Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan.

The association of socioeconomic status and lifestyle behaviours on mental health appears well-established in the literature, as several studies report that better socioeconomic status such as higher levels of disposable income and employment as well as practising healthy lifestyles can enhance mental well-being. However, the reliance on cross-sectional correlations and lack of adequate statistical controls are possible limitations. This study aims to add the evidence of longitudinal association to the literature by using Japanese representative longitudinal household panel data. We employed panel data analytical techniques such as the random-effects conditional logistic regression (RE-CLR) and the fixed-effects conditional logistic regression (FE-CLR) models with possible time variant confounders being controlled. Our sample was comprised of 14,717 observations of 3,501 individuals aged 22-59 years for five waves of the Japanese Household Panel Survey. We confirmed many of the factors associated with mental health reported in existing studies by analysing cross-sectional data. These significant associations are also longitudinal (within) associations estimated by the FE-CLR models. Such factors include unemployment, low household income, short nightly sleeping duration, and lack of exercise. However, we also found that several factors such as disposable income, living alone, and drinking habits are not significantly associated with mental health in the FE-CRL models. The results imply the reverse causality that poor mental health conditions cause lower disposal income, possibly due to the inability to exhibit higher productivity, but an increase in disposal income would not necessarily improve mental health conditions. In this case, aggressive policy interventions to increase the disposal income of people of lower socioeconomic backgrounds would not necessarily be effective to minimize health inequalities.

Introduction

Health inequalities have been receiving significant attention in many countries [1, 2]. The World Health Organisation has recently appealed to governments worldwide with this phrase, ‘Health inequalities are unfair,’ which served as a worldwide call to action to minimise health inequalities through governmental policies [3]. Subsequent studies have found that differences in socioeconomic status (SES) are one of the causes of health inequalities and are associated with health outcomes [2, 4, 5]. These studies usually examined participants’ income, educational levels, occupation and employment status which are the socioeconomic factors that have been proven to be main determinants related to individuals’ mental and physical health [1, 6–8]https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3783534/ - R36.

Mental illness has become a notable public health concern on a global scale, and has been directly related to physical disorders and suicides among those affected [9]. A previous study reported that mental illness and SES such as low income and unemployment were significantly associated with higher risks for committing suicide [5]. Japan was shown to be one of the countries with the highest suicide rates in the world [10, 11]; thus, it is important to investigate any possible associations between suicide and mental health, SES, and lifestyle-related factors [12].

Cross-sectional studies have shown that healthy lifestyle factors are positively associated with better mental health outcomes [13]. In that regard, the relationship between mental health conditions and lifestyle factors such as sleeping duration [14], habitual physical exercise [15], smoking habits [16, 17], and alcohol consumption [18] have been studied to find methods of preventing mental illness. For example, Golzier et al. analysed a cohort study and reported that shorter sleep duration is linearly associated with psychological distress [14]. The cross-sectional study and its follow-up study reported that habitual physical exercise was associated with better health outcomes [15, 19]. Additional cross-sectional studies found that smoking and alcohol problems are positively associated with depressive symptoms [16, 18]. These studies suggest that increasing healthy lifestyle practises may promote better health outcomes. However, the main limitation of these findings is that confounding remains possible when the study design was cross-sectional, unless all studies adjusted variables.

Deepening this discussion, there appear to be connections among lifestyle practises, mental disorders, SES, and poorer health conditions [20]. In particular, mental disorders are frequently associated with people with lower SES [5, 21]. Some studies report that people with low SES are more likely to practise unhealthy behaviours than those with high SES [4, 22]. For example, people with lower SES are more likely to be smokers, so they suffer from the effects of smoking [23]. This may help explain the background of income-related health inequality, and so it is important to determine causal links between poor health conditions and low levels of SES. Furthermore, it has been essential to discuss the probable causality running in both directions: poor SES may promote poor mental health conditions, and vice versa [24].

Therefore, instead of examining cross-sectional data, we analysed the panel (longitudinal) dataset by using random-effects and fixed-effects longitudinal regression models to avoid the potential bias from unobserved confounders. The random-effects models do not control for unobservable and time-invariant individual attributes, so the estimated results of these models should resemble those of cross-sectional studies. However, the fixed-effects models are able to control for unobservable and person-specific time-invariant heterogeneity and reverse causality caused by time-invariant factors, so the estimated results of these models may not resemble those of cross-sectional studies [25]. For example, some people may be resilient and cope with stress innately, but some may not. If such an unobservable heterogeneity factor is correlated with explanatory variables such as SES and lifestyle-related factors, the estimates from a cross-sectional regression or random-effects model would become inconsistent due to the omitted variable bias. Furthermore, in such cases, it often becomes impossible to interpret the causal relationships from the estimated results. However, the use of fixed-effects models often allows for the interpretation of the estimated results as reflections of a causal relationship as far as the time-invariant unobservable heterogeneity factor brings about reverse causality. Notwithstanding, it should be noted that, when performing this type of analysis, it is also important to use instrument variables to address the reverse causality caused by time-variant factors. However, it is not easy to find appropriate instrument variables that are dependent on explanatory variables but independent of dependent variables. We have not been able to find appropriate instrument variables for this study, but we suggest that future research should seek to use them to overcome the possible reverse causality related to time-variant factors.

To summarize, we utilised fixed-effects models to control time-invariant factors, so as to help in the identification of causal relationships and to eliminate possible biases, thereby leading to more consistent estimated results by analysing longitudinal panel data [26]. This study aims to build on the cross-sectional research cited which suggest positive associations between SES and lifestyle-related factors on participants’ mental health conditions, even after controlling for unobservable person-specific time-invariant individual attributes in the fixed-effects model. The objectives of this study are [1] to examine the prevalence of poor mental health conditions among Japanese individuals aged 22 to 59 years and [2] to analyse the differences between the estimates from random- and fixed-effects models, so as to help identify the SES and lifestyle-related risk factors for poor mental health conditions. Our findings may provide further evidence that incentivising people to practise healthier lifestyles may benefit mental health.

Methods

For this study, we used longitudinal panel data from the Japan Household Panel Survey (JHPS/KHPS). The JHPS/KHPS was developed by the Panel Data Research Centre at Keio University, Japan. This panel dataset is valuable because it has been providing nationally representative samples that allow for the study of Japanese individuals’ conditions related to their socioeconomic status, income and poverty dynamics, disparities in health quality, and health behaviours for the past two decades.

Sampling of the respondents

The JHPS/KHPS utilises a two-stage stratified random sampling method. According to the National Census Survey, Japan was stratified into 24 levels based on regional and city classifications. The number of samples for each level were calculated by using basic resident register population ratios with a range of 5 to 10 samples selected for each level.

The KHPS sample began with 4,005 respondents in 2004, and 1,400 and 1,000 respondents were added in 2007 and 2012, respectively. Subsequently, after being renamed JHPS, the sample began with 4,022 people in 2009. From 2004 to 2017, a total of 10,458 respondents were selected at random from the basic resident register system in Japan. It should be noted that the number of respondents gradually declined during the long study period. The JHPS/KHPS survey was carried out every year from February to March, and the questionnaire was distributed to respondents who had participated the previous year. This study utilised a 5-wave dataset that was collected from 2014 to 2018. In Japan, public pension benefits begin at the age of 60 to 65 years, so the socioeconomic status, such as disposable income, for this specific age group of the population is different from the other age groups. Thus, 10,185 observations of 2,204 individuals were excluded because the ages of the respondents were not between 22 and 59 years during the corresponding years of their possible participation. In total, we used data from 14,717 observations of 3,501 individuals aged 22 to 59. Table 1 details the number of respondents and response rates for each survey and the number of recruited participants from each survey for this study. All respondents received an informed consent form about the aims and details of the study, which also provided information about the anonymity and confidentiality of the replies. Before receiving the questionnaire, those selected were asked to participate in the study, and after agreeing, the questionnaire was sent or brought to their home by research assistants. Signed written consent by the participants was obtained for the study.

Table 1

The number of respondents and response rates for each survey and the number of recruited participants from each survey.

JHPS					KHPS					Total
Wave	Survey year	Number of respondents in total1)	Response rate (%)	Respondents in this study	Wave	Survey year	Number of respondents in total	Response rate (%)	Respondents in this study
6	2014	2,358	91.1	1,385	11	2014	3,312	92.6	2,102	3,487
7	2015	2,198	93.0	1,253	12	2015	3,124	98.8	1,939	3,192
8	2016	2,048	92.8	1,144	13	2016	2,945	94.0	1,776	2,920
9	2017	1,885	91.9	1,057	14	2017	2,741	92.7	1,628	2,685
10	2018	1,742	92.2	970	15	2018	2,549	93.0	1,463	2,433
Total		10,231	92.2	5,809	Total		14,671	94.2	8,908	14,717

1) Number of respondents in total: the number of collected questionnaires which were completed by the respondents.

Study variables

A self-administered questionnaire collected participants’ information related to socioeconomic factors, lifestyle-related factors, and mental health outcomes.

Socioeconomic factors

To assess participants’ SES, via the questionnaire, they were asked about their gender, age, employment status, the number of persons in the household (living status), and disposable income per household. Participants included only those aged 22 to 59 years at the time of participation. Employment status was divided into four groups: unemployed, self-employed, regular employee, and non-regular employee. The number of persons in the household was categorised into two groups: living with someone (≥ 2 people) and living alone. All participants were asked for their annual household disposable income for the year prior to participation, which included disposable incomes of all household members excluding tax and social insurance fees. The disposable income per household was divided into three groups: low level (< 2,000K yen), middle level (2,000K–6,000K yen), and high level (≥ 6,000K yen).

Lifestyle factors

To assess participants’ lifestyle practices, via the questionnaire, they were asked about their sleep duration during the week, physical exercise frequency, and smoking and drinking alcohol habits. For sleep duration, the question used was ‘How many hours do you usually sleep each weekday night?’ The responses were categorised into three groups: ≥ 7 hours, 6–7 hours, and < 6 hours. Regarding physical exercise frequency, the question used was: ‘Excluding work-related activities, how many days per week do you perform physical exercise in which you sweat?’ The answers were categorised into three groups: ≥ 3 days/week, ≤ 2 days/week, and no exercise. Regarding smoking habits, the question used was ‘Do you smoke?’ Participants were categorised as never-smokers, ex-smokers, and current smokers who smoked sometimes/every day. Drinking habits were assessed by the question: ‘How often do you drink alcohol?’ The answers were categorised into the following groups: Never, ≤ 2 times/week, ≥ 3 times/week.

