Abstract
Introduction:
Worldwide, an estimated 189 million adults smoke tobacco “occasionally” but not every day. Yet, few studies have examined the health risks of non-daily smoking.
Methods:
Data from the 1991, 1992, and 1995 US National Health Interview Surveys (NHIS), a nationally representative sample of 70,913 US adults (18–95 years) were pooled. Hazard ratios (HRs) and 95% confidence intervals (CIs) for death through 2011 were estimated from Cox proportional hazards regression using age as the underlying time metric and stratified by 5-year birth cohorts in 2017.
Results:
Non-daily smokers reported smoking a median of 15 days and 50 cigarettes per month in contrast to daily smokers who smoked a median of 600 cigarettes per month. Compared with never smokers, lifelong non-daily smokers who had never smoked daily had a 72% higher mortality risk (95%CI, 1.36–2.18): higher risks were observed for cancer, heart disease, and respiratory disease mortalities. Higher mortality risks were observed among lifelong non-daily smokers who reported 11–30 (HR, 1.34, 95%CI, 0.81–2.20), 31–60 (HR, 2.02, 95%CI, 1.17–3.29), and >60 cigarettes per month (HR, 1.74, 95%CI, 1.12–2.72) than never smokers. Median life-expectancy was about five years shorter for lifelong non-daily smokers than never smokers. As expected, daily smokers had even higher mortality risks (HR, 2.50, 95%CI, 2.35–2.66) and shorter survival (ten years).
Conclusions:
While the mortality risks of non-daily smokers are lower than daily smokers, they are still substantial. Policies should be specifically directed at this growing group of smokers.
Keywords: cigarette, social smoking, occasional smoking, mortality
INTRODUCTION
Approximately 17% of the world’s 1.1 billion smokers (189 million) are occasional smokers who smoke on only some days of the month.1 Such non-daily smokers constitute a growing proportion of current smokers in the United States (US). Whereas the number of daily cigarette smokers decreased from 37 million to 28 million from 2005 to 2015 in the US, the number of non-daily smokers who smoke on only some days of the month increased slightly.2 In 2015, an estimated 8.9 million people in the US were non-daily cigarette smokers.2 Because long-term non-daily smoking was once an uncommon pattern, few studies have assessed its health effects.3–11 Indeed, many non-daily smokers in the US and elsewhere consider themselves to be non-smokers and further believe that their level of smoking is harmless.12
The limited existing literature suggests that regular non-daily cigarette smokers may have higher mortality rates than never smokers, even if they have never smoked every day (Appendix Table 1).3–11, 13, 14 However, these prior studies are limited by small sample size, inconsistent definitions of non-daily smoking, lack of lifetime cigarette use information, and lack of detailed current usage patterns, precluding assessment of dose-response.
The National Health Interview Survey (NHIS) offers the advantage of detailed cigarette smoking assessment, including days smoked in the past 30 days, number of cigarettes used on days smoked, and whether non-daily smokers had ever smoked daily, a nationally representative US sample, and linked mortality follow-up data. These data allowed us to examine the association of non-daily cigarette smoking with mortality throughout the age range and across racial/ethnic groups.
METHODS
Study Population
The NHIS, an annual national household survey, uses a complex stratified multistage cluster sample design with sample weights to account for differential sampling and non-response rates of sampled persons and post-stratification adjustments to US population totals. The NHIS collects comprehensive information on demographics, behaviors, socioeconomic and health status among the US civilian non-institutionalized population. In each survey year, interviewers visit about 35,000 to 50,000 households nationwide and collect data from approximately 75,000 to 130,000 individuals. NHIS is designed by the Centers for Disease Control and Prevention (CDC)’s National Center for Health Statistics (NCHS) and administered by the US Census Bureau. Prior to the public release, the contents of the public-use data file went through an extensive review by the NCHS Disclosure Review Board. Data and detailed information on the NHIS including procedures for informed consent can be found at the NHIS website (https://www.cdc.gov/nchs/nhis.htm).
Exposure Assessment
In addition to the core questionnaires assessing health and demographic information, each NHIS includes one or more supplements, which either stand alone or are embedded in the core questionnaire, that respond to public health concerns. Tobacco use questions for adults aged 18 years and over have been included in the supplements since 1965.
Data on cigarette smoking from the 1991 Health Promotion and Disease Prevention Supplement, the 1992 Cancer Control Supplement, and the 1995 Year 2000 Objectives Supplement that included detailed questions regarding cigarette use during the past month and additional questions for non-daily smokers allowing determination of prior daily use were harmonized.15 Participants who reported that they had smoked at least 100 cigarettes in their lives were identified as “ever cigarette smokers”, and further categorized as ”former” or “current” smokers. Current smokers were those who had smoked either occasionally (“non-daily smokers”) or every day (“daily smokers”) in the last 30 days.16 Former smokers who reported smoking daily at some point were categorized as “former daily smokers” and former smokers who had never smoked daily were “former never-daily smokers”. Current non-daily smokers were further categorized into those who had previously smoked every day and those who were always non-daily smokers (“lifelong non-daily smokers”).
