Abstract
Purpose: The aim of this study was to test the hypothesis that exposure to second-hand smoke (SHS) would be positively associated with major depressive disorder (MDD) in perimenopausal women from a population-based perspective, after adjustment for all potential confounders.
Methods: This study used the National Health and Nutrition Examination Survey (NHANES) database, 2005–2012, to report on MDD in perimenopausal women.
Results: The odds ratio (OR) for MDD increased when there was a smoker was in the home, as compared to not having a smoker in the home (aOR = 2.97, 95% confidence interval [CI] = 1.15–7.67); however, in the non-poor group, the OR for MDD showed no difference between those who had or did not have a smoker in their home. For participants who self-rated their health condition as excellent, very good or good, the OR for MDD increased; it also increased if there were smokers in the home, as compared to those without smokers in the home (aOR = 2.58, 95% CI = 1.08–6.14).
Conclusions: The present study results augment our understanding of the clinical and public health significance of SHS, as well as the role of various socioeconomic and self-rated health conditions, in perimenopausal women.
An increasing OR for MDD was demonstrated with regard to health status such as CVD, chronic respiratory tract disease, arthritis, thyroid problems, lower eGFR, fair or poor self-rated health condition, and elevated CRP level.
Participants who self-rated their health condition as excellent, very good or good had an increased OR for MDD.
The OR also increased if the women had smokers in their home versus women who did not have smokers in the home.
Key messages
Introduction
An increasing number of reports have suggested that smokers are at a higher risk for depression, however, a potential association between exposure to second-hand smoke (SHS) and symptoms of depression remains unclear [Citation1–4]. Among non-smokers, women are more likely than men to be exposed to SHS at home [Citation3]. Furthermore, a school-based survey indicated an association between symptoms of depression and SHS in middle school and high school students [Citation3]. Interestingly, 60% of patients with a history of major depressive disorder (MDD) are current or past smokers, and smokers have increased rates of MDD, as compared with non-smokers [Citation1]. Smoking and SHS have other implications. One study found that 8284 current smokers died from either lung cancer or ischemic heart disease, and another 422 non-smokers who were exposed to SHS died because of lung cancer or ischemic heart disease [Citation3]. A study in China found that prolonged exposure to SES was associated with cognitive decline in women aged 55–64 [Citation5].
Furthermore, several studies have also found associations between SHS and depression in mid-life (age 38–44) women [Citation6], SHS and depression in middle-aged women (40–60) [Citation7], SHS and depression in pregnant women [Citation8], and SHS and post-partum depression [Citation9]. On the other hand, depressive disorders and symptoms are common in middle-aged women, and it is twice as common for women to experience depressive disorders than men [Citation10]. Women are also more likely to experience the symptoms and onset of depression at an earlier age, have symptoms more frequently, and have a higher risk of depression than men [Citation10]. Women are also more likely to experience symptoms of depression during perimenopause than in a premenopausal stage [Citation11–15]. Depression during perimenopause leads to a significant decrease in quality of life [Citation12].
Another study demonstrated that SHS in the work environment had a significant association with depressive symptoms [Citation7]. A few recent studies have shown a positive association between exposure to SHS and depressive symptoms [Citation3,Citation7,Citation16,Citation17]; however, to the best of our knowledge, the impact of SHS on mental health while a woman is perimenopausal has not been thoroughly evaluated. Hence, the aim of this study was to test the hypothesis that exposure to SHS would be positively associated with depression in perimenopausal women from a population-based perspective, even after an adjustment for all potential confounders.
Methods
Data source
This study was a secondary analysis of data from The National Health and Nutrition Examination Survey (NHANES) database (Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention) [Citation18]. These data are organized by subject matter, and were conducted on a nationwide probability sample in the United States. The NHANES data files comprise numerous demographic and socioeconomic variables such as age, gender, race, ethnicity, income, education and marital status [Citation18,Citation19]. Consequently, an evaluation of subjects within the NHANES database is reliable and multidimensional and can be equated to a population-level assessment [Citation19]. Institutional review board approval and informed consent from any subject were not required as all NHANES data are de-identified.
