Evaluating the Evidence on Sitting, Smoking, and Health: Is Sitting Really the New Smoking?

Abstract

Sitting has frequently been equated with smoking, with some sources even suggesting that smoking is safer than sitting. This commentary highlights how sitting and smoking are not comparable.

The most recent meta-analysis of sedentary behavior and health outcomes reported a hazard ratio of 1.22 (95% confidence interval [CI] = 1.09, 1.41) for all-cause mortality. The relative risk (RR) of death from all causes among current smokers, compared with those who have never smoked, is 2.80 (95% CI = 2.72, 2.88) for men and 2.76 for women (95% CI = 2.69, 2.84). The risk is substantially higher for heavy smokers (> 40 cigarettes per day: RR = 4.08 [95% CI = 3.68, 4.52] for men, and 4.41 [95% CI = 3.70, 5.25] for women). These estimates correspond to absolute risk differences of more than 2000 excess deaths from any cause per 100 000 persons per year among the heaviest smokers compared with never smokers, versus 190 excess deaths per 100 000 persons per year when comparing people with the highest volume of sitting with the lowest.

Conflicting or distorted information about health risks related to behavioral choices and environmental exposures can lead to confusion and public doubt with respect to health recommendations.

Sedentary behavior is any waking behavior characterized by an energy expenditure less than or equal to 1.5 metabolic equivalents (METs), while in a sitting, reclining, or lying posture.¹ Sedentary behaviors are pervasive, require minimal effort, and are accumulated throughout the week (particularly evenings and weekends) and across multiple domains (e.g., leisure, occupation, transport).² According to studies that used device-based measures of sedentary behaviors, adults typically spend 9 hours per day sitting. Older adults are sedentary, on average, 10 hours per day.³

Over the past decade, media coverage of sitting research has been widespread and the health consequences of sitting have often been compared with those of smoking (i.e., “sitting is the new smoking”). A recent analysis of news articles found almost 300 articles claiming that sitting is the new smoking.⁴ Here, we provide a brief perspective on recent findings on the relationships of sitting with health outcomes. We compare the sitting-related risk estimates and absolute risk differences for all-cause mortality and prevalent chronic diseases (incidence and mortality) to smoking-related risks, and argue that promulgating direct comparisons of the health consequences of sitting and smoking is not recommended.

THE EVIDENCE ON SITTING AND HEALTH

We do not discount the risks associated with sitting. Compared with lower volumes of sitting (e.g., < 4 hours/day), high volumes of sitting (e.g., > 8 hours/day) have been found to be associated with adverse health outcomes.⁵ The strongest risk estimates have been observed for type 2 diabetes, with the most recent meta-analysis reporting a hazard ratio of 1.91 (95% confidence interval [CI] = 1.64, 2.22).⁵ Associations with all-cause mortality, cancer mortality, and cardiovascular disease (CVD) mortality are smaller, with hazard ratios of 1.22 (95% CI = 1.08, 1.38) for all-cause mortality, 1.13 (95% CI = 1.06, 1.21) for all cancer mortality, and 1.15 (95% CI = 1.07, 1.24) for CVD mortality.⁵ A more recent meta-analysis reported an association between sitting time and CVD incidence, with a hazard ratio of 1.14 (95% CI = 1.09, 1.18).⁶ Stated differently, it appears that excessive sitting time almost doubles the risk of type 2 diabetes, but only increases incidence and mortality risk associated with other common chronic diseases by approximately 10% to 20%. In terms of absolute risk difference, these relative risk (RR) estimates correspond to an excess of around 33 CVD-related deaths, 27 cancer-related deaths, and 610 incident cases of diabetes per 100 000 persons per year in people with the highest volumes of sitting compared with those with the lowest volumes of sitting. Sitting has also been found to be adversely associated with the risk of depression (RR = 1.14; 95% CI = 1.06, 1.21)⁷ and physical health-related quality of life domains (average effect size = −0.15; 95% CI = −0.21, −0.10); however, negligible associations for mental and social health–related quality-of-life domains were reported.⁸

Studies have examined whether the health effects of sitting can be attenuated by physical activity. The largest study was a harmonized pooled analysis of 1 005 791 individuals.⁹ The results showed significant associations of daily sitting time and television viewing time with all-cause mortality in adults with low or moderate physical activity levels. A high volume of moderate–vigorous intensity activity (60–75 minutes per day) was necessary to fully attenuate all-cause mortality risk associated with sitting time (but failed to completely attenuate risks associated with television viewing time). Given that a large proportion of the population fails to attain 30 minutes of physical activity a day, 60 to 75 minutes per day may be unattainable for most.

