Introduction

Cervical cancer is the fourth most common cancer as well as the fourth most common cause of female cancer deaths worldwide. Cancer incidence and mortality rates vary geographically, and most cancers and associated deaths occur in less-developed regions [1]. Although the incidence of, and mortality from, cervical cancer have declined since the development of widespread screening programs, improved treatment, and most recently, the introduction of a vaccine against human papillomavirus, the disease burden remains high, and the numbers of long-term survivors are increasing [2]. In Korea, the incidence of cervical cancer has declined rapidly, and more cervical cancers are detected early because of the introduction of cervical cancer screening programs [3, 4]. However, cervical cancer remains the fifth most common cancer in the country and is associated with the largest number of long-term survivors [5].

Cancer survivors are at increased risk of sequelae of the cancer per se as well as of the cancer treatment: secondary cancer caused by radiotherapy, early ovarian failure attributable to surgery and/or radiotherapy, and non-cancer-related problems, including functional limitations, poor general health, and a reduced quality of life [6–9]. Cervical cancer survivors with known risk factors, such as human papillomavirus infection and smoking, are at a greater risk of secondary malignancies than are other cervical cancer survivors [10]. Healthful behaviors, such as smoking cessation, increased physical activity, consumption of a healthy diet, and weight control, can play roles in improving the long-term health-related outcomes of cancer survivors [11–13]. In addition, positive associations between the extent of physical activity and quality of life and between physical activity and other healthy behaviors have been found among cervical cancer survivors [14].

Although several population-based studies have evaluated the health behaviors of cancer survivors [6, 15–17], only a few focused on the behaviors of cervical cancer survivors in particular. As cancer survivors differ in terms of disease characteristics, risk factors, and health behaviors exhibited after diagnosis and treatment, a better understanding of the health behaviors and associated sociodemographic factors of cervical cancer survivors would allow physicians and public health professionals to provide effective individual- and population-based interventions to improve their health status and quality of life. Therefore, we used a population-based dataset to investigate the prevalence of smoking, alcohol consumption, physical activity, and obesity in cervical cancer survivors and to examine associations between sociodemographic factors and each health behavior.

Materials and Methods

Results

The sociodemographic characteristics of the 448 cervical cancer survivors and the 4,480 matched non-cancer controls are shown in Table 1. The cervical cancer survivors did not differ from the matched non-cancer controls with regard to age, educational attainment, and household income. The proportions of divorced, widowed, and unmarried subjects were higher (P = 0.030) in cervical cancer survivors compared with non-cancer controls. The groups did not significantly differ in terms of other sociodemographic characteristics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Comparison of basic characteristics between cervical cancer survivors and female respondents never diagnosed as having cancer.

https://doi.org/10.1371/journal.pone.0160682.t001

The prevalences of smoking, alcohol consumption, physical inactivity, and obesity were 2.7%, 23.9%, 62.0%, and 61.1%, respectively, in survivors, and these figures did not differ significantly from the comparable ones in matched non-cancer controls (2.2%. 26.4%, 64.6%, and 58.6%, respectively; Table 2). After adjusting for the matching variables and other covariates, the conditional logistical ORs for cervical cancer survivors compared with matched non-cancer controls were 1.11 (95% confidence interval [CI] = 0.60–2.06) in terms of smoking, 1.12 (95% CI = 0.94–1.52) in terms of alcohol consumption, 0.85 (95% CI = 0.69–1.05) in terms of physical inactivity, and 1.10 (95% CI = 0.89–1.35) in terms of obesity. Thus, the health behaviors of the two groups did not differ significantly.

The duration since diagnosis did not affect the prevalence of smoking, alcohol consumption, or physical inactivity. However, obesity decreased significantly in the 6–10 years after diagnosis, but it increased after 11 years (P = 0.004, not shown).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Health behaviors in cervical cancer survivors compared with female respondents never diagnosed as having cancer.

https://doi.org/10.1371/journal.pone.0160682.t002

Table 3 shows the associations of sociodemographic characteristics with other health-related factors and each health outcome in survivors. As current smokers were few in number (2.68%; 12/338), we excluded current smoking when performing multivariate logistic regression analysis. Current alcohol consumption decreased age (OR for 1-year age increment = 0.91, 95% CI = 0.88–0.95), with status as married or cohabitating (OR = 0.42, 95% CI = 0.23–0.78), and in highly educated survivors (OR = 0.49, 95% CI = 0.25–0.97). Physical inactivity was lower in survivors with a monthly household income ≥$2000 (OR = 0.61, 95% CI = 0.37–0.99), but higher in those who were currently employed and those with poor subjective health (OR = 2.16, 95% CI = 1.36–3.42; OR = 1.94, 95% CI = 1.23–3.05, respectively). Obesity increased with age (OR for 1-year age increment = 1.04, 95% CI = 1.01–1.08) and was less prevalent in cancer survivors diagnosed 6–10 years prior to the study, who were less likely to be obese (OR = 0.38, 95% CI = 0.20–0.72) compared with survivors diagnosed 5 years or less prior to the study.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Adjusted odds Ratio of health behaviors for each sociodemographic characteristic among the survivors of cervical cancer.

https://doi.org/10.1371/journal.pone.0160682.t003

Discussion

As cervical cancer survival rate has greatly improved and the number of long-term survivors has increased, problems with health behaviors and their long-term effects and management in cervical cancer survivors has been raised as important issue. In this study, we identified that cervical cancer survivors and matched non-cancer controls did not differ significantly with regard to smoking status, alcohol consumption, physical inactivity, or obesity.