Additional information varied by survey year, including age at first trying cigarettes (1995), age starting to smoke regularly (1992), number of days smoked in the past 30 days (all), cigarettes used on days smoked (all), number of months since last smoked daily (1991 and 1992), age at cessation (1992 and 1995), and number of years since cessation (1992 and 1995). Ever-use of cigar, pipe, and smokeless tobacco were assessed in 1991 and 1992. Ever-users were defined as participants who had used pipes at least 50 times, cigars 50 times, chewing tobacco 20 times, or snuff 20 times.
Mortality risks among non-daily smokers were assessed by number of days smoked in the past 30 days and number of cigarettes on days smoked among 961 lifelong non-daily smokers who provided such information. Four categories of cigarettes smoked per 30-day month were created: ≤10, 11–30, 31–60, and >60 cigarettes per 30-day month. Current non-daily smokers who previously smoked daily were not included in these analyses, because information on their prior daily smoking pattern, such as cigarettes smoked per day, was not available. Daily smokers were also categorized into ≤60, 61–300, 301–600, 601–900, and >900 cigarettes smoked per 30-day month.
Mortality Ascertainment
The NHIS has been linked to mortality data through probabilistic record matching with the National Death Index (NDI). Mortality through 2011 were identified using the NHIS-Linked Mortality Files (LMFs). In addition to all-cause mortality, mortality from smoking-related causes of death [International Classification of Diseases version 10 (ICD-10) codes: cancer, C00-C97; heart disease, I00-I09, I11, I13, I20-I51; cerebrovascular disease, I60-I69; and respiratory disease including chronic lower respiratory diseases, J40-J47 and influenza and pneumonia, J09-J18] and all other causes available in the public-use NHIS-LMFs were examined.17, 18
Statistical Analysis
A total of 73,084 individuals aged 18 years or older responded to the supplements and the core questionnaires in 1991, 1992, and 1995: overall response rates for the core questionnaire was about 95% and for the tobacco supplements were 87.8% (1991), 87.0% (1992), and 80.9% (1995). Of these individuals, participants who were not eligible for mortality follow-up (n=908), had unknown cause of death (n=82), had unknown cigarette smoking status (n=1,125), were younger than 18 years (n=8) or 96 years or older (n=45) at the time of survey, or reported smoking 96 or more cigarettes per day (n=3) were excluded, resulting in 70,913 individuals (42,454, 11,664, and 16,795 from 1991, 1992, and 1995, respectively).
Participants were followed from the date of interview through the date of death, the date before they turned 96 years old, or December 31st, 2011, whichever came first. Hazard ratios (HRs) and 95% confidence intervals (CIs) for each mortality outcome were estimated using Cox proportional hazards regression using age as the underlying time metric. Baseline hazards in the Cox regression were stratified by 5-year birth cohort. Covariates included sex, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and non-Hispanic other), education (less than high school, high school, some college/associate degree, and bachelor degree or higher), and survey year (1991, 1992, and 1995), based on literature review. Categories were created for missing values. Sub-group analyses were stratified by within or after five years of follow-up, sex, and race/ethnicity. Sensitivity analyses included 1) excluding participants who reported ever using other tobacco (available in 1991 and 1992) and additionally adjusting for 2) second-hand smoke exposure (1991 and 1992), 3) physical activity (1991 and 1995), and 4) family income and alcohol intake (1991).
Adjusted survival curves by smoking status were also created using these Cox proportional hazards models. However, for the survival curves to reflect the mortality rates of the combined birth cohorts of the US population represented by the surveys, the baseline hazards in these models were not stratified by 5-year birth cohort. Analyses were conducted in 2017 using SAS-callable SUDAAN release 11.0.1 to account for the sample weighted complex survey sample design for the NHIS.19 The adjusted survival curves were computed using the PHREG procedure in SAS/STAT Proc version 13.2 (SAS Institute Inc., Cary, NC) with the weight option including the NHIS sample weights.
RESULTS
The median age of the 70,913 participants (29,943 men and 40,970 women) was 41 years. There were 36,013 (51.0%) never, 16,315 (23.2%) former, and 18,585 (25.8%) current smokers. About 17% (n=3,166) of current smokers were non-daily smokers, of whom 2,039 (64.7%) reported previously smoking daily, and 1,127 (35.3%) reported never smoking daily. Among former smokers, 88.6% had ever smoked daily.
Per 30-day month, non-daily smokers reported smoking a median of 15 days and 50 cigarettes. In contrast, daily smokers smoked a median of 600 cigarettes per 30-day month (20 cigarettes per day). A broad range of cigarette usage patterns during the month were observed among lifelong non-daily smokers (Appendix Table 2). Those who smoked >60 cigarettes in the past 30 days smoked a median of eight cigarettes per day and 20 days per 30-day month, whereas those who smoked 10 or fewer cigarettes in the past 30 days smoked a median of two cigarettes per day and three days per 30-day month.