Health outcomes

Participants’ mental health status was measured by the General Health Questionnaire 12-items (GHQ-12) [27, 28], written in the Japanese language. This questionnaire served as a screening measure to detect nonpsychotic psychiatric diseases, and it was comprised of 12 questions about participants’ feelings over the previous few weeks. The questions included the following: Have you recently (1) been able to concentrate on whatever you’re doing, (2) lost much sleep over worry, (3) felt that you were playing a useful part in general, (4) felt capable of making decisions, (5) felt constantly under strain, (6) felt you couldn’t overcome your difficulties, (7) been able to enjoy your normal day-to-day activities, (8) been able to face up to problems, (9) been feeling unhappy or depressed, (10) been losing confidence in yourself, (11) been thinking of yourself as a worthless person, (12) been feeling reasonably happy, all things considered. In order to assess the severity of participants’ psychological distress, we utilised a scoring system described herein: the response categories (1, 2, 3, and 4) were converted into corresponding binary values (0, 0, 1, and 1) to calculate the total score of the 12 questions. The subjects were then divided into two groups: those with higher scores/poor mental health conditions: ≥ 4 points; and those with lower scores/good mental health conditions: ≤ 3 points [27].

Statistical methods

Participants’ demographic characteristics were analysed through mean and standard deviations values and percentages. Further, we analysed the association between socioeconomic and lifestyle factors by gender through adjusted prevalence odds ratios, and utilised 95% confidence intervals (95% CI) of participants’ scores (0: ≤ 3 points, 1: ≥ 4 points) from GHQ-12 by using both the random-effects conditional logistic regression (RE-CLR) and the fixed-effects conditional logistic regression (FE-CLR) models. These models are commonly used to analyse panel data [29]. Further, applying these two methods can help eliminate bias and improve consistency of the results [29], as well as help estimate the longitudinal association between socioeconomic- and lifestyle-related factors on participants’ mental health outcomes. We further evaluated the models by using the Hausman specification test to identify the better model for each analysis [29]. We controlled the results for all factors: age, number of persons in the household, employment status, annual disposable income per household, and lifestyle-related factors. Data was analysed separately by gender because some studies reported that the prevalence of mental health conditions and lifestyle factors were different between men and women [30, 31]. The statistical analysis was completed by using the SPSS 25.0 and STATA MP 15 computer package.

The Institutional Review Board, Institute for Economic Studies, Keio University approved this study (Reference number 15002).

Results

This study used a 5-wave panel dataset and analysed data for a total of 14,717 participants including 7,215 men (49.0%) and 7,502 women (51.0%). Participants’ descriptive statistics are detailed in Table 2.

Table 2

Participants’ demographic characteristics.

		Total		Men		Women
Variables	Group	N/mean	%/SD	N/mean	%/SD	N/mean	%/SD
Wave
	Men	7,215	49.0%
	Women	7,502	51.0%
Age (in years)		45.3	8.7	45.2	8.7	45.4	8.7
Number of persons in the household
	≥ 2 people	13,487	92.1%	6,415	89.4%	7,072	94.7%
	One person	1,155	7.9%	757	10.6%	398	5.3%
Employment status
	Unemployed	1,885	12.9%	271	3.8%	1,614	21.7%
	Self-employed	1,818	12.4%	1,106	15.4%	712	9.6%
	Regular employee	7,126	48.7%	5,332	74.3%	1,794	24.1%
	Non-regular employee	3,796	26.0%	466	6.5%	3,330	44.7%
Disposable income per household
	≥ 6,000K	4,856	36.9%	2,424	37.2%	2,432	36.6%
	2,000K-< 6,000K	7,373	56.0%	3,720	57.1%	3,653	54.9%
	< 2,000K	935	7.1%	370	5.7%	565	8.5%
	in Japanese Yen	542.8	311.3	542.2	295.9	543.3	325.6
Sleep duration. weekdays
	≥ 7 hours	5,450	37.6%	2,710	38.2%	2,740	37.0%
	6–7 hours	5,595	38.6%	2,769	39.1%	2,826	38.2%
	< 6 hours	3,448	23.8%	1,607	22.7%	1,841	24.9%
Physical exercise
	≥ 3 days/week	1,415	9.7%	729	10.2%	686	9.2%
	≤ 2 days/week	2,637	18.0%	1,518	21.2%	1,119	15.0%
	No exercise	10,568	72.3%	4,914	68.6%	5,654	75.8%
Smoking habit
	Never	7,888	53.7%	2,524	35.0%	5,364	71.6%
	Quit	3,423	23.3%	2,220	30.8%	1,203	16.1%
	Sometimes + everyday	3,379	23.0%	2,459	34.1%	920	12.3%
Drinking alcohol habit
	Never	5,192	35.4%	1,746	24.3%	3,446	46.2%
	≤ 2 times/week	5,130	35.0%	2,415	33.6%	2,715	36.4%
	≥ 3 times/week	4,328	29.5%	3,027	42.1%	1,301	17.4%
GHQ score
	≥ 4 points (poor)	5,713	39.1%	2,581	36.1%	3,132	42.0%
	≤ 3 points	8,891	60.9%	4,562	63.9%	4,329	58.0%
	0 point—12 points	3.4	3.4	3.2	3.4	3.6	3.4

In terms of participants’ mental health conditions, 36.1% of men and 42.0% of women were shown to have poor mental health conditions (≥ 4 GHQ-12 score). The results indicated statistically significant differences between men and women for every year and pooled data of all waves. Fig 1 shows participants’ GHQ-12 scores by gender.

Fig 1

Participants’ General Health Questionnaire 12-item scores by gender.

The levels of the General Health Questionnaire 12-item scores: 0 = ≤ 3 points, 1 = ≥ 4 points. Results of GHQ scores by gender: 2014: ***p < .001, 2015: ***p < .001, 2016: * p < .05, 2017: ***p < .001, 2018: **p < .01, Pooled data: n.s.

The results of the estimated associations between the GHQ scores and risk factors based on the random-effects conditional logistic regression model (RE-CLR), the fixed-effects conditional logistic regression (FE-CLR) model, and the results of the Hausman tests are shown in Table 3.

Table 3

Estimated associations between participants’ General Health Questionnaire 12-item scores and risk factors by gender based on the random-effects conditional logistic regression models, on the fixed-effects conditional logistic regression models and on the Hausman tests.

		All samples		Men		Women
		AOR(C.I.) 1)	AOR(C.I.)2)	AOR(C.I.)2)	AOR(C.I.)2)	AOR(C.I.)2)	AOR(C.I.)2)
Sex	Men	Ref.
	Women	1.419***
		(1.107–1.819)
Age	Under 39	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	40–49	0.971	1.117	0.857	1.121	1.146	1.120
		(0.791–1.190)	(0.815–1.532)	(0.623–1.179)	(0.692–1.815)	(0.880–1.492)	(0.738–1.700)
	50–59	0.836	1.082	0.766	0.935	0.936	1.184
		(0.666–1.049)	(0.693–1.691)	(0.538–1.092)	(0.474–1.846)	(0.697–1.256)	(0.653–2.149)
Number of persons in the household	≥ 2 people	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	One person	1.286	0.999	1.422*	1.085	0.994	0.905
		(0.946–1.749)	(0.634–1.575)	(0.946–2.139)	(0.612–1.924)	(0.620–1.594)	(0.430–1.904)
Employment status	unemployed	1.501***	1.727**	8.035***	5.852***	0.840	1.097
		(1.122–2.008)	(1.134–2.629)	(4.219–15.30)	(2.376–14.41)	(0.599–1.180)	(0.635–1.896)
	self-employed	0.832	1.193	0.896	1.134	0.608**	0.937
		(0.626–1.105)	(0.727–1.958)	(0.605–1.327)	(0.563–2.286)	(0.402–0.920)	(0.451–1.944)
	regular employee	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	non-regular employee	0.983	0.958	1.332	1.083	0.699**	0.764
		(0.774–1.247)	(0.680–1.350)	(0.808–2.196)	(0.565–2.077)	(0.527–0.928)	(0.485–1.203)
Disposable income per household	≥ 6,000K	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	2,000K-< 6,000K	1.254***	1.118	1.150	0.970	1.386***	1.273*
		(1.075–1.464)	(0.924–1.352)	(0.913–1.448)	(0.728–1.293)	(1.126–1.705)	(0.986–1.644)
	< 2,000K	1.977***	1.441*	1.714**	1.382	1.980***	1.464
		(1.475–2.650)	(0.993–2.092)	(1.082–2.713)	(0.782–2.440)	(1.351–2.901)	(0.881–2.434)
Sleep duration of the respondent	≥ 7 hours	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	6–7 hours	1.100	1.128	0.939	0.988	1.277**	1.285**
		(0.951–1.272)	(0.952–1.335)	(0.752–1.173)	(0.761–1.282)	(1.054–1.547)	(1.028–1.606)
	< 6 hours	1.644***	1.503***	1.613***	1.456**	1.662***	1.500**
		(1.355–1.993)	(1.177–1.919)	(1.203–2.161)	(1.012–2.095)	(1.283–2.151)	(1.070–2.104)
Physical exercise	≥ 3 days/week	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	≤ 2 days/week	1.175	1.066	1.128	0.979	1.265	1.184
		(0.897–1.538)	(0.792–1.435)	(0.768–1.656)	(0.646–1.485)	(0.864–1.851)	(0.767–1.828)
	No exercise	1.637***	1.418**	1.747***	1.531**	1.540**	1.326
		(1.284–2.088)	(1.062–1.895)	(1.227–2.488)	(1.013–2.315)	(1.099–2.160)	(0.874–2.012)
Smoking habit	Never	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
	Quit	1.135	0.929	1.119	1.054	1.176	0.856
		(0.909–1.419)	(0.579–1.490)	(0.794–1.578)	(0.507–2.193)	(0.880–1.570)	(0.476–1.540)
	Sometimes + everyday	1.217	0.940	1.201	1.176	1.242	0.714
		(0.950–1.560)	(0.520–1.700)	(0.850–1.697)	(0.503–2.750)	(0.854–1.805)	(0.302–1.685)
Drinking alcohol habit	Never
	≤ 2 times/week	0.893	1.033	0.817	0.928	0.945	1.073
		(0.745–1.069)	(0.809–1.320)	(0.602–1.108)	(0.609–1.413)	(0.758–1.179)	(0.793–1.452)
	≥ 3 times/week	0.785**	0.775	0.783	0.679	0.752*	0.819
		(0.629–0.979)	(0.552–1.088)	(0.564–1.087)	(0.407–1.133)	(0.547–1.035)	(0.503–1.334)
	Constant	0.189***		0.201***		0.309***
		(0.129–0.278)		(0.116–0.350)		(0.188–0.506)
	Observations	12,681	5,992	6,270	2,748	6,411	3,244
	Number of respondents	3,359		1,657		1,702
	Model Type	RE	FE*	RE	FE*	RE*	FE
	Hausman Test	0.0061		0.0586		0.4554

1) Adjusted odds ratios (AOR) with 95% CI (adjusted for sex, age, number of persons in the household, employment status, disposable income per household, sleep duration in weekdays, physical exercise, smoking habit, and drinking alcohol).