Study population, design and main outcome
This cross-sectional population-based study used data from four cycles of the NHANES database in 2005–2012 to report on MDD in perimenopausal women. The inclusion criteria were women aged 45–55 years who had complete data on the following items: Patient Health Questionnaire-9 (PHQ-9) and serum cotinine measures, as well as a self-report of the existence of smokers in their home, along with numerous potentially key confounders. Women aged 45–55 years are usually considered to be perimenopausal [Citation20–22].
To focus on the relationship between SHS exposure and depression, we excluded all current smokers, who responded “yes” to the question “smoked at least 100 cigarettes in their lifetime and were still smoking,” and participants whose serum cotinine (ng/mL) level was >3 [Citation23,Citation24].
Smokers were classified as a never, former or current smoker using the definitions of Benowitz et al. [Citation24]. Participants were defined as a current smoker if they responded “every day” or “some days” to the question, “Do you smoke now?” Subjects who had smoked <100 cigarettes in their lifetime were defined as non-smokers. Those participants who self-reported that they had smoked ≥100 cigarettes in their lifetime but were not current smokers and had serum cotinine levels of ≤3 ng/mL were considered former smokers. Only former and non-smokers were included within the study population.
Definitions of study variables
Major depressive disorder
The PHQ-9 was used to measure depressive symptoms, as it has been shown to be a reliable and valid questionnaire in community samples; moreover, it has high internal consistency and good sensitivity and specificity for identifying cases of MDD [Citation25–27]. Participants responded to the nine items in the PHQ-9 that measured depressive symptoms during the previous 2 weeks. Participants responded to questions on a four-point scale: 0, not at all; 1, on several days; 2, on more than half the day; or 3, nearly every day. The score for each item was added, and the sum was from 0 to 27 that was used to classify depressive symptoms. Participants whose PHQ‐9 scores were ≥10 were considered as having MDD.
Serum cotinine
The serum cotinine concentration was used to measure the extent of environmental tobacco smoke exposure [Citation28]. Serum cotinine concentrations were obtained from NHANES laboratory data. NHANES serum cotinine is measured using isotope dilution-high performance liquid chromatography together with atmospheric pressure chemical ionization tandem mass spectrometry, according to NHANES quality assurance and quality control (QA/QC) protocols [Citation28].
Menopausal status
Participants were considered menopausal if they responded “no” to the question “Have you had at least one menstrual period in the past 12 months?” Those who responded “yes” were considered to not be menopausal.
Demographic variables
Socio-demographic factors [Citation29], poverty [Citation30], self-rated health condition [Citation31] and menopause [Citation32] were all subject to sub-group analysis. Potential confounding variables were obtained from the demographic questionnaire data for each participant, including age, race/ethnicity, family poverty income ratio (shown as poor or non-poor according to the poverty income ratio <1 or ≥1), married/living with partner and occupation.
The term, smokers in the home, was defined as those who responded affirmatively to the following question: “Does anyone who lives here smoke cigarettes, cigars or pipes anywhere inside this home?”
Heavy drinkers were defined as those who indicated that they drank ≥4 times/week to the following question: “In the past 12 months, how often did you drink any type of alcoholic beverage?”
In this study, we used the definition of Ford et al. and the US Department of Health and Human Services for physical activity that was defined as the sum of the product of the time spent weekly in each activity reported by the participant multiplied by the metabolic equivalent of task (MET) value for that activity yielding a MET-h index [Citation33,Citation34]. One MET was the energy expenditure for every 1 kcal/kg of body weight per hour. Participants who had MET-min/week of ≥500 were referred to as physically active, but participants with <500 MET-min/week were referred to as non-active.
