This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on https://publichealth.jmir.org, as well as this copyright and license information must be included.
Previous studies have hardly explored the influence of pre-pregnancy smoking and smoking cessation during pregnancy on the health-related quality of life (HRQoL) of pregnant women, which is a topic that need to be addressed. In addition, pregnant women in China constitute a big population in the largest developing country of the world and cannot be neglected.
This study aims to evaluate the HRQoL of pregnant women in China with different smoking statuses and further estimate the association between pre-pregnancy smoking, smoking cessation, and the HRQoL.
A nationwide cross-sectional study was conducted to determine the association between different smoking statuses (smoking currently, quit smoking, never smoking) and the HRQoL in pregnant women across mainland China. A web-based questionnaire was delivered through the Banmi Online Maternity School platform, including questions about demographics, smoking status, and the HRQoL. EuroQoL Group’s 5-dimension 5-level (EQ-5D-5L) scale with EuroQoL Group’s visual analog scale (EQ-VAS) was used for measuring the HRQoL. Ethical approval was granted by the institutional review board of the First Affiliated Hospital of Sun Yat-sen University (ICE-2017-296).
From August to September 2019, a total of 16,483 participants from 31 provinces were included, of which 93 (0.56%) were smokers, 731 (4.43%) were ex-smokers, and 15,659 (95%) were nonsmokers. Nonsmokers had the highest EQ-VAS score (mean 84.49, SD 14.84), smokers had the lowest EQ-VAS score (mean 77.38, SD 21.99), and the EQ-VAS score for ex-smokers was in between (mean 81.04, SD 17.68). A significant difference in EQ-VAS scores was detected between nonsmokers and ex-smokers (
Smoking history is associated with a lower HRQoL in pregnant Chinese women. Pre-pregnancy smoking is related to a lower HRQoL (EQ-VAS) and a higher incidence of depression/anxiety problems. Smoking cessation during pregnancy does not significantly improve the HRQoL of pregnant Chinese women. Among ex-smokers, the more cigarettes they smoke, the lower HRQoL they have during pregnancy. We suggest that the Chinese government should strengthen the education on quitting smoking and avoiding second-hand smoke for women who have pregnancy plans and their family members.
Active smoking increases the risk of developing chronic diseases and malignancy, such as chronic obstructive pulmonary disease and lung cancer [
The World Health Organization reported that tobacco use is a major risk factor for cardiovascular diseases, respiratory diseases, and cancers [
The HRQoL is a multidimensional indicator for measuring people’s physical, mental, emotional, and social health states in their lives over time. The HRQoL not only benefits the health perception at the individual level but also enables health agencies in legislation, community health planning, and business health projects [
Considering its importance, we aim to explore the effect of pre-pregnancy smoking and smoking cessation during pregnancy on pregnant women’s HRQoL in mainland China and compare the effects of pre-pregnancy smoking on pregnant women’s HRQoL (5 health dimensions). Additionally, this study also explored the relationship between the number of cigarettes consumed and the HRQoL of pregnant women in mainland China.
A nationwide cross-sectional study was performed to investigate pregnant women’s HRQoL using a self-administrative questionnaire across mainland China. The questionnaire was designed based on the Global Tobacco Surveillance System and EuroQoL Group’s 5-dimension (EQ-5D) questionnaire [
Ethical approval was granted by the institutional review board of the First Affiliated Hospital of Sun Yat-sen University (ICE-2017-296). All procedures were conducted following the Declaration of Helsinki. All participants signed the informed consent documents before participation in this study.