2) Adjusted odds ratios with 95% CI (adjusted for age, number of persons in the household, employment status, sleep duration in weekdays, physical exercise, smoking habit, and drinking alcohol).

Bold ratios: statistically significant results.

The levels of the General Health Questionnaire 12-item scores: 0 = ≤ 3 points, 1 = ≥ 4 points.

Robust cieform in parentheses.

Ref.: Reference [1].

*** p<0.01

** p<0.05

* p<0.1.

The analysis made by computing adjusted odds ratios (AOR) among all participants and by gender indicated the differences among the potential risk factors. Among all participants, the Hausman test supported the FE-CLR model, and the estimated results indicated significant associations between poor mental health conditions and variables such as being unemployed (AOR 1.727 [95% CI: 1.134–2.629]), low level of disposable income per household (< 2,000K yen: AOR 1.441 [95% CI: 0.993–2.092]), having a short sleep duration (< 6 hours: AOR 1.503 [95% CI: 1.177–1.919]), and lack of physical exercise (AOR 1.418 [95% CI: 1.062–1.895]). It is important to note that the middle variables for disposable income per household (between 2,000K–6,000K yen: AOR 1.254 [95% CI: 1.075–1.464]) and for drinking habits (≥ 3 times/week: AOR 0.785 [95% CI: 0.629–0.979]) exhibited significant effects on poor mental health only in the RE-CLR model.

Among male participants, the Hausman test supported the FE-CLR model, and the estimated results indicated significant associations between poor mental health conditions and variables such as being unemployed (AOR 5.852 [95% CI: 2.376–14.41]), having a short sleep duration (AOR 1.456 [95% CI: 1.012–2.095]), and lack of physical exercise (AOR 1.531 [95% CI: 1.013–2.315]). Once again, only in the RE-CLR model, variables such as living alone (AOR 1.422 [95% CI: 0.946–2.139]) and low level of disposable income (AOR 1.714 [95% CI: 1.082–2.731]) showed significant effects on poor mental health. However, these variables did not show significant associations in the FE-CLR model.

Among female participants, the RE-CLR model was supported by the Hausman test, and the results indicated significant associations between poor mental health conditions and middle (AOR 1.386 [95% CI: 1.126–1.705]) and low levels of household disposable income (AOR 1.980 [95% CI: 1.351–2.901]), having a middle (6–7 hours: AOR 1.277 [95% CI: 1.054–1.547]) and a short sleep duration (AOR 1.662 [95% CI: 1.283–2.151]), and lack of physical exercise (AOR 1.540 [95% CI: 1.099–2.160]). However, variables such as self-employment (AOR 0.608 [95% CI: 0.402–0.920]), being a non-regular employee (AOR 0.699 [95% CI: 0.527–0.928]), and drinking alcohol more than 3 times/week (AOR 0.752 [95% CI: 0.547–1.035]) were inversely associated with poor mental health conditions.

Discussion

In this study, we utilised a 5-wave longitudinal panel survey to examine the prevalence of poor mental health conditions among men and women in Japan, which was measured by participants’ GHQ-12 scores. Among the AOR analysis both in the RE-CLR and FE-CLR models, for all participants, the results indicated significant associations between poor mental health conditions and being unemployed, low levels of disposable income, short sleep durations, and lack of physical exercise. Among men, there were significant associations between poor mental health conditions and being unemployed, short sleep duration, and lack of physical exercise. Among women, results indicated significant associations between poor mental health conditions and low levels of household disposable income, short sleep duration, and lack of physical exercise. Contrastingly, self-employment, being a non-regular employee, and drinking alcohol more than 3 times per week were inversely associated with poor mental health conditions. In general, these results were consistent with the results obtained from previous cross-sectional studies that addressed the risk factors of poor mental health conditions, with unemployment and low levels of disposable income being cited in two [33, 35], short sleep duration being cited in three [38, 42–45], and lack of physical exercise being cited in five studies [15, 19, 40]. Thus, it is confirmed that these factors are those most associated with effects on mental health conditions in terms of cross-sectional and longitudinal direction.

In terms of SES factors such as disposable income and employment status, previous studies reported that lower income groups showed an association with poor mental health conditions [32-35], and that the socioeconomically disadvantaged were more likely to experience poor mental health compared with the advantaged [32]. In corroboration, individuals who had faced financial hardship showed significantly poorer mental health conditions [32, 36]. Nevertheless, the results from our study based on the longitudinal panel data showed that variations within the same variable, such as between higher and lower levels of disposable income, did not cause any changes in mental health conditions. When comparing the FE-CLR model with the RE-CLR model analysis among all participants, the middle level of disposable income was significantly associated with a poor mental health condition exclusively in the RE-CLR model, which was not supported by the Hausman test. This implies that the estimated associations in the RE-CLR model were mainly caused by participants’ time-invariant factors. Those time-invariant factors were all controlled in the FE-CLR model where no significant association was found between disposable income and mental health condition. Similarly, among men, the low level of disposable income was significantly associated with a poor mental health condition exclusively in the RE-CLR model, which was not supported by the Hausman test. From these results, we can interpret that poor mental health conditions tend to be observed among men and women with middle levels of disposable income and among men with low levels of disposable income. While lower levels of disposable income may not be the cause of poor mental health conditions, it is a potential risk factor that can result in poor mental health.

Regarding employment status, the results of our study showed that both unemployed all samples and male samples had significant associations with poor mental health conditions. Lindstrom et al. presented that the unemployed and participants during periods of long-term sick leave had significantly higher ratios of poor psychological well-being among both men and women by using multiple analyses in a cross-sectional study [32]. A likely supposition is that the unemployed and those experiencing long-term sick leave may have low levels of disposable income. In our study, SES factors including unemployment and low levels of disposable income were associated with poor mental health. The results implied that clarifying the link between the complex socioeconomic factors such as disposable income that may lead to poor mental health is a clear requirement to improve mental health conditions and to minimise health inequalities.

However, regarding employment status, the females’ result was different from the males’. This study revealed that women who were self-employed or non-regular employees showed inverse associations with poor mental health conditions. In that regard, previous studies presented that non-permanent employees reported higher job dissatisfaction [33], but lower levels of stress than permanent employees [33, 34]. Further research is needed to identify the relationship between mental health and employment status by gender, particularly related to types of employment, gender roles, and their possible associations.

In terms of lifestyle factors, sleep duration on weekdays and physical exercise habits for men and women alike were significantly associated with poor mental health in both the RE-CLR and FE-CLR models. Specifically, sleeping less than 6 hours a night significantly contributed to poor mental health outcomes compared with sleeping more than 6 hours a night. In correlation, previous studies showed significant negative impacts of short sleep duration among generations currently in their ‘working years’ on depression [35, 36] and anxiety [37, 38]; poorer GHQ-12 scores were demonstrated as well [38]. Thus, short sleep duration could be a cause of poor mental health [37], and our findings suggest that sleeping 6 hours or more per day can be critical for improved mental health outcomes.

Regarding exercise, the lack of physical activity could be a risk factor for poor mental health conditions [39], and some studies report that physical activity could play a significant role in promoting better mental health outcomes [15, 19, 39–41]. Moreover, one previous study reported that adolescents who had low aerobic fitness were more likely to report poorer sleep quality [42]. Based on these findings, there is a need to promote regular practice of physical activities in order to improve quality of sleep and mental health conditions, mainly because it helps reduce stress levels.

Among men, living alone showed a significant association with poor mental health exclusively in the RE-CLR model. We can interpret these results as poor mental health conditions are likely to be observed in men who live alone, but the results do not allow for an understanding that a change from living with someone to living alone may influence mental health, because the FE-CLR model did not show a significant relationship between living alone and mental health. Some previous cross-sectional studies presented that depressive symptoms were significantly associated with living alone [35, 43, 44]. Still, according to our results based on longitudinal panel data, there was no significant longitudinal and possibly causal association between living alone and poor mental health.

Regarding drinking habits among all participants, the RE-CLR model indicated that drinking alcohol more than 3 times per week had significant inverse associations with poor mental health conditions in the RE-CLR model, but it did not show any significant association in the FE-CLR model, which was supported by the Hausman test. These results imply the reserve causality. That is, drinking habits (drinking alcohol more than 3 times per week) do not necessarily improve mental health, but participants with a good mental health condition tended to drink alcohol more than 3 times per week. The association between mental health and drinking habits are controversial. A study reported that problematic drinking habits are associated with depressive symptoms and suicidal ideation for men and women alike [35]. Notwithstanding, some studies, including a cohort study, reported that there was no association between mental health and drinking habits [15, 19, 39]. For this study, we controlled for unobservable and time-invariant variables, such as natural good mental health conditions, by using the FE-CLR models. Then, we found that there is no longitudinal association between mental health and drinking habits, a finding consistent with previous cohort studies [15, 19, 39]. Contrastingly, problem drinking may be a part of some stress coping strategies [45], so future studies are warranted to identify the relationship between stress coping and the amount of alcohol consumption.

Past research showed an association between practising several healthy lifestyle activities daily and reducing the risk of depression [13]. This same study suggested that sleeping for a proper duration, engaging in regular physical activity, and following abstemious drinking habits, which are all healthy lifestyle practices, could be essential for improving mental health. However, the underlying mechanisms that relate SES and lifestyle-related factors with mental health are complex [46], and studies that enhance our understanding related to these mechanisms are warranted, mainly studies that allow for finding causal relationships between variables, such as randomised control trials.