Compared with current daily smokers, current non-daily smokers were younger and more likely to belong to racial/ethnic groups other than non-Hispanic white, and have higher educational achievement (Table 1). These patterns were also observed when comparing current lifelong non-daily smokers to those who had previously smoked daily, as well as comparing former never-daily smokers to former daily smokers. Current non-daily smokers started smoking at a slightly older age than current daily smokers. Among current non-daily smokers, those who previously smoked daily smoked on more days in the past 30 days than lifelong non-daily smokers (15 vs. 10 days).
Table 1:
Demographic and smoking characteristics: The 1991, 1992, and 1995 National Health Interview Survey
| All | Never smokers | Current smokers | Former smokers | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Daily | Non-daily | Ever daily | Never daily | |||||||||||
| Previously daily | Lifelong non-daily | |||||||||||||
| N | 70,913 (100) | 36,013 (100) | 15,419 (100) | 2,039 (100) | 1,127 (100) | 14,580 (100) | 1,735 (100) | |||||||
| Survey year | ||||||||||||||
| Age at interview | 41 (30 – 57) | 38 (28 – 55) | 40 (31 – 52) | 38 (29 – 49) | 32 (25 – 41) | 52 (40 – 66) | 43 (32 – 59) | |||||||
| Sex | ||||||||||||||
| Racea | ||||||||||||||
| Educationb | ||||||||||||||
| Smoking characteristics | ||||||||||||||
| Age when first tried smoking cigarettes (1995; N = 7,862) | 16 (13 – 18) | - | 15 (13 – 18) | 16 (14 – 18) | 16 (14 – 18) | 16 (13 – 18) | 16 (14 – 18) | |||||||
| Number of days smoked in the past 30 daysc | 15 (5 – 20) | - | - | 15 (8 – 20) | 10 (4 – 15) | - | - | |||||||
| Number of cigarettes smoked per dayd | 20 (12 – 20) | - | 20 (12 – 20) | - | - | - | ||||||||
| Number of cigarettes smoked per day on days smokede | 4 (2 – 7) | - | - | 4 (2 – 10) | 3 (2 – 5) | - | - | |||||||
| Number of months since the last time smoking dailyf (1991 and 1992) | 12 (3 – 60) | - | - | 12 (3 – 60) | - | - | - | |||||||
| Age at cessationg (1992 and 1995) | 36 (26 – 49) | - | - | - | - | 37 (28 – 50) | 26 (21 – 36) | |||||||
| Number of years since cessationg (1992 and 1995) | 10 (4 – 20) | - | - | - | - | 10 (4 – 20) | 12 (4 – 25) | |||||||
| Ever use of other tobacco productsh (1991 and 1992) | 7,567 (15.7) | 1,733 (7.1) | 2,180 (21.1) | 245 (18.6) | 103 (12.5) | 3,003 (29.4) | 303 (26.0) | |||||||
Weighted median and interquartile range (IQR) for continuous variables; N (weighted %) for categorical variables.
Missing values were identified for 249 individuals (0.4%).
Missing values were identified for 171 individuals (0.3%).
Only for current non-daily smokers (N = 2,915)
Only for current daily smokers (N = 15,294)
Only for current non-daily smokers (N = 2,918)
Only for current non-daily smoker who ever smoked daily (N = 1,288)
Only for former smokers (N = 5,213)
Reporting ever-using cigars at least 50 times, pipe at least 50 times, chewing tobacco at least 20 times, or snuff at least 20 times on the 1991 and 1992 surveys (completed by 54,118 participants).
A total of 16,761 deaths were identified through the end of 2011, including deaths from cancer (n=4,114), heart disease (n=3,830), cerebrovascular disease (n=1,121), respiratory disease (n=1,361), and other causes (n=6,335). All-cause mortality risks were higher among current lifelong non-daily smokers than never smokers (HR, 1.72, 95% CI, 1.36–2.18) (Table 2). As expected, current daily smokers had even higher risk of all-cause mortality (HR, 2.50, % CI, 2.35–2.66). Higher risks of all-cause mortality were observed for current non-daily and daily smokers relative to never smokers across sub-groups of follow-up time (≤ or > five years), sex, and race/ethnicity, although the number of deaths were small in some strata. Associations were also similar when stratified by survey year (Appendix Table 3). In sensitivity analyses, similar associations were observed after excluding ever regular users of other tobacco products, and also after additionally adjusting for second-hand smoking, physical activity, or alcohol intake and income.
Table 2.