The web-based questionnaire was distributed through a national online platform (Banmi Online Maternity School) from August to September 2019. The Banmi Online Maternity School is a free platform that provides pregnancy knowledge for all internet users and serves more than 1 million users across China. The research group members of the Banmi Online Maternity School were the investigators. We advertised the survey with the wording “For providing you with more specific gestational health knowledge, we invite you to participate in this survey,” and no incentive was provided. A total of 16,811 questionnaires from pregnant women aged from 16 to 60 years were included, and 328 (1.95%) of them were excluded due to the living location not being mainland China. The final sample comprised 16,483 pregnant women from mainland China. According to the standards of the Chinese Center for Disease Control and Prevention, the research was performed in 7 administrative regions of mainland China: (1) the Northeast (Heilongjiang, Jilin, and Liaoning), (2) the North (Beijing, Tianjin, Hebei, Shanxi, and Inner Mongolia), (3) Central (Hubei, Hunan, and Henan), (4) the East (Shanghai, Shandong, Jiangsu, Anhui, Jiangxi, Zhejiang, and Fujian), (5) the South (Guangdong, Guangxi, and Hainan), (6) the Northwest (Shanxi, Gansu, Ningxia, and Xinjiang), and (7) the Southwest (Chongqing, Sichuan, Guizhou, Yunnan, and Tibet).
Participants’ sociodemographic information, including age, gestational age (weeks), address (provinces and cities), disposable income, smoking status, amount of cigarette consumption, smoking status of the spouse, and smoking duration (years), were collected. Previous studies have reported that maternal age, gestational age, and income level are related to people’s HRQoL [
We use the EQ-5D instrument, which consists of the EQ-5D-5L scale and the EQ-VAS, to evaluate the HRQoL of pregnant women. The EQ-5D-5L scale assesses 5 dimensions: mobility, self-care, usual activity, pain/discomfort, and anxiety/depression. Further, each dimension is addressed by 5 levels: (1) none, (2) slight problem, (3) moderate problem, (4) severe problem, and (5) extreme problem/unable. All dimension levels were converted into 1, 2, 3, 4, or 5 in the given order. Next, an EQ-5D index for each participant was calculated using the EQ-5D-5L Crosswalk Index Value Calculator. The possible maximal EQ-5D index is in the range of –0.224-1, where 1 indicates the highest health status, 0 represents death, and negative indices indicate the health status considered worse than death [
Data analysis was performed using STATA/SE version 14.0 for Windows (College Station, TX, USA). Normally distributed continuous variables were described using means and SDs. Nonnormal variables were presented as the median, and categorical variables were described using counts and percentages. Demographic data, including age, gestational age, address, smoking status, spouse’s smoking status, EQ-5D index, and EQ-VAS score, were included. The EQ-5D index and the EQ-VAS score were the outcome variables, and they were not normally distributed. A 1-way ANOVA test was performed to compare the continuous variables and analyze their variances. The Bartlett test was used to determine unequal variances. The Tamhane T2 method was used for pairwise comparison tests of EQ-VAS scores between groups, and the chi-square test performed to analyze the proportion of spouse smoking among groups. To estimate the relationship between independent variables (demographics) and dependent variables (EQ-5D index and EQ-VAS score), we also ran an ordinary least squares regression, which minimized the sum of the squared residuals to obtain adjusted values of the dependent variables. For nonsmokers and smokers, an ordered logistic regression with odds ratios (ORs) and 95% CIs was run to assess the effects of independent factors on each dimension of EQ-5D indices. In the ordered logistic regression analysis, pre-pregnancy smoking was a dichotomous variable consisting of no smoking behavior (nonsmoker) and quitting smoking during pregnancy (ex-smoker). All tests were 2-sided, and
Because the HRQoL is related to individuals’ perception of their position of life in the context of the culture and value systems in which they live, transnational culture differences will have an obvious impact on the HRQoL. Our study, which was conducted in China, avoided this potential difference [
From August to September 2019, a total of 16,483 participants from 30 provinces were included (
Demographics, EQ-5Da indices, and EQ-VASb scores of pregnant women with different smoking statuses (N=16,483).