This study has several limitations. First, since we utilised a panel dataset, there is a possibility of sample attrition bias. In that regard, Baltagi reported that the average attrition rate of panel studies revolves around 10% [25]. In corroboration, our study showed an attrition rate revolving less than 10%. Further, in our analysis, we need to consider the possibility that respondents who dropped out might have had unhealthy outcomes or have been in unfavourable situations [25]. Future studies that provide other types of longitudinal data and that ensure consistency in terms of participants are warranted. Second, the questionnaires about SES and lifestyle factors were self-reported. Although we analysed data using the RE-CLR and FE-CLR models, under- or over-reporting of socially desirable attitudes and recall bias are a reality for this type of measure, and future studies should address these problems. Third, regarding causal relationships, we only controlled for the time-invariant factors using fixed-effects models. Thus, if any time-variant factors brought about reverse causality, our estimated results do not allow for the understanding of a causal relationship between two variables. Future studies should address this issue and control for time-variant factors that can bring reverse causality.

However, this study successfully examined associations between mental health outcomes and their potential risk factors using a large-scale, 5-wave longitudinal panel dataset which has some benefits. First, panel data suggests that individuals are heterogeneous and enables control of unobservable and time-invariant variables. For that, the RE-CLR and FE-CLR models were employed to identify longitudinal associations and the results of the Hausman test identified which model was a better fit for analysing socioeconomic and lifestyle-related factors’ effects on mental health outcomes. Contrastingly, cross-sectional studies cannot control for heterogeneity and suffer from omitted variable biases [25]. Second, panel data are better to analyse changes and to identify associations among respondents. Supported by previous studies, our analyses utilised panel data to examine how respondents’ changes in SES- and lifestyle-related factors affected mental health outcomes [47]. Third, our respondents were taken from an age category of working adults, randomly sampled from a nationwide population. Our final number of observations exceeded 14,717, and participants’ response rates were over 90% for every year. After controlling for relevant covariates, this study revealed probable risk factors for poor mental health conditions and the significant associations between socioeconomic and lifestyle factors and mental health.

Conclusions

This study revealed that the prevalence of poor mental health conditions among women was higher compared to men during 2014–2018, the five years on which we focused. When comparing the RE-CLR and the FE-CLR models, poor mental health conditions tended to be observed among all respondents with middle levels of disposable income and among men with a low level of disposable income or who lived alone. Our results imply that there is no causal relationship between the levels of disposable income and mental health conditions, and that changes in disposable income between high and low levels would not cause any changes in mental health conditions. In this case, aggressive policy interventions for increasing disposal income for people with lower socioeconomic status would not be necessarily effective to minimize health inequalities. Regarding living status among men, we found no significant causal association between living alone and poor mental health conditions, so living alone was not shown to be a cause of poor mental health conditions. The results of this study showed that while lower levels of disposable income and living status are not significant causes, they are potential risk factors for poor mental health. Further research is needed to identify the effects of disposable income and living status on mental health conditions, as well as the role of gender.

SES factors such as unemployment showed significant associations with poor mental health conditions among men. However, among women, self-employed and non-regular employees showed inverse associations with poor mental health conditions. Future studies should focus on identifying the associations between mental health and employment status by gender, particularly related to the types of employment, cultural gender roles, and their possible associations.

Our results suggest that unhealthy lifestyle factors such as short sleeping duration and lack of physical exercise may be potential risk factors for poor mental health for both men and women. Also, we found no causal association between alcohol habits and mental health conditions. Based on our results and the results of previous studies, promoting healthy lifestyle practices would help improve mental health conditions. Thus, we recommend conducting more longitudinal studies to examine causal relationships.

There is a complex mechanism behind the association between socioeconomic and health-related factors that lead to health inequality. Further research should explore differences provoked by gender, age, SES, and healthy lifestyle practices on health outcomes. Determining the complex mechanisms that relate mental health conditions to socioeconomic and lifestyle factors can be beneficial, as this knowledge may allow us to develop effective social welfare policies and health promotion interventions that are adequately equipped to improve mental health conditions, thereby minimising health inequalities caused by socioeconomic factors.

15 May 2020

PONE-D-20-10688

Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan

PLOS ONE

Dear Dr. Nagasu,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The three reviewers addressed several major and minor concerns about your manuscript. Please revise your manuscript carefully.

We would appreciate receiving your revised manuscript by Jun 29 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Kenji Hashimoto, PhD

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements:

1.    Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2.PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Dear authors and editors

Thank you for the opportunity to review this study. The association between socioeconomic position and lifestyle and the impact they have on mental health are important issues in order to direct actions towards improving mental health. The authors use repeated answers to the GHQ from national representative surveys over a five-year period. There are two aims; one to examine the prevalence of poor mental health and secondly to analyse the differences between the estimates from random and fixed models of logistic regression, in order to identify risk factors for poor mental health.

The abstract teases with the statement, that the results imply the reverse causality that poor mental health conditions cause lower disposal income. Usually it is considered a fact, that serious mental disorders (schizophrenia/ bipolar disorder) causes low socioeconomic position whereas common mental disorders are caused by low socioeconomic position, therefor the study draw my attention.

Unfortunately, I find the manuscript hard to evaluate, due to many discrepancies in tables and the text as well as methodological problems. I cannot recommend publication. In the following you have my comments to the major problems in the study. I hope the comments are useful for you.

Abstract

The abstract gives a good introduction but should state the aims clearly and mirror them in the reported results – e.g. prevalence is not mentioned at all.

Introduction

The authors quote WHO for stating inequality in health is unfair and subsequent studies have found …. The studies referred to are much before the statement by WHO (ref 4 & 5).

Several longitudinal studies in this field are not included in the literature mentioned. It is claimed the cross-sectional studies are unadjusted – that is quite rare, at least for studies on common mental disorders.

The negative impact on mental health of economic changes are well documented – one cited here. 1

Methods

JPHS is short of Japanese Household Panel Survey, I guess KHPS is short of Keio Household Panel Survey?

It is not evident how you reach 3.501 participants out of – as I can calculate (4.005 + 10.458) – 2.204. Then you have a drop out of 8.758? Even though you have a response rate of 91-98% in 2014 – 2018 in table 1 – which I do not understand at all. Here you have much more respondents: 14.717? And we have a division of JHPS and KHPS – even though they have merged/renamed in 2009? We need a clearer presentation of how the sample ends at 3.501 and a revision of Table 1, where n respondents is replaced by received questionnaires, if that is so – “waves” need to be explained as well and why both KHPS and JHPS appears. If none are replaced in the five years, how many drop out?

The variables must have been defined by JHPS in the 2004, so not much to do about the questions and categorisation, but it would be informative to know if the income-groups are equally shared in Japan with 1/3 in each? Why these income-groups?

Disposable household income is a good measure, if the size of the household is known; here we only have one or more than one. In other words: we know the vehicle has 100 HP, but we do not know the load it has to carry - if it is a lorry or a car.

The method of measuring health behaviors seems fair from a health perspective, except for the question on drinking. This question is not comparable to other studies on drinking habits/alcohol use and does not seem relevant. To drink a beer three times a week is not considered a drinking problem in most OECD countries. It may be different in Japan, but then it needs explanation.

The chosen statistical methods are unconventional.

Results

Here 14.717 participants are presented as shown in table two. What to believe?

Table 2: %SD – do you mean %? Very few women have regular employment and drink alcohol three times a week, very few respondents have a low household income. It needs an explanation – the numbers after “in Japanese Yen” what are they? The “0 points – 12 points” at GHQ, what is that? An average? The GHQ- score is a result pooled over the years I guess, please write this, if so. 42% women have poor mental health? It makes you wonder if GHQ valid in this context? Please add reference on the validation of GHQ-12 in a Japanese context in the method section.

It is not mentioned in the method section, but the adjustments are done only for other variables than the one presented/analysed, isn’t it?

Table 3: The table is difficult to read. I would suggest the FE/RE was stated in the heading section. Bold is stated to indicate statistically significant findings but seems used at random. How can an AOR for RE-female alcohol >= 3 times a week at 0.752 (0.547 – 1.179) be significant? The non-significant results for men and women end up being 0.7885 (0.629 – 0.979)? Why are there fewer observations for the FE-anaysis than the RE-analysis?

Discussion

Limitations. Dropouts are discussed in general terms. What was the actual characteristics of the dropouts in this study? The income variable is very screwed – is that sampling bias or at true reflection of the study population?

Information bias is mentioned only as recall. However, the instrument is not (or is?) validated; again, when adjusting for the number of persons in household – which is very good -the variable is not covering that, but only one or more than one. This is a serious problem and makes any conclusion related to income invalid.

“Further research is needed to identify the effects of disposable income and living status on mental health conditions, as well as the role of gender”. They do exist in plenty – as do so for employment status and gender.

You do not reflect on the high prevalence of poor mental health, why? it is a central study objective. However, to examine the prevalence of poor mental health longitudinal data are not optimal, unless you want to give information on the development in mental health. Again, it is not evident if GHQ is validated in a Japanese population – and a pooled prevalence of 42% women in poor mental health does not seem reliable.

As for household income as socioeconomic index the validity is poor – first very few are in lowest category and even though it is stated the analyses are adjusted for number of persons in the household, when in fact it is only adjusting for one or more than one.

As for the FE logistic regression analyses vs RE logistic analyses none of them account for the time each individual contributes with in the time series (person-years), and thus less accurate than the traditional methods used for longitudinal studies.

1. Barbaglia MG, M. tH, Dorsselaer S, et al. Negative socioeconomic changes and mental disorders: a longitudinal study. J Epidemiol Community Health 2015;69(1):55-62.

Reviewer #2: In general, this study is well constructed to gain a rational outcome. I have some minor comments to be addressed for better understanding of the study.

In the introduction section, the author mentioned that Japan was shown to be one of the countries with the highest suicide rates in the world. It is partially acceptable, but a little over-represented. Some Eastern European nations as well as Russia has higher suicide rate. Also, in latest statistics, those of the US and Sweden are not so different from Japan's. I am afraid that outdated articles the author referred can be misleading.

In this study, the author classified some items the participants answered into some groups. How the author decide the thresholds of each group?

For example, are there any reasons that people taking 6-7 hours of sleep in a day is different to those taking >7 hours sleep, not >8? In my understanding, there are rich evidence suggesting sleeping under 6 hours in a day is harmful. But how many hours you should sleep is controversial.

Also, the author referred a Glozier's work. But its subjects were limited to young people recommended to take 8-9 hours sleep. The author should consider to show better preceding studies.

In this study, only 7.9% of the participants lived alone. According to recent official statistics in Japan, one-fourth of the household is composed of one person. In my calculation, single person household was 10.7% in 2014. Did the author compare the demographic data with those of contemporary official statistics? If there is a large discrepancy between them, the representativeness of the panel is doubtful.