Cigarette smoking and mortality among current smokers in the National Health Interview Survey
| Never | Current daily | Current non-daily | ||||||
|---|---|---|---|---|---|---|---|---|
| Previously daily | Lifelong non-daily | |||||||
| All-cause mortality | ||||||||
| All participants (N = 70,913) | N | 36,013 | 15,419 | 2,039 | 1,127 | |||
| Death | 6,808 | 4,178 | 366 | 147 | ||||
| HR (95% CI)a | 1.00 | 2.50 (2.35 – 2.66) | 1.62 (1.35 – 1.93) | 1.72 (1.36 – 2.18) | ||||
| Excluding ever users of cigars, pipes, chewing tobacco, or snuff (N = 46,551)b | N | 25,483 | 9,859 | 1,178 | 816 | |||
| Death | 4,961 | 2,594 | 200 | 116 | ||||
| HR (95% CI) | 1.00 | 2.47 (2.29 – 2.68) | 1.75 (1.39 – 2.21) | 1.66 (1.22 – 2.26) | ||||
| Additionally adjusted for second-hand smoke exposure (N = 53,834)c | N | 27,049 | 11,993 | 1,419 | 917 | |||
| Death | 5,410 | 3,366 | 272 | 129 | ||||
| HR (95% CI) | 1.00 | 2.19 (2.01 – 2.39) | 1.41 (1.11 – 1.79) | 1.59 (1.21 – 2.08) | ||||
| Additionally adjusted for physical activity level (N =55,034)d | N | 28,027 | 12,004 | 1,633 | 826 | |||
| Death | 4,779 | 3,057 | 258 | 103 | ||||
| HR (95% CI) | 1.00 | 2.49 (2.34 – 2.65) | 1.62 (1.37 – 1.92) | 1.72 (1.34 – 2.21) | ||||
| Additionally adjusted for income and alcohol intake (N = 42,454)e | N | 21,343 | 9,457 | 1,124 | 673 | |||
| Death | 4,324 | 2,648 | 212 | 97 | ||||
| HR (95% CI) | 1.00 | 2.39 (2.25 – 2.54) | 1.61 (1.35 – 1.92) | 1.62 (1.26 – 2.09) | ||||
| First 5 years of follow-up (N = 70,913) | N | 36,013 | 15,419 | 2,039 | 1,127 | |||
| Death | 1,545 | 850 | 1,963 | 1,094 | ||||
| HR (95% CI) | 1.00 | 2.23 (1.95 – 2.55) | 1.47 (1.09 – 1.97) | 2.05 (1.32 – 3.19) | ||||
| After 5 years of follow-up (N = 66,991) | N | 34,382 | 14,568 | 1,963 | 1,094 | |||
| Death | 5,263 | 3,328 | 1,673 | 980 | ||||
| HR (95% CI) | 1.00 | 2.58 (2.41 – 2.75) | 1.66 (1.36 – 2.02) | 1.65 (1.26 – 2.17) | ||||
| Men (N = 29,943) | N | 12,607 | 7,236 | 902 | 517 | |||
| Death | 1,864 | 2,144 | 181 | 51 | ||||
| HR (95% CI) | 1.00 | 2.54 (2.33 – 2.77) | 1.47 (1.13 – 1.92) | 1.99 (1.39 – 2.86) | ||||
| Women (N = 40,970) | N | 23,406 | 8,183 | 1,137 | 610 | |||
| Death | 4,944 | 2,034 | 185 | 96 | ||||
| HR (95% CI) | 1.00 | 2.49 (2.31 – 2.68) | 1.81 (1.46 – 2.24) | 1.55 (1.12 – 2.14) | ||||
| Non-Hispanic white (N = 52,707) | N | 25,303 | 11,962 | 1,379 | 586 | |||
| Death | 5,277 | 3,306 | 245 | 68 | ||||
| HR (95% CI) | 1.00 | 2.69 (2.51 – 2.88) | 1.71 (1.42 – 2.06) | 1.38 (1.02 – 1.88) | ||||
| Non-Hispanic black (N = 9,140) | N | 5,007 | 2,124 | 368 | 259 | |||
| Death | 982 | 637 | 91 | 51 | ||||
| HR (95% CI) | 1.00 | 1.82 (1.56 – 2.12) | 1.27 (0.86 – 1.89) | 2.21 (1.47 – 3.32) | ||||
| Hispanic (N = 6,735) | N | 4,227 | 958 | 227 | 227 | |||
| Death | 420 | 162 | 22 | 23 | ||||
| HR (95% CI) | 1.00 | 2.14 (1.67 – 2.75) | 1.76 (0.91 – 3.41) | 2.05 (1.27 – 3.32) | ||||
| Non-Hispanic other (N = 2,082) | N | 1,329 | 320 | 60 | 49 | |||
| Death | 111 | 61 | 8 | 5 | ||||
| HR (95% CI) | 1.00 | 2.02 (1.30 – 3.15) | 2.19 (0.87 – 5.49) | 0.64 (0.28 – 1.45) | ||||
| Cause-specific mortality | ||||||||
| Cancer (Death = 4,114) | Death | 1,343 | 1,323 | 118 | 31 | |||
| HR (95% CI)a | 1.00 | 3.25 (2.89 – 3.66) | 2.37 (1.79 – 3.14) | 1.50 (0.96 – 2.34) | ||||
| Heart disease (Death = 3,830) | Death | 1,669 | 782 | 75 | 37 | |||
| HR (95% CI) | 1.00 | 2.06 (1.81 – 2.35) | 1.26 (0.92 – 1.73) | 1.69 (1.08 – 2.64) | ||||
| Cerebrovascular disease (Death = 1,121) | Death | 574 | 196 | 14 | 13 | |||
| HR (95% CI) | 1.00 | 1.86 (1.49 – 2.30) | 1.14 (0.58 – 2.24) | 1.82 (0.85 – 3.90) | ||||
| Respiratory disease (Death = 1,361) | Death | 319 | 480 | 43 | 7 | |||
| HR (95% CI) | 1.00 | 7.55 (6.13 – 9.29) | 6.06 (3.84 – 9.58) | 3.04 (1.14 – 8.07) | ||||
| Other cause (Death = 6,335) | Death | 2,903 | 1,397 | 116 | 59 | |||
| HR (95% CI) | 1.00 | 1.94 (1.77 – 2.12) | 1.07 (0.80 – 1.42) | 1.69 (1.18 – 2.41) | ||||
HR, hazard ratio; CI, confidence interval
HR and 95% CI adjusted for sex, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, non-Hispanic other, and missing), education (< high school, high school, some college, college or higher, and missing), and survey year (1991, 1992, and 1995); age as the underlying time metric; baseline hazards were stratified by 5-year birth cohort; never smokers were the referent group.