There may be benefits to time spent sitting, including rest, socialization, and entertainment. Indeed, some research suggests harmful consequences to standing for prolonged periods of time.¹⁰ In relation to public health messaging on sitting, Hamilton et al.¹¹ suggested caution in warning against excessive exposure to the hazard, much like sun exposure, in which some exposure may be beneficial (e.g., stimulating vitamin D synthesis) but too much may result in hazardous health consequences (e.g., skin cancer). A meta-analysis found that the association between sitting and mortality may not be linear; the dose–response risk estimates heightened from 7 hours of sitting per day onward.¹² Overall, there is insufficient empirical evidence to make specific recommendations in relation to how much sitting time is hazardous.

IS SITTING REALLY THE NEW SMOKING?

When the potential health risk is one that is ubiquitous in society, such as sitting, media attention is often intensified and has indeed been widespread over the past decade.¹³ Media headlines around sitting have often equated sitting with smoking, with some sources even suggesting smoking is safer than sitting.¹⁴ Such headlines began appearing around 2010 and have been cited in high-profile outlets including the Los Angeles Times¹⁵ and Time magazine.¹⁶ It was recently estimated that news stories with “new smoking” or “as bad as smoking” comparisons increased 12-fold from 2012 to 2016.⁴ Similar headlines are permeating the academic context^17–21 and are being promoted by some respected clinical institutions.²²

News coverage of health risks influences public knowledge and perceptions of such risks,²³ and can influence individual behavioral choices and public policy.²⁴ Conflicting, conflated, and distorted information regarding health benefits and risks can lead to confusion and public doubt with respect to public health recommendations.²³ Such misinformation has consequences. For example, with respect to sitting, dialogue has emerged suggesting employers be held liable if indeed sitting is the new smoking.²⁵ The mass-media enthusiasm for condemning sitting by making comparisons to smoking has far outpaced the available scientific evidence. It is obvious from examination of smoking research that sitting and smoking are distinct behaviors with different levels of associated risk (Table 1).

TABLE 1—

Associations for Smoking and Sitting in Relation to Health Outcomes, From Most Recent Published Meta-analyses or Pooled Analysis

	Current vs Never Smoker		Highest vs Lowest Category of Smoking		Highest vs Lowest Category of Sitting		PAF
Outcome	RR (95% CI)	Absolute Risk Difference (per 100 000 Persons per Year)	RR (95% CI)	Absolute Risk Difference (per 100 000 Persons per Year)	RR (95% CI)	Absolute Risk Difference (per 100 000 Persons per Year)	% Attributable to Sitting	% Attributable to Smoking
All-cause mortality							3.826 to 6.927 (both sexes)
Men	2.80 (2.72, 2.88)a	1554 deaths	4.08 (3.68, 4.52)a	2659 deaths	1.22 (1.08, 1.38)b	190 deaths		2129
Women	2.76 (2.69, 2.84)a	1099 deaths	4.41 (3.70, 5.25)a	2129 deaths	1.22 (1.08, 1.38)b	137 deaths		1729
CVD mortality	2.07 (1.82, 2.36)c	238 deaths	2.63 (2.28, 3.04)c	363 deaths	1.15 (1.07, 1.24)b	33 deaths	N/A	40.931 (men); 11.931 (women)
All cancer mortality	2.19 (1.83, 2.63)d	192 deaths	2.88 (2.36, 3.50)d	303 deaths	1.13 (1.06, 1.21)b	21 deaths	N/A	28.833
Lung cancer	13.10 (9.90, 17.3)d	643 cases	21.70 (17.10, 27.40)d	1099 cases	1.17 (1.06, 1.52)e	9 cases	N/A	81.733
Type 2 diabetes	1.37 (1.33, 1.42)f	248 cases	1.57 (1.47, 1.66)f	382 cases	1.91 (1.64, 2.22)b	610 cases	N/A	2236

Notes. CI = confidence interval; CVD = cardiovascular disease; N/A = not available; PAF = population attributable fraction; RR = risk ratio. Absolute risk differences were calculated by multiplying the relevant background mortality or incidence rate in the United States by the relevant risk estimate minus 1. The background rates for determining absolute risk difference were based on the following references: all-cause mortality,³⁷ CVD mortality,³⁸ all-cancer mortality,³⁹ lung cancer incidence,⁴⁰ and diabetes.⁴¹

^aPooled data analysis, five prospective cohort studies; n = 2 222 223 participants.²⁸

^bSystematic review and meta-analysis⁵—all-cause mortality: 14 studies [n = 829 917 participants]; CVD mortality: 7 studies [n = 551 366 participants]; cancer, 8 studies [n = 744 706 participants]; type 2 diabetes: 5 studies [n = 26 700 participants].