Characteristic | Smoker (n=93) | Ex-smoker (n=731) | Nonsmoker (n=15,659) | |
Age (years), mean (SD) | 26.45 (5.43) | 26.18 (5.52) | 28.25 (4.91) | <.001c |
Gestational age (weeks), mean (SD) | 21.17 (8.87) | 20.50 (9.55) | 21.12 (9.09) | .19 |
Spouse smoking, n (%) | 87 (94) | 607 (83.0) | 8758 (56.0) | <.001c |
Disposable income (CN ¥d), mean (SD) | 28,589.01 (8680.59) | 28,247.57 (8126.92) | 29,978.01 (9321.93) | <.001c |
Smoking duration (years), mean (SD) | 19.53 (7.26) | 20.37 (5.84) | —e | .20 |
EQ-5D index, mean (SD) | 0.82 (0.14) | 0.80 (0.12) | 0.80 (0.13) | .16 |
EQ-VAS score, mean (SD) | 77.38 (21.99) | 81.04 (17.68) | 84.49 (14.84) | <.001c |
aEQ-5D: EuroQol Group’s 5-dimension.
bEQ-VAS: EuroQoL Group’s visual analog scale.
c
dA currency exchange rate of CN ¥1= US $0.13971 was applicable per OANDA Rates in September 1, 2019.
eNo result.
Geographical distribution of pregnant women’s EQ-VAS scores of (A) smokers, (B) ex-smokers, and (C) nonsmokers across the 7 administrative regions in mainland China. EQ-VAS: EuroQoL Group’s visual analog scale.
Of the 16,483 participants, 93 (0.56%) were smokers, 731 (4.43%) were ex-smokers, and 15,659 (95%) were nonsmokers (
Smokers, ex-smokers, and nonsmokers had an EQ-5D index of 0.82 (0.14), 0.80 (0.12), and 0.80 (0.13), respectively (
EQ-VAS distribution according to smoking status and IQR. EQ-VAS: EuroQoL Group’s visual analog scale.
EQ-5Da index and EQ-VASb scores among smokers, ex-smokers, and nonsmokers (N=16,483).
Smoking status | Unadjusted value, mean (SD) | Age, per capita disposable income, and spouse smoking status adjusted, mean (SD) | |
|
|||
|
Smoker | 0.82 (0.14) | 0.80 (0.00) |
|
Ex-smoker | 0.8 (0.12) | 0.80 (0.00) |
|
Nonsmoker | 0.8 (0.13) | 0.80 (0.00) |
|
0.16 | 0.82 | |
|
|||
|
Smoker | 77.38 (21.99) | 78.08 (0.13) |
|
Ex-smoker | 81.04 (17.68) | 80.86 (0.05) |
|
Nonsmoker | 84.49 (14.84) | 84.49 (0.01) |
|
<.001c | <.001c |
aEQ-5D: EuroQol Group’s 5-dimension.
bEQ-VAS: EuroQoL Group’s visual analog scale.
c
Pairwise comparisons of pregnant women’s smoking status and EQ-VASa score between groups.
Pairwise groups | Mean difference | SE | 95% CI | |
Smoker vs ex-smoker | 3.66 | 2.37 | –2.09 to 9.41 | .33 |
Smoker vs nonsmoker | 7.11 | 2.28 | 1.56-12.66 | .007b |
Ex-smoker vs nonsmoker | 3.45 | 0.66 | 1.86-5.04 | <.001b |
aEQ-VAS: EuroQoL Group’s visual analog scale.
b
Frequency (%) of the EQ-5Da index of pregnant women with different smoking statuses (N=16,483).