Participants of the 1991 or 1992 surveys after excluding 7,567 participants who reported having ever-used cigars at least 50 times, pipe at least 50 times, chewing tobacco at least 20 times, or snuff at least 20 times.
Participants of the 1991 or 1992 surveys who reported whether anybody in their household smoked cigarettes.
Participants of the 1991 or 1995 surveys who reported physical activity information. Additionally adjusted for physical activity (active, insufficient, and inactive).
Alcohol intake was assessed only in the 1991 survey.
In addition to all-cause mortality, associations for mortalities from cancer, heart disease, and particularly respiratory disease were observed. The HR for respiratory disease mortality was 6.06 (95% CI, 3.84–9.58) among current non-daily smokers who had previously smoked daily and 3.04 (95% CI, 1.14–8.07) among current lifelong non-daily smokers.
Next, survival from age 18 to 95 by cigarette smoking status was examined (Figure 1). Median survival was 85 years in never smokers, 80 years in lifelong non-daily smokers, and 75 years for daily smokers. Thus, relative to never smokers, median survival was about five years shorter for lifetime non-daily smokers in the cohort. Survival was ten years shorter for daily smokers than never smokers.
Figure 1.
Survival curves from 18 to 95 years old by smoking status among never and current cigarette smokers
Generally higher risks with higher numbers of cigarettes smoked per 30-day month were also observed (Table 3). Compared with never smokers, HRs (95% CIs) were 1.08 (0.62–1.90), 1.34 (0.81–2.20), 2.02 (1.17–3.49), and 1.74 (95% CI, 1.12–2.72) for those who smoked ≤10, 11–30, 31–60, and >60 cigarettes per month. Despite lack of statistical power due to a small number of daily smokers who smoked 60 or less cigarettes per month (≤2 cigarettes per day), this level of smoking was also associated with mortality in the context of daily smoking (HR, 1.42, 95% CI, 0.95–2.10), although not reaching statistical significance. The number of cigarettes smoked per 30-day month was also modeled as a continuous variable and found a statistically significant inverted quadratic relationship with the log hazard of mortality, such that the increase in mortality risk with each additional cigarette smoked per month was higher at low cigarettes per month than it was at higher cigarettes per month.
Table 3.
Number of cigarettes smoked in the past 30-day montha and all-cause mortality
| N | Death | HR (95% CI)b | |
|---|---|---|---|
| Never smokers | 36,013 | 6,808 | 1.00 |
| Current lifelong non-daily smokers | |||
| ≤ 10 | 248 | 21 | 1.08 (0.62 – 1.90) |
| 11 – 30 | 244 | 27 | 1.34 (0.81 – 2.20) |
| 31 – 60 | 206 | 27 | 2.02 (1.17 – 3.49) |
| > 60 | 263 | 35 | 1.74 (1.12 – 2.72) |
| Current daily smokers | |||
| ≤ 60 (≤ 2 CPD) | 160 | 45 | 1.42 (0.95 – 2.10) |
| 61 – 300 (3 – 10 CPD) | 3,748 | 878 | 2.07 (1.86 – 2.30) |
| 301 – 600 (11 – 20 CPD) | 7,690 | 1,934 | 2.34 (2.16 – 2.54) |
| 601 – 900 (21 – 30 CPD) | 2,043 | 648 | 3.28 (2.86 – 3.76) |
| > 900 (> 30 CPD) | 1,653 | 612 | 3.33 (2.96 – 3.76) |
HR, hazard ratio; CI, confidence interval; CPD; cigarettes per day
Based on reported number of days smoked in the past 30 days and number of cigarettes smoked per day on the days they smoked.
HR and 95% CI adjusted for sex, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, non-Hispanic other, and missing), education (< high school, high school, some college, college or higher, and missing), and survey year (1991, 1992, and 1995); age as the underlying time metric; baseline hazards were stratified by 5-year birth cohort; never smokers were the referent group.
Next, mortality risks among former smokers was examined, but no clear pattern with time since cessation was observed among former never-daily cigarette smokers (Appendix Table 4). However, it should be noted that there were few former never-daily smokers in our analytic dataset, nearly all of whom quit at a young age. Among never-daily smokers, participants who had quit smoking had lower risks than those who continued to smoke, with lower risks observed among those who quit smoking earlier.