Comorbidity and health status
Multiple comorbid conditions have been associated with depression in later life in both men and women [Citation35,Citation36]. Obesity was defined according to the standard body mass index (BMI) category. The BMI value was obtained from the NHANES laboratory measurements. Participants who had BMI ≥30 kg/m2 were referred to as obese, and those participants who had a BMI between ≥25 and <30 were referred to as overweight. Those with a BMI <25 were considered the underweight/normal weight group.
Dyslipidemia was self-reported and was indicated if the participant was told to take a blood lipid-lowering prescription or who had one of the following laboratory measurement data files: total serum cholesterol >200 mg/dL; serum triglyceride >200 mg/dL; or serum HDL <40 mg/dL.
Glomerular filtration rate (eGFR) was estimated from serum creatinine by the Modification of Diet in Renal Disease (MDRD) study equation: eGFR = 175 × ([calibrated serum creatinine in mg/dl] − 1.154) × age −0.203 × (0.742 if female) × (1.210 if African-American). Participants were grouped into an eGFR group of <60 or ≥60 mL/min/1.73 m2.
C-reactive protein (CRP) values were obtained from laboratory information in the NHANES dataset. A CRP level higher than 1 mg/dL was defined as elevated. Serum vitamin D levels (serum 25-hydroxyvitamin D level) were obtained from the NHANES laboratory data. A detailed description of the laboratory procedure manual can be found on the NHANES web page [Citation34]. For the present study, the cutoff values for serum vitamin D were defined as follows: >75 nmol/L as normal; 50–75 nmol/dL for insufficient; and <50 nmol/L for deficient as in previous studies.
Diabetes was self-reported and defined as having been told by a doctor or health care professional that the participant has diabetes or sugar diabetes, or they were currently taking diabetic pills or insulin. Cardiovascular diseases (CVD; including coronary heart disease, angina, congestive heart failure, myocardial infarction and stroke), cancer history, chronic respiratory tract disease (including emphysema, chronic bronchitis and asthma), arthritis and thyroid problems were defined as those who responded affirmatively to the following question: “Has a doctor or other health care professional ever told you that you have (disease).” Self-rated health condition provided personal interview data on an overall health assessment and was dichotomized into good (excellent, very good or good) and poor (fair or poor).
Statistical analysis
All statistical analyses were performed by considering the survey data structure. Categorical variables were summarized as frequency and weighted percentage, and continuous variables were summarized as the mean and standardized error of the mean. Analysis of variance was used to test differences in continuous variables, and the logistic regression model was performed for categorical variables. Logistic regression model considering the survey data structure was conducted to estimate the odds ratio (OR) for each factor with regard to MDD. Factors other than serum cotinine and smokers in the home were significant in the univariate logistic regression model and were further assessed in the multiple regression model for adjusting OR for MDD. All analyses were performed using SAS software (v. 9.4; SAS Institute Inc., Cary, NC). A two-tailed p-value of <.05 was considered significant.
Results
The study population included 2590 perimenopausal women. There were 178 in the MDD group and 2412 in the non-MDD group. (; ) More women in the non-MDD group had a higher concentration of serum cotinine (≥90 percentile, p = .007) and smokers in the home (p = .010). Distributions by race (p = .002), poverty income ratio (p < .001), education (p < .001), marital status (p < .001), occupation (p < .001), cardiovascular disease (p < .001), chronic respiratory tract disease (p < .001), arthritis (p < .001), thyroid problem (p < .001), eGFR (p = .005), self-rated health conditions (p < .001) and CRP level (p < .001) were significantly different between the two groups ().
Figure 1. Study flowchart of the population of perimenopausal women.
Table 1. Distribution of the characteristics of perimenopausal women, 2005–2012.
Crude OR for MDD were higher for women with a serum cotinine level of ≥90 percentile (OR = 1.96, 95% CI = 1.20–3.18) and with a smoker in the home (OR = 2.46, 95% CI = 1.24–4.88). Poverty, less education, and living alone also had a higher association with MDD. Health status such as CVD (OR = 4.39, p < .001), chronic respiratory tract disease (OR = 2.65, p < .001), arthritis (OR = 2.48, p < .001), thyroid problems (OR = 2.77, p < .001), lower eGFR (OR = 2.65, p = .004), fair or poor self-rated health condition (OR = 10.41, p < .001), and elevated CRP level (OR = 2.71, p < .001) showed an increased OR for MDD ().