EQ-5D dimension | Smoker, n (%) | Ex-smoker, n (%) | Nonsmoker, n (%) | Total, n (%) | ||||
|
||||||||
|
Level 1 | 75 (81) | 575 (78.7) | 12,116 (77.4) | 12,766 (77.45) | |||
|
Level 2 | 18 (19) | 125 (17.1) | 2969 (19.0) | 3112 (18.88) | |||
|
Level 3 | 0 | 27 (3.7) | 460 (2.9) | 487 (2.95) | |||
|
Level 4 | 0 | 4 (0.6) | 58 (0.4) | 62 (0.38) | |||
|
Level 5 | 0 | 0 (0.0) | 56 (0.4) | 56 (0.34) | |||
|
||||||||
|
Level 1 | 87 (94) | 689 (94.3) | 14,736 (94.1) | 15,512 (94.11) | |||
|
Level 2 | 4 (4) | 38 (5.2) | 843 (5.4) | 885 (5.37) | |||
|
Level 3 | 0 | 1 (0.1) | 58 (0.4) | 59 (0.36) | |||
|
Level 4 | 1 (1) | 2 (0.3) | 10 (0.1) | 13 (0.08) | |||
|
Level 5 | 1 (1) | 1 (0.1) | 12 (0.1) | 14 (0.07) | |||
|
||||||||
|
Level 1 | 80 (86) | 606 (82.9) | 12,460 (79.6) | 13,146 (79.75) | |||
|
Level 2 | 10 (11) | 116 (15.9) | 2875 (18.4) | 3001 (18.21) | |||
|
Level 3 | 0 | 7 (1.0) | 245 (1.6) | 252 (1.53) | |||
|
Level 4 | 1 (1) | 1 (0.1) | 27 (0.2) | 29 (0.18) | |||
|
Level 5 | 2 (2) | 1 (0.1) | 52 (0.3) | 55 (0.33) | |||
|
||||||||
|
Level 1 | 47 (51) | 302 (41.3) | 6836 (43.7) | 7185 (43.59) | |||
|
Level 2 | 42 (45) | 385 (52.7) | 8111 (51.8) | 8538 (51.79) | |||
|
Level 3 | 3 (3) | 33 (4.5) | 627 (4.0) | 663 (4.02) | |||
|
Level 4 | 1 (1) | 10 (1.4) | 68 (0.4) | 79 (0.48) | |||
|
Level 5 | 0 | 1 (0.1) | 17 (0.1) | 18 (0.11) | |||
|
||||||||
|
Level 1 | 39 (41) | 309 (42.3) | 7648 (48.8) | 7996 (48.51) | |||
|
Level 2 | 42 (45) | 346 (47.3) | 7086 (45.3) | 7474 (45.34) | |||
|
Level 3 | 11 (12) | 56 (7.7) | 750 (4.8) | 817 (4.96) | |||
|
Level 4 | 0 | 14 (1.9) | 129 (0.8) | 143 (0.87) | |||
|
Level 5 | 1 (1) | 6 (0.8) | 46 (0.3) | 53 (0.32) |
aEQ-5D: EuroQol Group’s 5-dimension.
Ordered logistic regression analysis for each dimension in the EQ-5Da index of nonsmokers and ex-smokers (n=16,390).
Dimension | ORb (95% CI) | |||
|
||||
|
Age | 1.02 (1.01-1.03) | <.001c | |
Spouse smoking | 1.00 (0.93-1.08) | .93 | ||
Disposable income | 1.00 (1.00-1.00) | .002c | ||
Gestational age | 1.04 (1.03-1.04) | <.001c | ||
Pre-pregnancy smokingd | 0.99 (0.82-1.18) | .88 | ||
|
||||
|
Age | 0.98 (0.97-1.00) | .02c | |
Spouse smoking | 0.79 (0.69-0.90) | <.001c | ||
Disposable income | 1.00 (1.00-1.00) | .15 | ||
Gestational age | 1.08 (1.07-1.09) | <.001c | ||
Pre-pregnancy smokingd | 1.01 (0.73-1.41) | .93 | ||
|
||||
|
Age | 1.01 (1.00-1.02) | .01c | |
Spouse smoking | 0.88 (0.81-1.95) | .001c | ||
Disposable income | 1.00 (1.00-1.00) | 0.37 | ||
Gestational age | 1.03 (1.03-1.04) | <.001c | ||
Pre-pregnancy smokingd | 0.86 (0.70-1.05) | .13 | ||
|
||||
|
Age | 0.98 (0.97-0.99) | <.001c | |
Spouse smoking | 1.00 (0.94-1.06) | .89 | ||
Disposable income | 1.00 (1.00-1.00) | .001c | ||
Gestational age | 1.02 (1.02-1.02) | <.001c | ||
Pre-pregnancy smokingd | 1.09 (0.94-1.27) | .24 | ||
|
||||
|
Age | 0.98 (0.97-0.98) | <.001c | |
Spouse smoking | 1.09 (1.03-1.16) | .006c | ||
Disposable income | 1.00 (1.00-1.00) | .001c | ||
Gestational age | 1.01 (1.00-1.01) | <.001c | ||
Pre-pregnancy smokingd | 1.29 (1.12-1.50) | .001c |
aEQ-5D: EuroQol Group’s 5-dimension.