The same thing is also adapted to employment status, but it seems consistent with the official statistics, as far as my checking.

Table 2 is difficult to read at a glance because each number and the mean share the same column. The author should rearrange it.

Is the mean income of Japanese only 542.8 JPY?

What does "0 point - 12 points" mean in the GHQ score section?

Figure 1. is hardly understandable and space-killing. What does the vertical axis mean? I reckon it shows the percentage. Anyway, it seems better to choose another type of graph (polygonal line graph, or a chart maybe).

Reviewer #3: This epidemiological survey is important for mental healthy field, worth reading. The authors however shoud propose the hypothesis of the study clearly, and enphasize the new findings in the present study, and the differences from the previous reports throuout the manuscript.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Aake Packness, PhD, MPH, RN

Reviewer #2: Yes: Akihiro Shiina

Reviewer #3: Yes: Yoshimura Reiji

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

10 Sep 2020

Authors’ response to reviewers:

PLOS ONE

PONE-D-20-10688

Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan

Reviewer #1:

Thank you very much for reviewing our manuscript. We have revised the manuscript according to your comments and suggestions.

Abstract

1. The abstract gives a good introduction but should state the aims clearly and mirror them in the reported results – e.g. prevalence is not mentioned at all.

Response:

We have rewritten many parts of the abstract and added the following sentences. Further, the results of this study are added in the abstract.

‘This study aims to reveal the prevalence of poor mental health conditions among Japanese individuals and to identify SES- and lifestyle-related risk factors that might lead to these conditions.’

‘The prevalence of poor mental health conditions, represented by a GHQ-12 score of 4 or more, was 36.1% and 42.0% of men and women, respectively.’

‘Various factors, such as unemployment, low household income, short nightly sleeping duration, and lack of exercise, showed significant longitudinal (within) associations with mental health conditions estimated by the FE-CLR models.’

Introduction

2. The authors quote WHO for stating inequality in health is unfair and subsequent studies have found …. The studies referred to are much before the statement by WHO (ref 4 & 5).

Response:

Thank you for pointing this out. We have corrected the sentence as follows:

‘Several studies have found that differences in socioeconomic status (SES) are one of the causes of health inequalities and are associated with health outcomes (2, 4, 5).’

We wanted to cite these references because, during the 20 years, surprisingly, we have not solved health inequality and still WHO has to write ‘Health inequality is unfair’. As you mentioned, a study (Reference No. 5, 2003) reported that suicide risk is strongly associated with factors related to low SES, such as unemployment and low income. Lynch et al (Reference No. 4, 1997) also mentioned that socioeconomic inequalities in health have often made reference to the observation that poor health behaviours and psychosocial characteristics cluster in low SES groups. Since then, even in 2017, we should realise that still WHO had to write ‘Health inequalities are unfair’ on their website. We believe that these references should be included to discuss this issue.

3. Several longitudinal studies in this field are not included in the literature mentioned. It is claimed the cross-sectional studies are unadjusted – that is quite rare, at least for studies on common mental disorders.

Response:

Thank you for pointing this out. We added some references of longitudinal studies as follows:

‘Thus, any possible associations between suicide and mental health, SES, and lifestyle-related factors should be investigated (11, 12).’

‘Cross-sectional and longitudinal studies have demonstrated that healthy lifestyle factors are positively associated with better mental health outcomes (13). Thus, the relationship between mental health conditions and lifestyle factors, such as sleeping duration (14, 15), habitual physical exercise (16, 17), smoking habits (18, 19), and alcohol consumption (20), have been investigated to identify methods for preventing mental illness.’

As the mentioned by the reviewer that it is claimed the cross-sectional studies are unadjusted – that is quite rare, at least for studies on common mental disorders. We partially agree with the comment. We believe that this does not imply ‘the cross-sectional studies’, about ‘the logistic regression model used in the cross-sectional studies’, but this implies that the variables are adjusted in the logistic regression models and so on.

As we explained, we applied fixed-effects regression models that automatically adjust for all time-constant unobserved confounders and help reduce the risk of omitted variable bias, as well as adjust for identified time-varying confounders. These characteristics are different from those of logistic regression models.

Accordingly, we added the following sentence as an explanation about the fixed-effects model:

‘The fixed-effects regression models are used to adjust for all time-constant unobserved confounders and decrease the risk of omitted variable bias, as well as adjust for identified time-varying confounders.’

4. The negative impact on mental health of economic changes are well documented – one cited here.

Response:

Thank you for the suggestion. We added the recommended paper as follows:

‘Barbaglia et al. reported that negative socioeconomic changes, such as substantial reduction in household income and job loss, significantly increased the risk of incident mental disorders (23).’

Methods

5. JPHS is short of Japanese Household Panel Survey, I guess KHPS is short of Keio Household Panel Survey? It is not evident how you reach 3.501 participants out of – as I can calculate (4.005 + 10.458) – 2.204. Then you have a drop out of 8.758? Even though you have a response rate of 91-98% in 2014 – 2018 in table 1 – which I do not understand at all. Here you have much more respondents: 14.717? And we have a division of JHPS and KHPS – even though they have merged/renamed in 2009? We need a clearer presentation of how the sample ends at 3.501 and a revision of Table 1, where n respondents is replaced by received questionnaires, if that is so – ‘waves’ need to be explained as well and why both KHPS and JHPS appears. If none are replaced in the five years, how many drop out?

Response:

As we mentioned in abbreviation list, the JHPS/KHPS stands for the Japan Household Panel Survey and Keio Household Panel Survey. The JHPS/KHPS has two series of samples: the KHPS and the JHPS started in 2004 and in 2009, respectively. In 2014, the KHPS and JHPS were integrated in one survey called the KHPS/JHPS. Given that the JHPS and KHPS comprise the same questionnaire from 2014, we can use both samples for the analysis. Hence, we added the sentence below mentioned in the method section:

‘The JHPS/KHPS is a panel (longitudinal) data in which there are multi-dimensional data involving measurements over time. It contains time-series observations of each participant, and multiple phenomena (answers for questionnaires in this case) are obtained over multiple time periods (for 5 years in this study) for the same participants’.

We also wrote the following in the abstract: ‘Our sample comprised 14,717 observations of 3,501 individuals aged 22–59 years for five waves of the Japanese Household Panel Survey.’. Regarding the number of participants who dropped out from the survey, we added the numbers in brackets in Table 1.

We discussed the number of respondents who participated in the studies from 2014 to 2018 in Table 1. Regarding your calculation, we apologize, but we cannot understand how you derived these numbers in Table 1: (4.005 + 10.458) – 2.204. In the table, we also added information on how to calculate the response rate (%).

When using panel data or longitudinal data in Public Health 1,2) and Economics, researchers commonly use technically the word ‘wave’ instead of ‘year’. However, we added an explanation about ‘wave’ in sampling of the respondents section as follows: ‘In panel survey, ‘wave’ is generally used as the same meaning as ‘year’

Reference:

1: Carroll SJ, Dale MJ, Niyonsenga T, Taylor AW, Daniel M. Associations between area socioeconomic status, individual mental health, physical activity, diet and change in cardiometabolic risk amongst a cohort of Australian adults: A longitudinal path analysis. PloS one. 2020;15(5):e0233793.

2: Wang S, Mak HW, Fancourt D. Arts, mental distress, mental health functioning & life satisfaction: fixed-effects analyses of a nationally-representative panel study. BMC public health. 2020;20(1):208.

Table 1. Number of respondents and response rates for each survey and number of recruited participants from each survey.

JHPS KHPS Total

Wave Survey year Total number of respondents1) Response rate (%)2) Respondents in this study Wave Survey year Total number of respondents Response rate (%) Respondents in this study

6 2014 2,358 (238) 91.1 1,385 11 2014 3,312 (275) 92.6 2,102 3,487

7 2015 2,198 (186) 93.0 1,253 12 2015 3,124 (229) 98.8 1,939 3,192

8 2016 2,048 (163) 92.8 1,144 13 2016 2,945 (207) 94.0 1,776 2,920

9 2017 1,885 (175) 91.9 1,057 14 2017 2,741 (229) 92.7 1,628 2,685

10 2018 1,742 (147) 92.2 970 15 2018 2,549 (206) 93.0 1,463 2,433

Total 10,231 (909) 92.2 5,809 Total 14,671 (1,146) 94.2 8,908 14,717

1) Total number of respondents: the number of collected questionnaires that were completed by the respondents. The numbers in brackets show the number of people who dropped out.

2) Response rate (%) = (number of completed questionnaires – number of people who restarted answering questionnaire)/number of completed questionnaires in the previous year) × 100)

6. The variables must have been defined by JHPS in the 2004, so not much to do about the questions and categorisation, but (1) it would be informative to know if the income-groups are equally shared in Japan with 1/3 in each? (2) Why these income-groups?

Response:

Thank you for your question. The categorisation is the same as the Japanese government survey, the National Survey Health and Nutrition (Ministry of Health, Labour and Welfare, Japan) in 2014. The income groups are not equally shared with 1/3 in both the National Survey Health and Nutrition and JHPS, as presented in Table 2. It should be noted that the table shows that the proportion of each income group is different between the government survey and JHPS. The partial possible reason for this difference is that JHPS sample include more single person household because the participants of JHPS were selected at the individual level while those of the government survey were at the household level.

Table 2. Categories of disposable income

National Survey Health and Nutrition (2014) This study (JHPS)

Households All

n % n %

≥ 6,000 K 717 22.0 4856 36.9

2,000 K-< 6,000 K 1,765 54.0 7373 56.0

< 2,000 K 784 24.0 935 7.1

Total 3,266 100.0 13164 100.0

To account for the possible bias due to the smaller sample of single persons, we conducted a weighted logit regression using the inverse of the number of households as weight. As presented in Table 3, although some of the statistical significance vary, the sign (more than 1 or not) and magnitude of the odds ratio seem not to be so different between the estimates with and without weight. Thus, we regard that although there is a sampling bias in our data because less persons living alone were selected, it does not have large effects on the estimation results.