Table 2. Odds ratio for major depressive disorder in perimenopausal women, 2005–2012 by univariate logistic regression model.
After adjusting the model for race, poverty income ratio, education, marital status, CVD, chronic respiratory tract disease, arthritis, thyroid problems, eGFR, self-rated health problems and CRP level, there was a slight, but not significant, positive association between having a smoker in the home and MDD. To clarify their importance, subgroup analysis was conducted for the variables of poverty income ratio, self-rated health conditions, smoking history and menopause. In the subgroup analysis, the OR for MDD increased for poor women who had a smoker in the home, as compared to those women who had no smokers in the home (aOR = 2.97, 95% CI = 1.15–7.67). As for participants who self-rated their health condition as excellent, very good, or good, the OR for MDD increased, and it also increased if there were smokers in the home, as compared with no smokers in the home (aOR = 2.58, 95% CI = 1.08–6.14) ().
Table 3. Odds ratio for major depressive disorder in perimenopausal women using a multiple logistic regression model, 2005–2012.
Discussion
This study was conducted to test the hypothesis that SHS exposure would be positively associated with depression in perimenopausal women from a population-based perspective, despite adjustments for all potential confounders. Crude ORs for MDD were greater for women with a serum cotinine level of ≥90 percentile (OR = 1.96, 95% CI = 1.20–3.18) and with a smoker in the home (OR = 2.46, 95% CI = 1.24–4.88). An increased OR for MDD was demonstrated with regard to health status such as CVD (OR = 4.39, p < .001), chronic respiratory tract disease (OR = 2.65, p < .001), arthritis (OR = 2.48, p < .001), thyroid problems (OR = 2.77, p < .001), lower eGFR (OR = 2.65, p = .004), fair or poor self-rated health condition (OR = 10.41, p < .001) and elevated CRP level (OR = 2.71, p < .001). Further, a slight, but insignificant, positive relationship was observed between women who had a smoker in the home and MDD. Interestingly, participants who self-rated their health condition as excellent, very good, or good had an increased OR for MDD. The OR also increased if the women had smokers in their home versus women who did not have smokers in the home (aOR = 2.58, 95% CI = 1.08–6.14).
Women who never had premenopausal depression had an increased risk of psychiatric disorders during the transitional period to menopause [Citation10]. Furthermore, perimenopausal women had a four times greater likelihood of clinically apparent symptoms of depression and a two and one-half times greater likelihood of an MDD diagnosis, as compared with when they were premenopausal [Citation10]. Moreover, Steinberg et al. [Citation37] investigated the plasma hormone levels, symptoms, and reproductive status of 116 women and demonstrated that incidents of depression were grouped later during the menopausal transition and first year after menopause.
Other studies have found an increased association with SHS and MDD. Kim et al. [Citation38] demonstrated that exposure to SHS in the home was associated with depression and stress. A meta-analysis performed by Zeng and Li [Citation39] on a total of eleven studies involving 86,739 participants concluded that exposure to SHS was related to depressive symptoms and psychological distress in non-smokers. Another study conducted in Spain of 21,007 adults, of whom 11,214 never smoked, demonstrated that SHS exposure at home was related to psychological distress [Citation40]. In the United Kingdom, however, a study found an insignificant relationship between SHS exposure and poor mental health status [Citation41]. Additionally, a study of Chinese-American smokers had symptoms of depression similar to other US populations; moreover, women and the unemployed had more distinctive symptoms of depression [Citation42]. Major depressive incidents were found to be associated with SHS in a general Canadian population [Citation43].
