Prevalence of Cigarette Smoking among Adult Cancer Survivors in Korea

The Korea National Health and Nutrition Examination Survey (KNHANES)

The KNHANES is a population-based, cross-sectional survey to assess the health-related behavior, the health condition, and the nutritional state of Koreans. It consists of the Health Interview Survey (HIS), the Nutrition Survey, and the Health Examination Survey. This study included 23770 adults, aged 18 years or older, who participated in the HIS from 2007 to 2010. The detailed descriptions of the plan and operation of the survey have been described on the KNHANES website (http://knhanes.cdc.go.kr/).

Ascertainment of cancer diagnosis history

Data on the cancer history were obtained via the HIS. The interview was performed by trained personals asking a specific question "Have you ever been diagnosed with cancer by a physician?" If the respondent concurred with the question, he or she was considered as a "cancer survivor" regardless of time duration since diagnosis, whereas those who denied were classified as "non-cancer control". Cancer survivors were further questioned on the type of cancer and the age at cancer diagnosis. In case of multiple cancer diagnoses, the latest one was used as an index cancer. We excluded cancer survivors who were diagnosed with cancer at the age of 18 or younger. Time duration since cancer diagnosis was estimated by subtracting the age at cancer diagnosis from the current age and categorized into "<2 years", "2 to <5 years", "5 to <10 years", and "more than 10 years". The type of cancer was classified into smoking-related and smoking-unrelated cancers. Smoking-related cancers included lung, mouth, lips, nasal cavity, and paranasal sinuses, larynx, pharynx, esophagus, stomach, pancreas, kidney, bladder, uterine cervix, colon/rectum, ovary cancers, and acute myeloid leukemia.10

Smoking history

Smoking history was revealed by interviews. It included current smoking status, smoking cessation duration, and whether other smoker(s) existed in the household. Smoking status was classified into "never-smoker", "former-smoker", and "current-smoker". Individuals who reported to smoke less than 100 cigarettes in their lifetime were considered "never-smokers". Those who smoked at least 100 cigarettes in their lifetime and stopped smoking prior to interview were considered "former-smokers". The former-smokers were further asked the time point of the smoking cessation. The time intervals between index cancer diagnosis and smoking cessation were calculated to check if the participants stopped smoking before or after cancer diagnosis. The "current-smokers" were further questioned on smoking amount and plans for smoking cessation.

Covariates measurement

We collected data on the demographic (age and marital status) and socioeconomic (household income and education) factors, and health perception (self-perceived health status).

Marital status was classified into "married" and "single". The household income level was classified into quartiles and education level into ≤9, 10-12, and ≥13 years of education. The self-reported perceived health status was categorized into "good", "moderate", and "poor".

Statistical analysis

Analytic sample included 23777 Koreans aged 18 years old or older, including 650 cancer survivors

Descriptive statistics of the demographic characteristics and smoking patterns were presented according to the cancer history. The differences in characteristics between cancer survivors and non-cancer controls were assessed using the logistic regression for categorical variable and the general linear modeling for continuous variable. Smoking rates were presented according to the cancer subtypes if there were more than 30 cases of cancer survival for a specific subtype.

Multivariate logistic regression analyses were performed to examine the relationship between the smoking status and related factors after adjustment for covariates. The outcome variable was the smoking status. Factors known to affect the likelihood of being a persistent smoker were adjusted. Those factors were gender, age, education, household income level, marital status, self-perceived health status, years since cancer diagnosis, smoking household member, and cancer type (smoking-related or not). We additionally adjusted for the survey year.

All analyses were weighted to incorporate sampling weight considering the multistage probability sampling design of KNHANES and non-response with SPSS 18.0 statistical package (SPSS Inc., Chicago, IL, USA). The study protocol was approved by the Institutional Review Board (IRB No. IIT-2012-203). Informed consent was waived.

The study population

The baseline characteristics of the study population are shown in Table 1. In this representative population, 2.1% of the Korean adults were cancer survivors. Cancer survivors were older, more likely to be female, married, and less educated. They had lower income level than non-cancer controls (all p<0.001). As for the time interval since cancer diagnosis, 19.0% of the cancer survivors had a time interval of 2 years or less, 34.4% between 2 and 5 years, 23.0% between 5 and 10 years and 23.5% longer than 10 years ago.

Table 1
General Characteristics of Study Population

• Click for larger image
• Click for full table
• Download as Excel file

Smoking status

Cancer survivors were less likely to be current smokers (7.8±1.3% vs. 26.4±0.4%), but more likely to be former smokers (29.2±2.3% vs. 19.5±0.3%) or never smokers (63.0±2.3% vs. 54.2±0.4%) than non-cancer controls (p<0.001) (Table 2). The difference in current smoking rate between cancer survivors and non-cancer controls was observed only among males (46.3±0.6% vs. 14.5±2.9%, p<0.001), but not among females (6.5±0.3% vs. 4.2±1.3%, p=0.144). Among the cancer survivors, 14.6% reported being current-smokers at the time of cancer diagnosis. Interestingly, 53.4% among them continued to smoke even after their cancer diagnosis. Only one fourth of the current smokers planned to quit smoking within 6 months.