Table 3. Results of RE and estimation with weight

Results of RE　 Results of estimation with weight

odds ratios odds ratios

Sex Men Refe Refe

Women 1.419*** 1.236***

(1.107–1.819) (1.122–1.362)

Age Under 39 Refe Refe

40–49 0.971 0.951

(0.791–1.190) (0.866–1.045)

50–59 0.836 0.861***

(0.666–1.049) (0.782–0.948)

Number of persons in the household ≥ 2 people Refe Refe

One person 1.286 1.248***

(0.946–1.749) (1.091–1.428)

Employment status unemployed 1.501*** 1.165**

(1.122–2.008) (1.022–1.328)

self-employed 0.832 0.832***

(0.626–1.105) (0.737–0.940)

regular employee Refe Refe

non-regular employee 0.983 0.987

(0.774–1.247) (0.887–1.098)

Disposable income per household ≥ 6,000 K Refe Refe

2,000 K-< 6,000 K 1.254*** 1.198***

(1.075–1.464) (1.106–1.299)

< 2,000 K 1.977*** 1.770***

(1.475–2.650) (1.517–2.064)

Sleep duration of the respondent ≥ 7 h Refe Refe

6–7 h 1.100 1.028

(0.951–1.272) (0.945–1.119)

< 6 h 1.644*** 1.418***

(1.355–1.993) (1.288–1.561)

Physical exercise ≥ 3 days/week Refe Refe

≤ 2 days/week 1.175 1.228***

(0.897–1.538) (1.054–1.429)

No exercise 1.637*** 1.449***

(1.284–2.088) (1.270–1.652)

Smoking habit Never Refe Refe

Quit 1.135 1.076

(0.909–1.419) (0.979–1.183)

Sometimes + everyday 1.217 1.126**

(0.950–1.560) (1.020–1.243)

Drinking alcohol Never Refe Refe

≤ 2 times/week 0.893 0.879***

(0.745–1.069) (0.806–0.959)

≥ 3 times/week 0.785** 0.894**

(0.629–0.979) (0.812–0.985)

Constant 0.189*** 0.360***

(0.129–0.278) (0.301–0.431)

Observations 12,681 12,681

Number of respondents 3,359

*** p<0.01, ** p<0.05, * p<0.1

To explain the above points, we added the following sentences into the method and limitation part:

‘The categorisation method is the same as the government survey, National Survey Health and Nutrition (Ministry of Health, Labour and Welfare, Japan) in 2014.’

‘First, for the proportion of single person household, we compared the demographic data of this study with those of the national census taken on Oct. 2015. The result of this study seems to have considerably fewer participants lived alone than the results of the national census. Because the participants of this study were selected at the individual level, that of the national census were at the household level. We also calculated the composition of single-person living at the individual level as 14.4% (the number of single person household is 18,420, which amounts to 14.4% of the Japanese population of 128,000 thousands). Therefore, there is about a double difference between population (14.4%) in the national census and our sample (7.9%). To account for the possible bias due to the smaller sample of single-person, we conducted a weighted logit regression using the inverse of the number of households as weight. Although there are some statistical significance between RE-CLR and RE-CLR with weight, the sign (more than 1 or not) and magnitude of the odds ratios seem not to be so different between the estimates with and without weight. Thus, we regard that although there is a sampling bias in our data in the sense that less persons living alone were selected, it does not largely affect the estimation results.’

7. Disposable household income is a good measure, if the size of the household is known; here we only have one or more than one. In other words: we know the vehicle has 100 HP, but we do not know the load it has to carry - if it is a lorry or a car.

Response:

Thank you for your comment. Our intention to categorise into two groups was to distinguish between married or unmarried. As the number of household member is actually available in JHPS, we confirmed that the estimation results did not change significantly even if we used the dummy variables indicating the size of household: a one-person, two-person, three-person, and four-person dummies or more. Please find the comparison of the estimation results depending on two types of the variable about ‘the number of household members.’

To explain this, we added the following sentence in the Methods section.

‘Although the number of household members is actually available in JHPS, it was categorised into two groups: living with someone (≥ 2 people) and living alone. We confirmed that the estimation results did not change significantly even if we used the dummy variables indicating the size of household: a one-person, two-person, three-person, and four or more person dummies.’

Table 4. Estimated associations between the GHQ scores and risk factors with two types of the variable ‘the number of household member.

All samples All samples

AOR(C.I.) 4) AOR(C.I.)4) 　 AOR(C.I.) 5) AOR(C.I.)5)

Sex Men Refe Refe

Women 1.419*** 1.419***

(1.107 - 1.819) (1.108 - 1.818)

Age Under 39 Refe Refe Refe Refe

40-49 0.971 1.117 0.968 1.093

(0.791 - 1.190) (0.815 - 1.532) (0.790 - 1.186) (0.800 - 1.495)

50-59 0.836 1.082 0.819 1.067

(0.666 - 1.049) (0.693 - 1.691) (0.651 - 1.029) (0.685 - 1.661)

Number of persons in the household 1 person 1.286 0.999 1.396 0.860

(0.946 - 1.749) (0.634 - 1.575) (0.962 - 2.026) (0.495 - 1.492)

2 people Refe Refe 1.184 0.854

(0.873 - 1.608) (0.529 - 1.378)

3 people 1.167 0.926

(0.884 - 1.542) (0.618 - 1.387)

4 people 1.052 0.924

(0.813 - 1.361) (0.643 - 1.328)

5 people Refe Refe

6 people 1.038 1.076

(0.676 - 1.594) (0.595 - 1.945)

≥ 7 people 0.852 0.940

(0.486 - 1.496) (0.406 - 2.173)

Employment status Unemployed 1.501*** 1.727** 1.491*** 1.692**

(1.122 - 2.008) (1.134 - 2.629) (1.115 - 1.994) (1.112 - 2.574)

Self-employed 0.832 1.193 0.830 1.163

(0.626 - 1.105) (0.727 - 1.958) (0.625 - 1.102) (0.712 - 1.898)

Regular employee Refe Refe Refe Refe

Non-regular employee 0.983 0.958 0.985 0.950

(0.774 - 1.247) (0.680 - 1.350) (0.776 - 1.250) (0.675 - 1.337)

Disposable income per household ≥ 6,000 K Refe Refe Refe Refe

2,000 K-< 6,000 K 1.254*** 1.118 1.243*** 1.109

(1.075 - 1.464) (0.924 - 1.352) (1.065 - 1.451) (0.918 - 1.339)

< 2,000 K 1.977*** 1.441* 1.940*** 1.423*

(1.475 - 2.650) (0.993 - 2.092) (1.446 - 2.602) (0.981 - 2.065)

Sleep duration of the respondent ≥ 7 h Refe Refe Refe Refe

6-7 h 1.100 1.128 1.094 1.116

(0.951 - 1.272) (0.952 - 1.335) (0.947 - 1.265) (0.943 - 1.321)

< 6 h 1.644*** 1.503*** 1.624*** 1.466***

(1.355 - 1.993) (1.177 - 1.919) (1.339 - 1.970) (1.148 - 1.871)

Physical exercise ≥ 3 days/week Refe Refe Refe Refe

≤ 2 days/week 1.175 1.066 1.174 1.074

(0.897 - 1.538) (0.792 - 1.435) (0.897 - 1.537) (0.797 - 1.447)

No exercise 1.637*** 1.418** 1.640*** 1.432**

(1.284 - 2.088) (1.062 - 1.895) (1.286 - 2.091) (1.072 - 1.912)

Smoking habit Never Refe Refe Refe Refe

Quit 1.135 0.929 1.130 0.927

(0.909 - 1.419) (0.579 - 1.490) (0.905 - 1.412) (0.577 - 1.488)

Sometimes + everyday 1.217 0.940 1.215 0.943

(0.950 - 1.560) (0.520 - 1.700) (0.948 - 1.557) (0.521 - 1.708)

Drinking alcohol Never Refe Refe Refe Refe

≤ 2 times/week 0.893 1.033 0.898 1.032

(0.745 - 1.069) (0.809 - 1.320) (0.750 - 1.076) (0.807 - 1.319)

≥ 3 times/week 0.785** 0.775 0.792** 0.774

(0.629 - 0.979) (0.552 - 1.088) (0.635 - 0.989) (0.550 - 1.088)

Constant 0.189*** 0.176***

(0.129 - 0.278) (0.114 - 0.271)

Observations 12,681 5,992 12,737 6,025

Number of respondents 3,359 3,359

0.0061

1) Level of the GHQ 12 score (0: ≤ 3 points, 1: ≥ 4 points)

2) Robust cieform in parentheses

3) *** p<0.01, ** p<0.05, * p<0.1

4) Adjusted odds ratios (AOR) with 95% CI (adjusted for sex, age, number of persons in the household (2 categories: one person or ≥　2 people), employment status, disposable income per household, sleep duration in weekdays, physical exercise, smoking habit and drinking alcohol).

5) Adjusted odds ratios with 95% CI (adjusted for sex, age, number of persons in the household (7 categories: 1 person, 2 people, 3 people, 4 people, 5 people, 6 people, and 7 people and more), employment status, sleep duration in weekdays, physical exercise, smoking habit and drinking alcohol).

6) Refe: Reference 1

8. The method of measuring health behaviours seems fair from a health perspective, except for the question on drinking. This question is not comparable to other studies on drinking habits/alcohol use and does not seem relevant. To drink a beer three times a week is not considered a drinking problem in most OECD countries. It may be different in Japan, but then it needs explanation.

Response:

There are various approaches used to determine drinking habits. It would be ideal to ask the amount of alcohol directly and obtain an answer for the amount of alcohol content in a beverage. However, almost people do not know the alcohol content in liquor. Further, as it is well known that the alcohol content in beer varies, nutritionists attempted to develop simple questions such as frequency of alcohol intake1). Asking for the frequency of alcohol intake is a common method in nutrition. Compared to other approaches, such as asking for the amount of alcohol in 24 hours, this approach has some benefits. Some studies on alcohol content and frequency of alcohol intake have already been conducted2,3). In particular, the National Health and Nutrition Survey conducted by the Japanese government used questions about the frequency of alcohol intake. Researchers have been attempting to validate the food frequency questionnaire (FFQ)3). Therefore, we believe that this question is comparable to other studies and relevant to measure alcoholism of a person by using self-administrated questionnaire in this study.

References:

1: Willett W. Food frequency methods. Nutr Epidemiol. Third edition. Oxford, New York: Oxford University Press; 2012.

2:Sadakane A, Gotoh T, Ishikawa S, Nakamura Y, Kayaba K, Jichi Medical School Cohort Study G. Amount and frequency of alcohol consumption and all-cause mortality in a Japanese population: the JMS Cohort Study. Journal of epidemiology. 2009;19(3):107-15.