In the current study, the OR for MDD increased when there was a smoker in the home, as compared to not having s smoker in the home; however, in those with a poverty income ratio ≥1, the OR for MDD demonstrated no difference between those who have and those who did not have a smoker in their home. Another study using NHANES data found that black women who were poor and smoked had higher risk of MDD symptoms [Citation44], although this study looked only at smoking status, not exposure to the smoke of others. Another US study sought to determine the temporal relationship between smoking and MDD in low-income black women, which found that both smokers and ex-smokers had higher levels of MDD than women who had never smoked [Citation45]. Not only did they find that beginning smoking preceded MDD symptoms, but each year of delay in starting smoking lowered the odds of MDD symptoms by 8.2%. Smoking within a household with low income may be an early reaction to stressors which later trigger MDD or other mental health conditions.
Another study by Alagiyawanna et al. [Citation46] conducted in Sri Lanka for 30 groups of 25 women from 750 households who were divided into control and intervention groups to assess the impact of educational activities focused on increasing awareness of the detriments and effects of SHS, as well as on the right to a smoke-free living environment, concluded that these interventional educational sessions decreased SHS exposure in the home. Wang et al. conducted a study in Jilin, China, and found that the prevalence of mental distress among non-smokers was 24.5% and that the prevalence of SHS exposure among this group of non-smokers with mental distress was 65.0% [Citation47]. In the United States, Goodwin et al. [Citation48] found that, over the previous decade, smoking durig pregnancy was four times more common for women with MDD than for women without MDD.
A study using the 1999–2010 NHANES data of 24,791 participants assessed SHS and found that serum cotinine levels were as follows: no exposure to SHS geometric mean followed by CI, respectively, of 0.047 (0.044–0.050) ng/mL; exposure to SHS at work only 0.055 (0.047–0.064) ng/mL; exposure to SHS at home only 0.522 (0.401–0.678) ng/mL; and exposure to SHS at home and at work 0.485 (0.280–0.0840) ng/mL [Citation49]. In the present study, a higher concentration of serum cotinine was found in women in the non-MDD group than in the MDD group (≥90 percentile, p = .007).
Despite the strengths of this study using a NHANES database indicative of population level, this study had some limitations. First, this assessment was a cross-sectional analysis, thus limiting the ability to draw any inferences regarding causality, and the onset and progression of depression could not be determined. Second, non-menopausal women were not included in the present study, so a comparison between perimenopausal and non-menopausal women has not been explored. Third, the interview (questionnaire) data were based on self-reports and, therefore, could be subject to recall bias and misunderstanding of the question, as well as a variety of other factors. The NHANES database is made up of data that are representative of the proportions of the different ethnic, socioeconomic status, feelings about self-rated health conditions within the United States and thus should be validated in other countries. As such, additional investigation and well-designed longitudinal studies are needed to confirm the findings.
The present study results augment our understanding of the clinical and public health significance of SHS, as well as the various socioeconomic and self-rated health conditions of women that affect the occurrence of depression in perimenopausal women. This information may help to create more focused interventions directed at improving women’s health.
Acknowledgments
The authors acknowledge the efforts of the US National Center for Health Statistics (NCHS) for the creation of the National Health and Nutrition Examination Survey Data. The interpretation and reporting of these data are the sole responsibility of the authors.
Disclosure statement
No potential conflict of interest was reported by the authors.
References
- Bakhshaie J, Zvolensky MJ, Goodwin RD. Cigarette smoking and the onset and persistence of depression among adults in the United States: 1994–2005. Compr Psychiatry. 2015;60:142–148.
- Mendelsohn C. Smoking and depression - a review. Aust Fam Physician. 2012;41:304–307.
- Jung SJ, Shin A, Kang D. Active smoking and exposure to secondhand smoke and their relationship to depressive symptoms in the Korea national health and nutrition examination survey (KNHANES). BMC Public Health. 2015;15:1053.
- Taha F, Goodwin RD. Secondhand smoke exposure across the life course and the risk of adult-onset depression and anxiety disorder. J Affect Disord. 2014;168:367–372.
- Pan X, Luo Y, Roberts AR. Secondhand smoke and women’s cognitive function in China. Am J Epidemiol. 2018;187:911–918.