Among current non-daily smokers who reported previously smoking daily, no clear pattern was observed for the association of time since last smoking daily with mortality. Participants who had switched from daily to non-daily smoking at least 10 years before baseline had 1.77 (95%CI, 1.31–2.40) times the mortality risk of never smokers.
DISCUSSION
In our nationally representative US study, current non-daily smokers smoked a median of 50 cigarettes per month, far fewer than the median 600 cigarettes per month smoked by current daily smokers. Nevertheless, lifelong non-daily smokers had a 72% higher risk of mortality during follow-up than never smokers. Associations were observed across a range of smoking-related causes of death, in men and women, and among different racial/ethnic groups. Median survival was about five years shorter for current non-daily smokers than never smokers, illustrating the substantial impact of non-daily smoking on mortality.
Results from the current study are generally consistent with the limited number of previous studies of the association (Appendix Table 1).5, 6, 9–11, 14 Four previous cohort studies reported increased mortality risks of non-daily smoking, although were limited by a small sample size, lack of information on the number of days smoked and the number of cigarettes smoked per month, and inconsistent definitions for non-daily smoking; most lacked information as to whether non-daily smokers had ever previously smoked daily5, 9, 11, 14 In the NIH-AARP study, participants who reported consistently smoking <1 cigarette per day over their lifetime had 1.64 times higher mortality risk than never smokers.9 In the US National Longitudinal Mortality Surveys, current non-daily cigarette smokers had 1.6 times higher all-cause mortality compared with never tobacco users.14 Doll & Peto previously reported 10 years shorter survival among active smokers than never smokers in a large British cohort20 and similar findings have been observed elsewhere21, 22, which are consistent with our finding for daily smokers.
Our observed association between non-daily cigarette smoking and mortality is plausible. Cigarette smoke contains at least 7,000 chemical compounds; of these, hundreds are harmful, including many carcinogens.23 Smoking even one cigarette exposes a user to substantial levels of these chemicals. Despite smoking substantially fewer cigarettes per month than most daily smokers, non-daily smokers in the US population still typically smoke for many years and thus smoke a substantial number of cigarettes over their lifetime. For example, a person smoking 50 cigarettes per month for 35 years will end up smoking more than 20,000 cigarettes in their lifetime. Long-term exposure to cigarette smoke is known to be an important determinant of disease, both among active smokers 24–26 and among people exposed to second-hand cigarette smoke.27
Among lifelong non-daily smokers, increased mortality risks was obsered even at very low levels of cigarette use. As the number of non-daily smokers in our study was relatively small, future larger studies are needed to refine the associations, particularly among non-daily smokers using ≤30 cigarettes per month. Yet, our findings provide additional evidence that even very low levels of smoking are hazardous and that all smokers, no matter how few cigarettes they smoke, should quit, as recommended by current guidelines.28
In our study, more than a third of current non-daily smokers reported never smoking daily. These lifelong non-daily smokers were more likely to be young and racial/ethnic minorities than current non-daily smokers who reported previously smoking daily, as well as current daily smokers. A higher prevalence of non-daily smoking among racial/ethnic minorities has been noted previously.29, 30 Racial/ethnic minorities have been shown to have relatively higher risks of smoking-related disease at the same cigarettes per day, although few such studies have examined occasional, non-daily, smoking.29, 31 It will be important to carefully compare a dose-response association among non-daily smokers across ethnic groups in future studies. Unfortunately, we lacked sufficient case numbers to perform such analyses in the current study. Relatively less is known about other important aspects of non-daily smoking including levels of nicotine addiction and dependency 32, as well as optimal strategies for cessation. Future studies are needed to answer these questions.
Key strengths of the current study are the availability of detailed data on smoking frequency and intensity, the nationally representative sample of over 70,000 US adults, and nearly complete follow-up. High response rates in the NHIS also strengthened the study design: overall response rates in 1991, 1992 and 1995 were approximately 95% to the core questionnaires and 81–88% for the tobacco supplement questionnaires. Appropriate survey weights were applied in our analysis, making our findings representative of the US civilian non-institutionalized adult population. Our analysis included a wide age-range and allowed us to examine associations by sex and across racial/ethnic groups. Associations in each survey were also assessed separately and found consistent results.
Limitations
Our smoking data relied on participants’ recalling their tobacco use, and therefore potentially subject to misclassification. However, the validity of both non-daily and daily self-reported smoking have been shown to be good, correlating with biomarkers such as nicotine and its metabolites.33–36 Non-daily smokers may have taken up daily smoking during the follow-up. However, similar patterns were observed in mortality associations within and after five years of follow-up. Additionally, in the previous analysis in the NIH-AARP cohort, older smokers who reported consistently smoking <1 cigarette per day throughout their lifetime had increased risk of mortality.9 Yet, the statistical power was limited for some analyses. Finally, as in all observational studies, uncontrolled and residual confounding is a potential limitation.