^cMeta-analysis, 25 prospective cohort studies; n = 503 905 participants.³⁰

^dMeta-analysis, 19 prospective cohort studies; n = 897 021 participants.³²

^eMeta-analysis, 17 prospective cohort studies; n = 857 581 participants.³⁴

^fSystematic review and meta-analysis, 88 prospective cohort studies; n = 5 898 795 participants.³⁵

EVIDENCE ON SMOKING AND HEALTH

Smoking has been defined as one of the greatest public health disasters of the 20th century.⁴² The 2014 US Surgeon General’s Report on smoking summarizes thousands of studies linking smoking to a host of adverse health consequences including CVD, chronic obstructive pulmonary disease, dementia and Alzheimer’s disease, respiratory diseases (e.g., sleep apnea, asthma exacerbations), adverse reproductive outcomes, and cancers of at least 12 sites.⁴² Smoking will cause one billion deaths in the 21st century.⁴³ The annual global cost of smoking-attributable disease was estimated at $467 billion in 2012.⁴⁴ It was estimated that tobacco companies spent $8.7 billion in 2016 marketing cigarettes in the United States alone.⁴⁵ Estimates have suggested that the cost of physical inactivity (not achieving 150 minutes of at least moderate-intensity physical activity per week) for international health care systems was $53.8 billion in 2013.⁴⁶ The economic burden of sitting is unknown and the relevant interest groups and amount of money spent are not comparable.

Risk estimates and absolute risk differences for smoking far outweigh those for sitting (Table 1), except for type 2 diabetes. For example, the RR for all-cause mortality among smokers who smoked more than 40 cigarettes per day was 4.08 (95% CI = 3.68, 4.52) for men, and 4.41 (95% CI = 3.70, 5.25) for women.²⁸ These RR estimates correspond to absolute risk differences of more than 2000 excess deaths from any cause per 100 000 persons per year among the heaviest smokers compared with never smokers. The corresponding absolute risk difference for the highest volume of sitting versus the lowest volumes is 190 excess deaths per 100 000 persons per year. Even when all smokers are collapsed into one group (i.e., current smokers), risk estimates far exceed those for sitting. For example, the RR of death from all causes among current smokers compared with those who have never smoked was 2.80 (95% CI = 2.72, 2.88) for men (absolute risk difference of 1554 excess deaths per 100 000 persons per year) and 2.76 for women (95% CI = 2.69, 2.84; absolute risk difference of 1099 excess deaths per 100 000 persons per year),²⁸ compared with 1.22 (95% CI = 1.09, 1.41) for sitting.⁵ Thus, any level of smoking increases risk of dying from any cause by approximately 180% versus a 25% risk increase for sitting. Even light smoking (1–4 cigarettes per day) has been associated with a higher risk of mortality compared with sitting.⁴⁷ Smoking also increases the risk of other health outcomes including depression (odds ratio = 1.62; 95% CI = 0.10, 2.40)⁴⁸ and poor quality of life (standardized mean difference = 0.22; 95% CI = 0.09, 0.36).⁴⁹ Both the magnitude of these associations and the strength of the evidence are greater than those observed for sitting time to date.