bOR: odds ratio.
c
dPre-pregnancy smoking is a dichotomous variable consisting of no smoking behavior (nonsmoker, represented as 0) and quitting smoking during pregnancy (ex-smoker, represented as 1).
EQ-5Da index and EQ-VASb score for pregnant ex-smokers (n=731).
|
Mild smoker (n=638) | Moderate smoker (n=79) | Heavy smoker (n=14) | |
EQ-5D index, mean (SD) | 0.80 (0.12) | 0.77 (0.10) | 0.73 (0.16) | .007 |
EQ-VAS, mean (SD) | 81.53 (17.45) | 79.38 (17.82) | 67.93 (22.79) | .01 |
aEQ-5D: EuroQol Group’s 5-dimension.
bEQ-VAS: EuroQoL Group’s visual analog scale.
c
Smoking history (whether before or during pregnancy) is related to a worse HRQoL for of pregnant women. Smoking cessation during pregnancy does not significantly improve pregnant women’s HRQoL. Pre-pregnancy smoking is related to a worse HRQoL (EQ-VAS score) and a higher risk of anxiety/depression problems. In mainland China, pregnant smokers tend to have partners who are smokers. Moreover, the more cigarettes pregnant ex-smokers consume per day, the lower their HRQoL.
We found a few limitations of our study. First, the study was conducted online, so pregnant women without access to the internet were not included. Second, this study divided participants only into 3 groups according to their smoking history. Although we adjusted the impact of age, spouse smoking rate, and disposable income on the HRQoL of participants, there are still many other endogenous factors that can affect pregnant women’s HRQoL (eg, years of schooling, body mass index, chronic disease, abortion history) [
Our results revealed that pregnant women with a smoking history, whether ex-smokers or smokers, have a lower HRQoL (EQ-VAS score) compared to nonsmokers. This result is similar to previous studies that reported that among women, smokers have a lower HRQoL compared to never-smokers [
The significant difference between EQ-VAS scores of ex-smokers and nonsmokers revealed the negative effect of pre-pregnancy smoking on pregnant women’s HRQoL in China. Before our study, few studies have addressed the effect of pre-pregnancy smoking on pregnant women’s health. However, the only study in this field included only 3 months prior to conception as pre-pregnancy smoking, did not explore a wide range of mental issues, and reported that women who smoked during the 3 months prior to conception were more likely to report poor vitality than nonsmokers [
The insignificant EQ-VAS score difference between smokers and ex-smokers revealed that smoking cessation cannot significantly improve the HRQoL of pregnant women in China. This was similar to a previous study that investigated people but not pregnant women and concluded that quitting smoking alone does not improve an individual’s HRQoL [
At the same time, the average EQ-5D index of pregnant women who were ex-smokers was not significantly different from that of other groups. Further analysis of the EQ-5D index revealed that the major problems for pregnant women in China are pain/discomfort problems. For smokers and nonsmokers, the anxiety/depression limitation is the most bothersome problem. Therefore, future policy planning in China should consider pain/discomfort care and mental health care during pregnancy.