CONCLUSIONS
In conclusion, in the US nationally representative NHIS-mortality follow-up study, lifelong non-daily cigarette smokers had 72% higher mortality risk than never smokers, with a median survival about five years shorter. Our findings provide additional evidence that even very low amouts of cigarette smoking are hazardous and support public health recommendations that all smokers, including low-intensity and non-daily smokers, should quit.28
Supplementary Material
Appendix Table 1. Summary of previous and current prospective cohort studies on non-daily or occasional cigarette smoking and disease risk
Appendix Table 2. Cigarette smoking patterns and causes of deaths by number of cigarettes smoked in the past 30 days among lifelong non-daily smokers
Appendix Table 3. Cigarette smoking and all-cause mortality in the 1991, 1992, and 1995 National Health Interview Survey
Appendix Table 4. Time since a) cessation and b) switching from daily to non-daily smoking and all-cause mortality
ACKOWLEDGMENTS
Funding
This study was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology & Genetics. The funding source reviewed and approved final submission but had no role in the design and conduct of the study; the collection, analysis, and interpretation of the data; the preparation of the manuscript; or the decision to submit the manuscript for publication.
Footnotes
Financial disclosures: No financial disclosure were reported by the authors of this paper.
Conflict of interest: None of the authors have conflicts of interest to declare.
REFERENCES
- 1.U.S. National Cancer Institute and World Health Organization. The Economics of Tobacco and Tobacco Control. National Cancer Institute Tobacco Control Monograph 21. NIH Publication No.16-CA-8029A: Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute; and Geneva, CH: World Health Organization; 2016. [Google Scholar]
- 2.Jamal A, King BA, Neff LJ, Whitmill J, Babb SD, Graffunder CM. Current Cigarette Smoking Among Adults - United States, 2005–2015. MMWR Morb Mortal Wkly Rep 2016;65(44):1205–1211. [DOI] [PubMed] [Google Scholar]
- 3.Schane RE, Ling PM, Glantz SA. Health effects of light and intermittent smoking: a review. Circulation 2010;121(13):1518–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rosengren A, Wilhelmsen L, Wedel H. Coronary heart disease, cancer and mortality in male middle-aged light smokers. J Intern Med 1992;231(4):357–62. [DOI] [PubMed] [Google Scholar]
- 5.Luoto R, Uutela A, Puska P. Occasional smoking increases total and cardiovascular mortality among men. Nicotine Tob Res 2000;2(2):133–9. [DOI] [PubMed] [Google Scholar]
- 6.Bjerregaard BK, Raaschou-Nielsen O, Sorensen M, Frederiksen K, Tjonneland A, Rohrmann S, et al. The effect of occasional smoking on smoking-related cancers: in the European Prospective Investigation into Cancer and Nutrition (EPIC). Cancer Causes Control 2006;17(10):1305–9. [DOI] [PubMed] [Google Scholar]
- 7.Prescott E, Scharling H, Osler M, Schnohr P. Importance of light smoking and inhalation habits on risk of myocardial infarction and all cause mortality. A 22 year follow up of 12 149 men and women in The Copenhagen City Heart Study. J Epidemiol Community Health 2002;56(9):702–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bjartveit K, Tverdal A. Health consequences of smoking 1–4 cigarettes per day. Tob Control 2005;14(5):315–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Inoue-Choi M, Liao LM, Reyes-Guzman C, Hartge P, Caporaso N, Freedman ND. Association of Long-term, Low-Intensity Smoking With All-Cause and Cause-Specific Mortality in the National Institutes of Health-AARP Diet and Health Study. JAMA Intern Med 2017;177(1):87–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Inoue-Choi M, Hartge P, Liao LM, Caporaso N, Freedman ND. Association between long-term low-intensity cigarette smoking and incidence of smoking-related cancer in the national institutes of health-AARP cohort. Int J Cancer 2018;142(2):271–280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lochen ML, Gram IT, Mannsverk J, Mathiesen EB, Njolstad I, Schirmer H, et al. Association of occasional smoking with total mortality in the population-based Tromso study, 2001–2015. BMJ Open 2017;7(12):e019107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Amrock SM, Weitzman M. Adolescents’ perceptions of light and intermittent smoking in the United States. Pediatrics 2015;135(2):246–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hackshaw A, Morris JK, Boniface S, Tang JL, Milenkovic D. Low cigarette consumption and risk of coronary heart disease and stroke: meta-analysis of 141 cohort studies in 55 study reports. BMJ 2018;360:j5855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Christensen CH, Rostron B, Cosgrove C, Altekruse SF, Hartman AM, Gibson JT, et al. Association of Cigarette, Cigar, and Pipe Use With Mortality Risk in the US Population. JAMA Intern Med 2018;178(4):469–476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Korn EL, Graubard BI. Wiley Series in Probability and Statistics In: Analysis of Health Surveys. Hoboken, NJ, USA: John Wiley & Sons, Inc.; 1999. p. 383–390. [Google Scholar]