Table 2
Smoking Status of Cancer Survivors and Non-Cancer Controls

• Click for larger image
• Click for full table
• Download as Excel file

Stratifying the smoking status according to gender and age, in males the current smoking rates in cancer survivor groups were lower than non-cancer control group throughout all age groups (Fig. 1A). In contrast, in females the smoking rate of cancer survivors in middle-aged group (age between 45-64 years) was not different from that of non-cancer controls (5.8±2.7 vs. 4.7±0.4, p=0.603) (Fig. 1B).

Fig. 1
Percentage of current smokers among cancer survivors and remaining Korean population by age, 2007-2010. (A) Male. (B) Female.

• Click for larger image
• Download as PowerPoint slide

Types of cancers diagnosed and the site specific smoking rate among the cancer survivors

Fig. 2 shows the smoking rate of the survivors with cancer types that have 30 and more cases. Liver cancer survivors had the highest smoking rate (16.7%), followed by stomach cancer (14.0%), colon cancer (13.3%), and cancer of the urinary tract (12.1%). The smoking rate was 9.8% in survivors of smoking-related cancers, whereas the smoking rate was only 4.0% in those of smoking-unrelated cancer (Table 3).

Fig. 2
Cancer site specific smoking rate among Korean cancer survivors, 2007-2010. Only cancer types which had more than 30 cases of prevalence among cancer survivors were presented.

• Click for larger image
• Download as PowerPoint slide

Table 3
Odds of Being Current Smoker among Cancer Survivors

• Click for larger image
• Click for full table
• Download as Excel file

Predictors of persistent-smoking among cancer-survivors

As for the risk factor for persistent-smoking among the cancer survivors, male gender [odds ratio (OR), 6.34; 95% confidence interval (CI), 2.62-15.31], middle-aged group (OR, 2.74; 95% CI, 1.12-6.72), the low income (lowest income: OR, 4.10; 95% CI, 1.19-14.15), living with smoking family members (OR, 5.49; 95% CI, 2.42-12.48), and the poor self-perceived health status (OR, 2.78; 95% CI, 1.01-7.71) were associated independently with persistent smoking in Korean cancer survivors (Table 3).

Risk factors of persistent smoking among the Korean cancer survivors

We showed that male gender, middle age, lowest household income, living with smoking family members, and poor health self-perception were associated independently with the persistent smoking among the cancer survivors. This result is in line with Western studies.15, 16, 17 Among the risk factors, the male gender and the living with smoking family members were strong risk factors. Concerning the male gender, it was associated with six-fold increased risk for persistent smoking (OR, 6.34; 95% CI, 2.62-15.31). Living with a smoking family member was associated with five-fold increased odd (OR, 5.49; 95% CI, 2.42-12.48) of persistent smoking.7, 18 Smoking family members significantly lower the motivation to quit smoking.7 Thus, it is essential to involve family members for successful smoking cessation among the cancer survivors.19

Persistence smoking according to the cancer types

The smoking rate of cancer survivors varies with cancer type. We showed that smoking-related cancer survivors had a higher smoking rate than that of smoking-unrelated cancer survivors (9.8% vs. 4.0%, p=0.080) with a marginal significance, exposing smoking-related cancer survivors to increased risks for the development of secondary cancers. In particular, the smoking rate of cervical cancer-a smoking-related cancer-survivors was three times higher than that of breast cancer-a smoking-unrelated cancer-(5.9% vs. 1.8%) in our study population, as was in the US National Health Interview Survey (46% vs. 14.1%),20 suggesting that some of the health behaviors promoting cancer development persist still after cancer diagnosis.

The need for intervention

Smoking cessation after cancer diagnosis increases survival rate and the quality of life,21 decreases the development of a secondary malignancy, and relieves psychological distress.14 Despite many beneficial effects of smoking cessation, less than 50% of patients quit smoking after cancer diagnosis in this study. Western studies reported that the smoking cessation rate after cancer diagnosis ranged between 22-69%.7 Strikingly, 50% of ex-smokers reported to restart cigarette smoking,22 suggesting a high addictive potential of cigarette smoking and the need for comprehensive and systematic prevention programs for smoking cessation.

The optimal timing for such intervention may be as soon as after cancer diagnosis,6 when the patients are highly motivated to improve their health-related behavior.23 Thus, healthcare providers, especially those involved in the initial cancer treatment, should take a primary role in counseling patients and encouraging smoking cessation. Furthermore, smoking cessation programs should be accessible to all patient groups regardless of socioeconomic status. Currently, smoking cessation therapy is not reimbursed by the National Health Insurance in Korea, necessitating changes in the national health policy.

Study limitations

There are several limitations in this study. First, KNHANES is not designed to elucidate the reasons why cancer survivors continue smoking after cancer diagnosis. Second, the data on smoking history were self-reported. Underreporting of smoking history is common among patients.24 Third, although adjusted for significant clinical variables in Korean adults, there could be residual or unmeasured confounders which may influence smoking behavior.

In conclusion, Korean cancer survivors have lower smoking rate than Western counterparts. However, despite hazardous effect of cigarette smoking, more than half of the patients continue smoking after cancer diagnosis, and especially male gender, middle-aged group, the low income level, living with smoking family members, and the poor perceived health status are risk factors for the persistent smoking. The high cigarette smoking rate after diagnosis mandates comprehensive and systematic intervention for smoking cessation among cancer survivors.