The factors influencing unhealthy behaviors differed by behavior. Alcohol drinking was negatively correlated with educational level, which is consistent with previous data on cancer survivors who had undergone hematopoietic cell transplantation [25]. Physical inactivity was also inversely correlated with income, as noted previously [17]. In addition, married or cohabiting survivors drank less, and those who were employed or considered themselves unhealthy were more likely to be physically inactive, suggesting that survivors with lower education, lower income, unhealty subjective health status, being unmarried/uncohabitated, or employed were particularly vulnerable to the effects of unhealthy lifestyles. However, no sociodemographic factor studied was associated with obesity. Although previous studies identified several unhealthy behaviors were correlated in general populations [26] and in colon cancer [27] and gynecological cancer survivors [14], no interrelationship among unhealthy behaviors was observed in cervical cancer survivors in the current study, confirming modification of each health behaviors needs to be accessed individually.

An earlier study found that survivors of gynecological cancer had a higher current smoking rate than did non-cancer controls, but neither alcohol consumption nor level of physical activity differed between the two groups [28]. Another study found no significant difference with respect to smoking, drinking, or physical activity between all cancer survivors and non-cancer controls but reported that cervical cancer survivors generally exhibited more unhealthy behaviors [17]. Thus, cervical cancer survivors might be expected to exhibit more unhealthy behaviors than other cancer survivors.

Several previous studies reported that the smoking prevalence among cervical cancer survivors was higher than that among survivors of other types of cancers or in general populations [14–16]. According to research conducted in Western countries, survivors of cervical or gynecological cancer exhibited higher smoking rates (21.5–46.0% in the USA [15, 16, 29]; 20.9% in Australia [14]) than did our study population (2.7%). In Korea, the female smoking rate is the lowest (4.3%) of all countries in the Organization for Economic Cooperation and Development [30]. Additionally, our study subjects were voluntary health examinees and thus were health-conscious, which explains the lower smoking rates compared with those in other studies of both the cervical cancer survivors and the matched non-cancer controls. One previous study reported that 56.9% of cancer survivors consumed alcohol regularly [15]. This figure is higher than that in the present study, which is more consistent with the finding reported by another study: that the prevalence of risky drinking in cervical cancer survivors was only 7.2% [16]. The proportions of cervical cancer survivors engaging in physical activity have varied among studies as have the definitions of such activity. About 50% of survivors engaged in the recommended level of physical activity [19]; this was lower than the proportion noted in the general population. In other studies, 20–30% of survivors were physically active [15, 16]. Our figure is intermediate, as it is similar to that of Beesely et al. [14]. The prevalence of obesity and overweight status in Korea noted in the present and previous studies are comparable, at 20–40% [14, 29].

A cancer diagnosis may motivate an individual to adopt healthy behaviors [31]; a higher prevalence of positive behaviors was evident in cancer survivors than in cancer-free subjects in several earlier studies [17, 32, 33]. However, consistent with other studies [17, 28], we found that such behaviors were not evident in cervical cancer survivors. Thus, the “teachable moment” used to trigger healthy behavior in cancer survivors may be difficult to identify in cervical cancer survivors.

To the best of our knowledge, this population-based study is the first to examine the health behaviors, associated sociodemographic factors, and other health-related issues in cervical cancer survivors, a population that is vulnerable to effects of unhealthy behaviors.

Although we have gathered valuable information, several limitations should also be noted. First, our cross-sectional approach using a baseline questionnaire of cohort study can measure the prevalence of various health behaviors and differences between groups, but it is difficult to infer causality. Second, all cervical cancer survivors and matched non-cancer controls had voluntarily visited health examination centers for health check-ups. Therefore, they may be more concerned about their health status and practice better health behaviors than cancer survivors in general or general populations. For example, one previous study found that cervical cancer survivors recruited from tertiary hospitals had more comorbid chronic diseases than did general female populations [34]. However, in the present study, the proportion of cervical cancer survivors with comorbidities was not significantly different from controls. Thus, our results may not be readily generalizable. However, patients who complete treatment and undergo medical check-ups in health examination centers rather than cancer treatment hospitals may more readily return to daily life and thus may be prime targets for community-level efforts to provide management for cervical cancer survivors. To ensure the comparability of health behaviors, we matched survivors and controls by age, educational level, and income. We defined cervical cancer survivors using a self-report method. Although some information bias may have been affected our results, a previous study found that a self-reported cancer history was reliable [35].

In conclusion, we found that the health behaviors of cervical cancer survivors were no worse than those of matched non-cancer controls. The findings of a previous report [17] suggest that the health behaviors of cervical cancer survivors may, in fact, be worse than those of other cancer survivors. Several sociodemographic characteristics were associated with unhealthy behaviors among cervical cancer survivors, but we found no interrelationship among such behaviors. Thus, those at high risk for each unhealthy behavior differed, and interventions should be designed to modify lifestyles and interventions targeting each unhealthy behavioral factor of survivors with particular sociodemographic characteristics may be more appropriate for cervical cancer survivors. Efforts to induce healthy behavior in cervical cancer survivors are unfortunately not routine during cancer treatment and rehabilitation. Yet, health behaviors are modifiable. The identification of cervical cancer survivors at higher risk of an unhealthy lifestyle who would thus benefit from interventions would allow individual- and population-based programs to use their limited resources more effectively.