- 16.Centers for Disease C, Prevention. Cigarette smoking among adults--United States, 1992, and changes in the definition of current cigarette smoking. MMWR Morb Mortal Wkly Rep 1994;43(19):342–6. [PubMed] [Google Scholar]
- 17.NCHS Surveys, 2011 Linked Mortality Files, Public-Use Data Dictionary. November 17, 2015. [cited 2017 April 17]; Available from: https://www.cdc.gov/nchs/data/datalinkage/public_use_data_dictionary_11_17_2015.pdf
- 18.Anderson RN, Minino AM, Hoyert DL, Rosenberg HM. Comparability of cause of death between ICD-9 and ICD-10: Preliminary estimates. National Vital Statistics Reports; Vol 49 No 2: National Center for Health Statistics, Center for Disease Control and Prevention; 2001. [PubMed] [Google Scholar]
- 19.Research Triangle Institute. SUDAAN Language Manual, Volumes 1 and 2, Release 11. Research Triangle Park, NC.; 2012. [Google Scholar]
- 20.Doll R, Peto R, Boreham J, Sutherland I. Mortality in relation to smoking: 50 years’ observations on male British doctors. BMJ 2004;328(7455):1519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jha P, Ramasundarahettige C, Landsman V, Rostron B, Thun M, Anderson RN, et al. 21st-century hazards of smoking and benefits of cessation in the United States. N Engl J Med 2013;368(4):341–50. [DOI] [PubMed] [Google Scholar]
- 22.Pirie K, Peto R, Reeves GK, Green J, Beral V, Million Women Study C. The 21st century hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet 2013;381(9861):133–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-Attributable Disease: A Report of the Surgeon General. Atlanta, GA: Centers for Disease Control and Prevention; 2010. [PubMed] [Google Scholar]
- 24.Doll R, Peto R. Cigarette smoking and bronchial carcinoma: dose and time relationships among regular smokers and lifelong non-smokers. J Epidemiol Community Health 1978;32(4):303–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lubin JH, Caporaso NE. Cigarette smoking and lung cancer: modeling total exposure and intensity. Cancer Epidemiol Biomarkers Prev 2006;15(3):517–23. [DOI] [PubMed] [Google Scholar]
- 26.Flanders WD, Lally CA, Zhu BP, Henley SJ, Thun MJ. Lung cancer mortality in relation to age, duration of smoking, and daily cigarette consumption: results from Cancer Prevention Study II. Cancer Res 2003;63(19):6556–62. [PubMed] [Google Scholar]
- 27.The health consequences of involuntary exposure to tobacco smoke: a report of the Surgeon General: U.S. Dept. of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2006. [PubMed]
- 28.Treating Tobacco Use and Dependence: 2008. Update. June 2015. Agency for Healthcare Research and Quality, Rockville, MD: Available from: http://www.ahrq.gov/professionals/clinicians-providers/guidelines-recommendations/tobacco/index.html [DOI] [PubMed] [Google Scholar]
- 29.Tobacco use among U.S. racial/ethinic minority groups. Atlanta, GA: U.S. Department of Health and Human Services, Center for Disease Control; 1998. [Google Scholar]
- 30.Reyes-Guzman CM, Pfeiffer RM, Lubin J, Freedman ND, Cleary SD, Levine PH, et al. Determinants of Light and Intermittent Smoking in the United States: Results from Three Pooled National Health Surveys. Cancer Epidemiol Biomarkers Prev 2017;26(2):228–239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Haiman CA, Stram DO, Wilkens LR, Pike MC, Kolonel LN, Henderson BE, et al. Ethnic and racial differences in the smoking-related risk of lung cancer. N Engl J Med 2006;354(4):333–42. [DOI] [PubMed] [Google Scholar]
- 32.Shiffman S Light and intermittent smokers: background and perspective. Nicotine Tob Res 2009;11(2):122–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Krall EA, Valadian I, Dwyer JT, Gardner J. Accuracy of recalled smoking data. Am J Public Health 1989;79(2):200–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Brigham J, Lessov-Schlaggar CN, Javitz HS, McElroy M, Krasnow R, Swan GE. Reliability of adult retrospective recall of lifetime tobacco use. Nicotine Tob Res 2008;10(2):287–99. [DOI] [PubMed] [Google Scholar]
- 35.Caraballo RS, Giovino GA, Pechacek TF, Mowery PD. Factors associated with discrepancies between self-reports on cigarette smoking and measured serum cotinine levels among persons aged 17 years or older: Third National Health and Nutrition Examination Survey, 1988–1994. Am J Epidemiol 2001;153(8):807–14. [DOI] [PubMed] [Google Scholar]
- 36.Shiffman S, Dunbar MS, Benowitz NL. A comparison of nicotine biomarkers and smoking patterns in daily and nondaily smokers. Cancer Epidemiol Biomarkers Prev 2014;23(7):1264–72. [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
Appendix Table 1. Summary of previous and current prospective cohort studies on non-daily or occasional cigarette smoking and disease risk
Appendix Table 2. Cigarette smoking patterns and causes of deaths by number of cigarettes smoked in the past 30 days among lifelong non-daily smokers
Appendix Table 3. Cigarette smoking and all-cause mortality in the 1991, 1992, and 1995 National Health Interview Survey
Appendix Table 4. Time since a) cessation and b) switching from daily to non-daily smoking and all-cause mortality