Age was related to worsening mobility, which might be due to the decreasing mobility as people get older [
The spouse smoking rate was related to pregnant women’s smoking status, in which the smoker group had the highest spouse smoking rate, the ex-smoker group had the second-highest spouse smoking rate, and the nonsmoker group had the lowest spouse smoking rate. This was similar to a previous study that concluded that smokers are more likely to have partners who smoke [
For ex-smokers, we found that the more cigarettes the women consumed before pregnancy, the lower their HRQoL, which is a new finding. Although a previous study explored the correlation between cigarette number and fetus health, no study has investigated the correlation between the cigarette consumption per day and the HRQoL of pregnant women [
The Banmi Online Maternity School is a free platform for all internet users and serves more than 1 million users in all the 31 provinces/municipalities across mainland China. Basically, it covers all pregnant women regardless of age, occupation, living location, past medical history, individual income, and other individual characteristics, except those who did not use the internet or pay attention to pregnancy care knowledge. Therefore, the characteristics of pregnant women in our study were not different from those of the regular pregnant women in China, except that our study did not include women who could not access the internet or did not pay attention to pregnancy care knowledge. Considering this information, the representativeness of the sample population is high. As of June 2019, the internet penetration rate in China was 61.2%, which was relatively low compared to that of South Korea and Japan, which ranged over 90% [
This study had a large sample size, with a total of 16,483 participants from 31 provinces/municipalities across mainland China. This study filled the gap, as the effect of pre-pregnancy smoking and smoking cessation on pregnant women’s HRQoL was hardly addressed before, especially in China. This study is the first, to date, that horizontally compares pregnant Chinese women’s HRQoL among smoking-before-pregnancy, smoking-during-pregnancy, and never-smoking groups and provides statistical evidence that the more cigarettes pregnant Chinese women consume, the lower their HRQoL. This study revealed that pregnant Chinese women who stop smoking after pregnancy are more likely to suffer from depression or anxiety compared to nonsmokers.
This study systematically explored the effect of the smoking period (whether before or during pregnancy), nicotine source (whether pregnant women themselves or their spouses), and the number of cigarettes consumed on the HRQoL of pregnant women. Smoking cessation during pregnancy does not significantly improve pregnant women’s HRQoL. Pre-pregnancy smoking is related to a better HRQoL (EQ-VAS score). Pre-pregnancy smoking is also related to a higher risk of anxiety/depression problems. The more cigarettes pregnant ex-smokers consume per day, the lower their HRQoL. This study provides scientific guidance for the education of pregnant women and their families about protection of both mother and baby during pregnancy. Although nicotine might benefit pregnant women’s physical health through the pain relief mechanism, its overall harmfulness for pregnant women’s HRQoL (both physical and mental health) should not be neglected. We suggest that women who have labor plans or have already conceived quit smoking and do not resume smoking and avoid an environment with nicotine, especially if their spouses or other family members smoke. However, it is common sense that quitting smoking requires a strong mind and perseverance. Therefore, for those who cannot ban smoking at home, we suggest that they separate smoking family members from pregnant women to reduce the amount of nicotine to which the pregnant women are exposed. This can be achieved by establishing a contemporary smoking room or a pregnant woman room at home.
EuroQol Group’s 5-dimension
EuroQol Group’s 5-dimension 5-level
EuroQoL Group’s visual analog scale
health-related quality of life
odds ratio
The authors would like to gratefully acknowledge the participants, collaborators, and the Banmi Online Maternity School.
KH and SZ contributed equally. KH and W-KM contributed to the study design. KH and SZ analyzed the data and drafted the manuscript. HW contributed to the conduct of work and manuscript writing. SZ, CJPZ, BA, and ZW revised the manuscript. All authors approved the submission of the final manuscript.
None declared.