Studies have calculated the population attributable fraction for sitting-related all-cause mortality. In a meta-analysis that included almost 600 000 adults, it was estimated that 5.9% of deaths could be attributed to total daily sitting time.¹² In a larger study of 54 countries (approximately 1.2 million people), sitting time was responsible for 3.8% of all-cause mortality.²⁶ Eliminating sitting more than three hours per day increased average life expectancy by 0.23 years (0.13 years in the United States). With respect to smoking, it has been estimated that 21% of deaths among men and 17% among women were attributable to smoking.²⁹ An analysis of the World Health Organization Mortality Database concluded that eliminating smoking would increase life expectancy by 2.4 years for men and 1 year for women (2.9 years for women in the United States).⁵⁰

SECONDHAND EXPOSURE

Unlike sitting, smoking has serious physical health consequences for others. In the United States alone, approximately 2.5 million nonsmokers died from secondhand smoke–related problems between the years 1964 and 2014.⁴² Despite advances in legislation regarding smoke-free spaces, it is difficult to avoid exposure to secondhand and thirdhand smoke. An individual can most often decide to sit, stand, or move. However, an individual often cannot simply choose to avoid second- or thirdhand smoke. One recent meta-analysis found a disproportionately high risk of stroke even at low levels of secondhand smoke exposure.⁵¹ For pregnant smokers, nicotine presents serious health risks to the developing fetus, including poor neurodevelopment, cell damage, and impaired synaptic activity.⁵² Sitting has no comparable capacity to affect the health of others. Furthermore, there is no research to suggest that an individual’s sedentary behaviors provide harmful and unavoidable health consequences for another individual.

ADDICTION VS HABIT

Caution is needed when one is comparing an addictive behavior (smoking) to a habit (sitting). Addiction is defined as “an inability to consistently abstain, impairment in behavioural control, craving, diminished recognition of significant problems with one’s behaviours and interpersonal relationships, and a dysfunctional emotional response.”⁵³ Cigarette smoke contains nicotine, a highly addictive chemical that creates a physiologic dependency, which can create of a spectrum of withdrawal symptoms.^54,55

Whereas nicotine addiction is a direct result of smoking, sitting, as is the case for other behaviors (such as eating),⁵⁶ is not an addiction. Sitting is consistent with the definition of habit. Habits are actions or behaviors triggered automatically in response to contextual cues.⁵⁷ Habits emerge from the gradual learning of associations between responses and the features of performance contexts.⁵⁸ It has been suggested that sitting is habitual because individuals develop associations between sitting and common activities or contexts in their daily lives (e.g., at work, watching television).⁵⁹ Arguing that an addiction to sitting exists is implausible given the nature of addiction. It seems improbable that such neural pathways for reinforcement of a behavior are triggered with sitting. Furthermore, studies suggest that environmental cues to smoke (e.g., pictures of cigarettes, being around a smoker) lead to stronger cravings and shorter latency times (i.e., time leading up to actually smoking a cigarette).⁶⁰ At the very least, to support the notion that sitting is addictive, evidence must show that sitting elicits the signs and symptoms of addiction including irritability, cravings, depressed mood, anhedonia, restlessness, and anxiety.⁶¹

CONCLUSIONS

It has been suggested the mass media is the most important and significant source of health information for the general public.⁶² News coverage of any health risk influences the knowledge and perceptions not only of the general public but also health care professionals.²³ Conflicting or distorted information regarding health benefits and risks of personal behavioral choices can lead to confusion and doubt with respect to important health recommendations.⁶³ Researchers must work closely with their affiliated institutions to communicate their findings in the most responsible fashion. The scientific community should also take efforts to transmit clear and accurate messages to the public and avoid sensationalized headlines that favor their positions or findings.⁶⁴

There are limitations to our arguments that deserve mention. The body of sedentary behavior literature is small compared with the extensive research on smoking. Furthermore, measurement error is considerably higher for self-reported sitting. Objective assessment of sitting (using devices such as the activPAL [PAL Technologies Ltd, Glasgow, Scotland]) and more methodologically rigorous studies (e.g., longer follow-up studies) will allow for more conclusive and valid inferences to be drawn with respect to the health hazards of too much sitting. Smoking rates have significantly decreased in high-income countries, and these gains are a result of several decades of public health initiatives. Given the habitual nature of sitting, public health efforts to change population trends in sitting may be more successful than smoking efforts given sitting habits may be more amenable to change than addictive behaviors such as smoking. Unlike the smoking context, to date, there have been few documented public health initiatives to reduce sedentary behavior.

Given the current state of the evidence, equating sitting with smoking is unwarranted, misleading for the public, and may serve to distort and trivialize the ongoing and serious risks of smoking. The magnitude of the associations between sitting and health risks and corresponding absolute risk differences are small in comparison with the risks and risk differences associated with smoking. Betteridge’s Law of Headlines states that any headline that ends in a question mark can be answered by the word no. Is sitting the new smoking? No.