3: Harmouche-Karaki M, Mahfouz M, Obeyd J, Salameh P, Mahfouz Y, Helou K. Development and validation of a quantitative food frequency questionnaire to assess dietary intake among Lebanese adults. Nutrition Journal. 2020;19(1):65.

9. The chosen statistical methods are unconventional.

Response:

Thank you for the comment. However, we believe that the fixed-effects or random-effects model based on panel data are becoming standard statistical methods. These methods have been commonly and frequently used in economics literature and are used in other fields recently, including public health. For example, many published scientific paper used the fixed-effects and random-effects models to determine the association among potential risk factors in public health, such as the following.

References:

1. Wang S, Mak HW, Fancourt D. Arts, mental distress, mental health functioning & life satisfaction: fixed-effects analyses of a nationally-representative panel study. BMC public health. 2020;20(1):208.

2. Wang J, Xu J, An R. Effectiveness of backward walking training on balance performance: A systematic review and meta-analysis. Gait & Posture. 2019;68:466-75.

3. Oshio T, Inoue A, Tsutsumi A. Associations among job demands and resources, work engagement, and psychological distress: fixed-effects model analysis in Japan. Journal of occupational health. 2018;60(3):254-62.

4. Oshio T. The association between individual-level social capital and health: cross-sectional, prospective cohort and fixed-effects models. Journal of epidemiology and community health. 2016;70(1):25.

5. Jin Q, Shi G. Meta-Analysis of SNP-Environment Interaction with Heterogeneity. Hum Hered. 2019;84(3):117-26.

6.　　　Kuroda S, Yamamoto I. Why Do People Overwork at the Risk of Impairing Mental Health? Journal of Happiness Studies. 2019;20(5):1519-38.

7.　　　Ocean N, Howley P, Ensor J. Lettuce be happy: A longitudinal UK study on the relationship between fruit and vegetable consumption and well-being. Social science & medicine. 2019;222:335-45.

8. Oshio T. The association between individual-level social capital and health: cross-sectional, prospective cohort and fixed-effects models. Journal of epidemiology and community health. 2016;70(1):25.

Results

10. Here 14.717 participants are presented as shown in table two. What to believe?

Response:

Would you please check the number in total (bottom of right corner) in Table 1? In the Abstract and Methods section, we wrote that 14,717 implies 14,717 observations/participants/questionnaires in total for 5-year questionnaire survey from 2014 to 2018. In econometrics, this is called 14,717 observations.

We added the sentence as follows: ‘In total of this study, we used data from 14,717 observations of 3,501 individuals aged 22 to 59.’

11. 1) Table 2: %SD – do you mean %? 2) Very few women have regular employment and drink alcohol three times a week, very few respondents have a low household income. It needs an explanation – the numbers after “in Japanese Yen” what are they? 3) The “0 points – 12 points” at GHQ, what is that? 4) An average? 5) The GHQ- score is a result pooled over the years I guess, please write this, if so. 42% women have poor mental health? It makes you wonder if GHQ valid in this context? Please add reference on the validation of GHQ-12 in a Japanese context in the method section.

Response:

1) Thank you for the comment. %SD implies % and SD. We changed the layout accordingly as follows:

Table 2. Participants’ demographic characteristics.

Total Men Women

Variables Group n % N % n %

Wave

Men 7,215 49.0%

Women 7,502 51.0%

Number of persons in the household

≥ 2 people 13,487 92.1% 6,415 89.4% 7,072 94.7%

1 person 1,155 7.9% 757 10.6% 398 5.3%

Employment status

Unemployed 1,885 12.9% 271 3.8% 1,614 21.7%

Self-employed 1,818 12.4% 1,106 15.4% 712 9.6%

Regular employee 7,126 48.7% 5,332 74.3% 1,794 24.1%

Non-regular employee 3,796 26.0% 466 6.5% 3,330 44.7%

Disposable income per household

≥ 6,000 K 4,856 36.9% 2,424 37.2% 2,432 36.6%

2,000 K-< 6,000 K 7,373 56.0% 3,720 57.1% 3,653 54.9%

< 2,000 K 935 7.1% 370 5.7% 565 8.5%

Sleep duration. weekdays

≥ 7 h 5,450 37.6% 2,710 38.2% 2,740 37.0%

6–7 h 5,595 38.6% 2,769 39.1% 2,826 38.2%

< 6 h 3,448 23.8% 1,607 22.7% 1,841 24.9%

Physical exercise

≥ 3 days/week 1,415 9.7% 729 10.2% 686 9.2%

≤ 2 days/week 2,637 18.0% 1,518 21.2% 1,119 15.0%

No exercise 10,568 72.3% 4,914 68.6% 5,654 75.8%

Smoking habit

Never 7,888 53.7% 2,524 35.0% 5,364 71.6%

Quit 3,423 23.3% 2,220 30.8% 1,203 16.1%

Sometimes + everyday 3,379 23.0% 2,459 34.1% 920 12.3%

Drinking alcohol habit

Never 5,192 35.4% 1,746 24.3% 3,446 46.2%

≤ 2 times/week 5,130 35.0% 2,415 33.6% 2,715 36.4%

≥ 3 times/week 4,328 29.5% 3,027 42.1% 1,301 17.4%

GHQ score

≥ 4 points (poor) 5,713 39.1% 2,581 36.1% 3,132 42.0%

≤ 3 points 8,891 60.9% 4,562 63.9% 4,329 58.0%

Mean (SD) Mean (SD) Mean (SD)

Age (years old) 45.3 (8.7) 45.2 (8.7) 45.4 (8.7)

Disposable income per household 5,428K (3,113K) 5,422K (2,959K) 5,433K (3,256K)

GHQ score 3.4 (3.4) 3.2 (3.4) 3.6 (3.4)

2) It was a unit of Japanese currency, and we deleted it.

3) We deleted ‘The 0 points – 12 points’.

4) It was not an average. It was a range of the GHQ score from 0 points to 12 points.

5) We deleted Figure 1 and made Table 3 as follows:

Table 3. Differences of GHQ scores of 4 points or more by sex, per year and pooled data.

Years Females Males P value 1)

2014 41.9% 35.8% ***

2015 41.9% 35.7% ***

2016 41.8% 37.9% *

2017 43.6% 36.9% ***

2018 40.7% 34.3% **

Pooled data 42.0% 36.1% n.s.

Table 3 shows the differences of GHQ scores of 4 points or more by sex per year and pooled data. As the reviewer recommended, we added ‘pooled data’ in the sentence as follows:

‘In terms of pooled data of participants’ mental health conditions, 36.1% of men and 42.0% of women were shown to have poor mental health conditions (≥ 4 GHQ-12 score).’

According to the results of this study, 42% of female subjects obtained 4 or more points of GHQ. This implies 42% of women have poor mental health conditions. The results during 5 years, around 40% of women showed 4 and more points consistently. We have not found any reference in which GHQ is not valid for Japanese people. For example, Hori et al. wrote as follows1):

The GHQ-12 is a widely used, self-administered questionnaire that was originally designed as a screening tool for mental illness. The GHQ is also used in primary health care screening in the general population survey2). The GHQ-12 score was first applied to adults and then validated for adolescents as well3). Every item on the GHQ-12 describes a symptom and has four possible responses: the two answers which indicate the absence of the symptom are given a score of 0, and the other two which indicate the presence of the symptom receive a score of 1. The overall score on the scale will fall into a range of 0–12, with higher scores indicating more psychological distress. Good mental health was defined as a GHQ-12 score <4 and poor mental health as a score ≥4 4).

We added the sentences as follows into the method section:

‘Previous study reported that they assessed the factor structure of the GHQ-12 for the Japanese general adult population. Data came from a sample of 1808 Japanese aged 20 years or older who were randomly selected based on the 1995 census (897 men and 911 women). Cronbach’s α coefficients of GHQ-12 were 0.83 for Japanese men and 0.85 for Japanese women (32).’

References:

1. Hori D, Tsujiguchi H, Kambayashi Y, Hamagishi T, Kitaoka M, Mitoma J, et al. The associations between lifestyles and mental health using the General Health Questionnaire 12-items are different dependently on age and sex: a population-based cross-sectional study in Kanazawa, Japan. Environ health prev med. 2016;21(6):410-21.

2. Doi Y, Minowa M. Factor structure of the 12-item General Health Questionnaire in the Japanese general adult population. Psychiatry Clin Neurosci. 2003;57(4):379-83.

3. Baksheev GN, Robinson J, Cosgrave EM, Baker K, Yung AR. Validity of the 12-item General Health Questionnaire (GHQ-12) in detecting depressive and anxiety disorders among high school students. Psychiatry research. 2011;187(1-2):291-6.

4. Goldberg DP, Oldehinkel T, Ormel J. Why GHQ threshold varies from one place to another. Psychological medicine. 1998;28(4):915-21.

12. It is not mentioned in the method section, but the adjustments are done only for other variables than the one presented/analysed, isn’t it?

Response:

Yes, it is. We have tried some adjustments with some models, but we presented only the model.

13. 1) Table 3: The table is difficult to read. I would suggest the FE/RE was stated in the heading section.

2) Bold is stated to indicate statistically significant findings but seems used at random. 3) How can an AOR for RE-female alcohol >= 3 times a week at 0.752 (0.547 – 1.179) be significant? 4) The non-significant results for men and women end up being 0.7885 (0.629 – 0.979)? 5) Why are there fewer observations for the FE-analysis than the RE-analysis?

Response:

1) About the table design, we changed the layout based on your suggestions as follows:

Table 4. Estimated associations between participants’ General Health Questionnaire 12-item scores and risk factors by gender based on the random-effects conditional logistic regression models, on the fixed-effects conditional logistic regression models and on the Hausman tests.

All samples Men Women

AOR(C.I.) 1) AOR(C.I.)2) AOR(C.I.)2) AOR(C.I.)2) AOR(C.I.)2) AOR(C.I.)2)

Model Type RE FE* RE FE* RE* FE

Sex Men Ref.

Women 1.419***

(1.107 - 1.819)

Age Under 39 Ref. Ref. Ref. Ref. Ref. Ref.

40-49 0.971 1.117 0.857 1.121 1.146 1.120

(0.791 - 1.190) (0.815 - 1.532) (0.623 - 1.179) (0.692 - 1.815) (0.880 - 1.492) (0.738 - 1.700)

50-59 0.836 1.082 0.766 0.935 0.936 1.184

(0.666 - 1.049) (0.693 - 1.691) (0.538 - 1.092) (0.474 - 1.846) (0.697 - 1.256) (0.653 - 2.149)

Number of persons in the household ≥ 2 people Ref. Ref. Ref. Ref. Ref. Ref.

