the researchers note, aging effects curiously vanish. The study suggests different possible causes for diminished effects in older smokers. It could be possible, for example, that tobacco initiates bodily repair mechanisms, which some experts have suggested is the case for smoking and the body’s defense against Parkinson’s. But it could also be true that the effects of aging, in the long run, are random and psychologically taxing in ways that produce widely divergent individual outcomes, or that many smokers die earlier and aren’t accounted for in studies of older individuals.
Notably, this study is also consistent with recent approaches to aging. Mamoshina et. al’s work here reflects the researchers’ previous studies on how the advanced technology they use beats chronological age in predicting all-cause mortality3, and how it can generally track age-related alterations.4 Other researchers have found the effects of smoking on aging, including one study that used DNA methylation clocks,5 but this study is the first using a deep learning-based approach to measure tobacco use and aging.
Why are these findings important?
We often assume an intuitive connection between smoking and accelerated aging, given noticeable changes in smokers’ appearance, speech, mobility, and a number of other easily spotted alterations in observed behavior. And we’ve known for decades that smoking is highly damaging to individual health across a range of indicators. Smokers' cough or excessive skin wrinkles on the face or discolorations on the fingers are easily detected without genomic or other sophisticated testing. But having scientific research confirm the age-accelerating effects of tobacco use is still very important. Advanced biological aging isn’t just “one more” health problem caused by smoking. As the study notes, the populations of many countries are aging as part of a global demographic trend with a number of underlying causes. This increases the number of individuals requiring medical care overall and, because the elderly often have more severe health problems, the number of individuals requiring more intense care. As a result, aging puts considerable stress on available health care resources, and aging caused by tobacco use only exacerbates these bourgeoning pressures. In other words, instead of representing just “one more” problem caused by smoking, tobacco-related advanced aging is a problem of growing urgency for many societies.
Mamoshina et. al’s study also has broader relevance for health care practitioners because information that relies on individual self-reporting of smoking status is unreliable. We have found that as well in asking about smoking in the home or during pregnancy.
It is not clear whether advanced aging is solely due to smoking cigarettes or exposure to cigarette smoke, or something found with many drugs of abuse. Volkow and her colleagues reviewed the literature on pathophysiological processes that may hasten aging and are particularly relevant to smoking, second hand smoke, drug use, and addictions.6 They focused on oxidative stress and cellular aging, inflammation in periphery and brain, decline in brain volume and function, and early onset of cardiac, cerebrovascular, kidney, and liver disease. Drug use may trigger early onset of age-related disease, due to drug-induced multi-system toxicity and perilous lifestyle, which remains mostly undetected and untreated. Learning more about biological aging effects should help us, over time, develop better preventative approaches and early detection practices and treatments, but as for now some evidence on aging is very clear: smokers have another disturbingly pressing reason to quit.
National Center for Chronic Disease Control and Prevention. (2019) Smoking & Tobacco Use: Fast Facts, Diseases and Death. Centers for Disease Control and Prevention
Merlo L.J., Sutton J.A., Gold M.S. (2014) Brief educational intervention to improve medical student competence in managing patients exposed to secondhand smoke. Substance Abuse
Mamoshina, P. et al. (2018) Population-specifc biomarkers of human aging: a big data study using South Korean, Canadian and Eastern European patient populations. J. Gerontol
Putin, E. et al. (2016) Deep biomarkers of human aging: Application of deep neural networks to biomarker development. Aging
Lei, M.-K., Beach, S. R. H., Dogan, M. V. & Philibert, R. A. (2017) A pilot investigation of the impact of smoking cessation on biological age. The American Journal on Addictions
Bachi K., Sierra S., Volkow N.D., Goldstein R.Z., Alia-Klein N. (February 2017) Is biological aging accelerated in drug addiction? Current Opinion in Behavioral Sciences