- Elmasry H, Goodwin RD, Terry MB, et al. Early life exposure to cigarette smoke and depressive symptoms among women in midlife. Nicotine Tob Res. 2014;16:1298–1306.
- Ye X, Li L, Gao Y, et al. Dose-response relations between second-hand smoke exposure and depressive symptoms among middle-aged women. Psychiatry Res. 2015;229:533–538.
- Huang J, Wen G, Yang W, et al. The association between second-hand smoke exposure and depressive symptoms among pregnant women. Psychiatry Res. 2017;256:469–474.
- Khan S, Arif AA, Laditka JN, et al. Prenatal exposure to secondhand smoke may increase the risk of postpartum depressive symptoms. J Public Health (Oxf). 2015;37:406–411.
- Llaneza P, Garcia-Portilla MP, Llaneza-Suarez D, et al. Depressive disorders and the menopause transition. Maturitas. 2012;71:120–130.
- de Kruif M, Spijker AT, Molendijk ML. Depression during the perimenopause: a meta-analysis. J Affect Disord. 2016;206:174–180.
- Soares CN, Zitek B. Reproductive hormone sensitivity and risk for depression across the female life cycle: a continuum of vulnerability? J Psychiatry Neurosci. 2008;33:331–343.
- Wariso BA, Guerrieri GM, Thompson K, et al. Depression during the menopause transition: impact on quality of life, social adjustment, and disability. Arch Womens Ment Health. 2017;20:273–282.
- Zsido RG, Villringer A, Sacher J. Using position emission tomography to investigate hormone-mediated neurochemical changes across the female lifespan: implications for depression. Int Rev Psychiatry. 2017;29:580–596.
- Mulhall S, Andel R, Anstey KJ. Variation in symptoms of depression and anxiety in midlife women by menopausal status. Maturitas. 2018;108:7–12.
- Kim NH, Kim HC, Lee JY, et al. Association between environmental tobacco smoke and depression among Korean women. BMJ Open. 2015;5:e007131.
- Bandiera FC, Arheart KL, Caban-Martinez AJ, et al. Secondhand smoke exposure and depressive symptoms. Psychosom Med. 2010;72:68–72.
- National Health and Nutrition Examination Survey. Available from: http://www.cdc.gov/nchs/nhanes/
- Zipf G, Chiappa M, Porter KS, et al. National health and nutrition examination survey: plan and operations, 1999–2010. Vital Health Stat 1. 2013;1–37.
- Brett KM, Cooper GS. Associations with menopause and menopausal transition in a nationally representative US sample. Maturitas. 2003;45:89–97.
- Richard A, Rohrmann S, Mohler-Kuo M, et al. Urinary phytoestrogens and depression in perimenopausal US women: NHANES 2005–2008. J Affect Disord. 2014;156:200–205.
- Kozinoga M, Majchrzycki M, Piotrowska S. Low back pain in women before and after menopause. Prz Menopauzalny. 2015;14:203–207.
- Kim S. Overview of cotinine cutoff values for smoking status classification. Int J Environ Res Public Health. 2016;13:1236.
- Benowitz NL, Bernert JT, Caraballo RS, et al. Optimal serum cotinine levels for distinguishing cigarette smokers and nonsmokers within different racial/ethnic groups in the United States between 1999 and 2004. Am J Epidemiol. 2009;169:236–248.
- Kroenke K, Spitzer RL. The PHQ-9: a new depression and diagnostic severity measure. Psychiatr Ann. 2002;32:509–521.
- Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16:606–613.
- Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA 1999;282:1737–1744.
- National Health and Nutritional Examination Survey 2003-2004. Data Documentation, Codebook, and Frequencies. Available from: https://wwwn.cdc.gov/Nchs/Nhanes/2003-2004/L06BMT_C.htm
- Holma IA, Holma KM, Melartin TK, et al. Depression and smoking: a 5-year prospective study of patients with major depressive disorder. Depress Anxiety. 2013;30:580–588.
- Baiyewu O, Yusuf AJ, Ogundele A. Depression in elderly people living in rural Nigeria and its association with perceived health, poverty, and social network. Int Psychogeriatr. 2015;27:2009–2015.
- Chen Y, While AE, Hicks A. Self-rated health and associated factors among older people living alone in Shanghai. Geriatr Gerontol Int. 2015;15:457–464.
- Vivian-Taylor J, Hickey M. Menopause and depression: is there a link?. Maturitas. 2014;79:142–146.
- Ford ES, Wheaton AG, Chapman DP, et al. Associations between self-reported sleep duration and sleeping disorder with concentrations of fasting and 2-h glucose, insulin, and glycosylated hemoglobin among adults without diagnosed diabetes. J Diabetes. 2014;6:338–350.
- National Health and Nutritional Examination Survey 2005–2006. Data, documentation, codebook, and frequencies. Available from: https://wwwn.cdc.gov/Nchs/Nhanes/2005-2006/VID_D.htm#LBDVIDMS
- Heeres RH, Hoogeveen EK, Geleijnse JM, et al. Kidney dysfunction, systematic inflamation and mental well-being in elderly post-myocardial infarction patients. BMC Psychol. 2015;5:1.
- Kim WK, Shin D, Song WO. Depression and its comorbid conditions more serious in women than in men in the United States. J Womens Health (Larchmt). 2015;24:978–985.
- Steinberg EM, Rubinow DR, Bartko JJ, et al. A cross-sectional evaluation of perimenopausal depression. J Clin Psychiatry. 2008;69:973–980.
- Kim NH, Choi H, Kim NR, et al. Secondhand smoke exposure and mental health problems in Korean adults. Epidemiol Health. 2016;38:e2016009.
- Zeng YN, Li YM. Secondhand smoke exposure and mental health in adults: a meta-analysis of cross-sectional studies. Soc Psychiatry Psychiatr Epidemiol. 2016;51:1339–1348.
- Ballbe M, Martinez-Sanchez JM, Gual A, et al. Association of second-hand smoke exposure at home with psychological distress in the Spanish adult population. Addict Behav. 2015;50:84–88.
- Lam E, Kvaavik E, Hamer M, et al. Association of secondhand smoke exposure with mental health in men and women: cross-sectional and prospective analyses using the U.K. Health and Lifestyle Survey. Eur Psychiatry. 2013;28:276–281.
- Tsoh JY, Lam JN, Delucchi KL, et al. Smoking and depression in Chinese Americans. Am J Med Sci. 2003;326:187–191.
- Patten SB, Williams JVA, Lavorato DH, et al. Major depression and secondhand smoke exposure. J Affect Disord. 2018;225:260–264.
- Amutah-Onukagha NN, Doamekpor LA, Gardner M. An examination of the sociodemographic and health determinants of major depressive disorder among black women. J Racial Ethn Health Disparities. 2017;4:1074–1082.
- Scarinci IC, Thomas J, Brantley PJ, et al. Examination of the temporal relationship between smoking and major depressive disorder among low-income women in public primary care clinics. Am J Health Promot. 2002;16:323–330.
- Alagiyawanna A, Rajapaksa-Hewageegana N, Gunawardena N. The impact of multiple interventions to reduce household exposure to second-hand tobacco smoke among women: a cluster randomized controlled trial in Kalutara district, Sri Lanka. BMC Public Health. 2017;17:810.
- Wang R, Zhang P, Lv X, et al. Association between passive smoking and mental distress in adult never-smokers: a cross-sectional study. BMJ Open. 2016;6:e011671.
- Goodwin RD, Cheslack-Postava K, Nelson DB, et al. Smoking during pregnancy in the United States, 2005–2014: the role of depression. Drug Alcohol Depend. 2017;179:159–166.
- Jain RB. Exposure to second hand smoke at home and work among nonsmokers. Chemosphere. 2015;135:225–232.