One person 1.286 0.999 1.422* 1.085 0.994 0.905

(0.946 - 1.749) (0.634 - 1.575) (0.946 - 2.139) (0.612 - 1.924) (0.620 - 1.594) (0.430 - 1.904)

Employment status unemployed 1.501*** 1.727** 8.035*** 5.852*** 0.840 1.097

(1.122 - 2.008) (1.134 - 2.629) (4.219 - 15.30) (2.376 - 14.41) (0.599 - 1.180) (0.635 - 1.896)

self-employed 0.832 1.193 0.896 1.134 0.608** 0.937

(0.626 - 1.105) (0.727 - 1.958) (0.605 - 1.327) (0.563 - 2.286) (0.402 - 0.920) (0.451 - 1.944)

regular employee Ref. Ref. Ref. Ref. Ref. Ref.

non-regular employee 0.983 0.958 1.332 1.083 0.699** 0.764

(0.774 - 1.247) (0.680 - 1.350) (0.808 - 2.196) (0.565 - 2.077) (0.527 - 0.928) (0.485 - 1.203)

Disposable income per household ≥ 6,000K Ref. Ref. Ref. Ref. Ref. Ref.

2,000K-< 6,000K 1.254*** 1.118 1.150 0.970 1.386*** 1.273*

(1.075 - 1.464) (0.924 - 1.352) (0.913 - 1.448) (0.728 - 1.293) (1.126 - 1.705) (0.986 - 1.644)

< 2,000K 1.977*** 1.441* 1.714** 1.382 1.980*** 1.464

(1.475 - 2.650) (0.993 - 2.092) (1.082 - 2.713) (0.782 - 2.440) (1.351 - 2.901) (0.881 - 2.434)

Sleep duration of the respondent ≥ 7 hours Ref. Ref. Ref. Ref. Ref. Ref.

6-7 hours 1.100 1.128 0.939 0.988 1.277** 1.285**

(0.951 - 1.272) (0.952 - 1.335) (0.752 - 1.173) (0.761 - 1.282) (1.054 - 1.547) (1.028 - 1.606)

< 6 hours 1.644*** 1.503*** 1.613*** 1.456** 1.662*** 1.500**

(1.355 - 1.993) (1.177 - 1.919) (1.203 - 2.161) (1.012 - 2.095) (1.283 - 2.151) (1.070 - 2.104)

Physical exercise ≥ 3 days/week Ref. Ref. Ref. Ref. Ref. Ref.

≤ 2 days/week 1.175 1.066 1.128 0.979 1.265 1.184

(0.897 - 1.538) (0.792 - 1.435) (0.768 - 1.656) (0.646 - 1.485) (0.864 - 1.851) (0.767 - 1.828)

No exercise 1.637*** 1.418** 1.747*** 1.531** 1.540** 1.326

(1.284 - 2.088) (1.062 - 1.895) (1.227 - 2.488) (1.013 - 2.315) (1.099 - 2.160) (0.874 - 2.012)

Smoking habit Never Ref. Ref. Ref. Ref. Ref. Ref.

Quit 1.135 0.929 1.119 1.054 1.176 0.856

(0.909 - 1.419) (0.579 - 1.490) (0.794 - 1.578) (0.507 - 2.193) (0.880 - 1.570) (0.476 - 1.540)

Sometimes + everyday 1.217 0.940 1.201 1.176 1.242 0.714

(0.950 - 1.560) (0.520 - 1.700) (0.850 - 1.697) (0.503 - 2.750) (0.854 - 1.805) (0.302 - 1.685)

Drinking alcohol habit Never

≤ 2 times/week 0.893 1.033 0.817 0.928 0.945 1.073

(0.745 - 1.069) (0.809 - 1.320) (0.602 - 1.108) (0.609 - 1.413) (0.758 - 1.179) (0.793 - 1.452)

≥ 3 times/week 0.785** 0.775 0.783 0.679 0.752 0.819

(0.629 - 0.979) (0.552 - 1.088) (0.564 - 1.087) (0.407 - 1.133) (0.547 - 1.035) (0.503 - 1.334)

Constant 0.189*** 0.201*** 0.309***

(0.129 - 0.278) (0.116 - 0.350) (0.188 - 0.506)

Observations 12,681 5,992 6,270 2,748 6,411 3,244

Number of respondents 3,359 1,657 1,702

Hausman Test 0.0061 0.0586 0.4554

Bold ratios: the significant results of Hausman’s test.

*** p<0.01, ** p<0.05, * p<0.1

2) Response:

Thank you for the comment. This was my careless mistake. Bold does not show ‘statistically significant results’. Bold shows the results of Hausman test. We corrected it.

3) Response:

Thank you for finding this careless mistake. We corrected the result of the AOR for RE-female alcohol >= 3 times a week as no significant one.

4) Response:

The results ‘0.785** (0.629 – 0.979)’ is a significant result, as we put ‘**’. In addition, it is 0.785, not 0.7885.

5) Response:

In estimation, the fixed-effects model uses only information of the observations that experienced changes within an individual. In other words, fixed-effects model removes the observations doesn't change over time. That is why the number of observations in FE and RE models are different.

Discussion

14. Limitations. Dropouts are discussed in general terms. What was the actual characteristics of the dropouts in this study? The income variable is very screwed – is that sampling bias or at true reflection of the study population?

Response:

As we wrote that ‘the dropouts might have had unhealthy outcomes or have been in unfavourable situations’, there is a possibility of sample attrition bias and dropouts. However, we explained our attrition rate in Table 1, indicating less than 10% attrition rates for 5 years. This is the same as previous studies, which revealed that the average attrition rate of panel studies revolves around 10%1). We believe that the attrition bias is not so serious, because of the low attrition rates of JHPS.

Regarding income variable, as we noted in the reply to your comment 6, although the income distribution of our sample is not necessarily similar to the population, we confirmed that the results of weighted regression using the inverse of the number of households as weight were not so much different from the original results.

Reference:

1) Baltagi BH. Econometric analysis of panel data. UK: John Wiley & Sons Ltd; 2005. p. 1-9.

15. Information bias is mentioned only as recall. However, the instrument is not (or is?) validated; again, when adjusting for the number of persons in household – which is very good -the variable is not covering that, but only one or more than one. This is a serious problem and makes any conclusion related to income invalid.

Response:

As we noted in the reply to your comment 7, we confirmed that the estimation results did not change significantly even if we used the dummy variables indicating the size of household: a one-person, two-person, three-person, and four or more person dummies. Furthermore, as we noted in the replies to your comment 6, we confirmed the results of weighted regression using the inverse of the number of households as weight were not so much different from the original results. Therefore, we are afraid that this might not be a serious problem to make a conclusion about income.

16. “Further research is needed to identify the effects of disposable income and living status on mental health conditions, as well as the role of gender”. They do exist in plenty – as do so for employment status and gender.

Response:

As you mentioned, in the next study, I would like to analyse unemployed and homemaker females separately. This would be an issue in Japanese gender studies.

17. You do not reflect on the high prevalence of poor mental health, why? it is a central study objective. However, to examine the prevalence of poor mental health longitudinal data are not optimal, unless you want to give information on the development in mental health. Again, it is not evident if GHQ is validated in a Japanese population – and a pooled prevalence of 42% women in poor mental health does not seem reliable.

Response:

Thank you for your comment. We added the sentence into discussion as follows:

‘The results of GHQ-12 indicated that 36.1% of men and 42.0% of women were shown to have poor mental health conditions (≥ 4 GHQ-12 score). The results indicated statistically significant differences between men and women for five years and pooled data of all waves. Hori et al. reported that 24.2% of men and 32.0% of women were shown to have GHQ score with 4 points or more (36). The prevalence of poor mental health conditions in this study was higher than those of Hori’s study. However, Matsuba et al. reported that 43.2% of Japanese people in Thailand indicated poor mental health conditions (37). According to the results of previous studies, scores for the female respondents tended to be higher than those for the male respondents. Further research is needed to determine the cause of the high prevalence rate.’

As we wrote, Matsuba et al. also reported that 43.2% of Japanese people indicated poor mental health conditions. It would be inappropriate to say that a pooled prevalence of 42% women in poor mental health does not seem reliable.

References:

1. Hori D, Tsujiguchi H, Kambayashi Y, Hamagishi T, Kitaoka M, Mitoma J, et al. The associations between lifestyles and mental health using the General Health Questionnaire 12-items are different dependently on age and sex: a population-based cross-sectional study in Kanazawa, Japan. Environ health prev med. 2016;21(6):410-21.

2. G. Matsuba, D. Suzuku, Y. Inaba. Physical practice and mental stress among Japanese people living in Thailand Juntemdo Igaku. 2007;53(4):581-7.

18. As for household income as socioeconomic index the validity is poor – first very few are in lowest category and even though it is stated the analyses are adjusted for number of persons in the household, when in fact it is only adjusting for one or more than one.

Response:

Thank you for the comment. However, we do not think that more than 350 subjects are very few as one category, while data analysis with 2 methods worked properly. As we wrote why we made three groups of disposable income for the comment 6, the categorisation method is the same as the government survey, National Survey Health and Nutrition (Ministry of Health, Labour and Welfare, Japan) in 2014. Moreover, there are previous studies in public health, which used a variable ‘household annual income’. Lei et al. (2020) included a model average household income (<1500, 1500-3000, 3000-6000, 6000-9000, <9000) 1). We also published two research papers with the variable household annual income 2, 3). As we explained several times here, the key points of our study were the relationship between whether people live alone or live with somebody and mental health conditions and impact of household annual income on mental health conditions. Please understand that as we noted in the reply to your comment 7, we confirmed that the estimation results did not change significantly even if we used the dummy variables indicating the size of household: a one-person, two-person, three-person, and four or more person dummies.

References:

1. Lei L, Huang X, Zhang S, Yang J, Yang L, Xu M. Comparison of prevalence and associated factors of anxiety and depression among people affected by versus people unaffected by quarantine during the COVID-19 epidemic in Southwestern China. Med Sci Monit. 2020;26:e924609.

2. Muto K, Yamamoto I, Nagasu M, Tanaka M, Wada K. Japanese citizens' behavioral changes and preparedness against COVID-19: An online survey during the early phase of the pandemic. PloS one. 2020;15(6):e0234292.

3. Nagasu M, Kogi K, Yamamoto I. Association of socioeconomic and lifestyle-related risk factors with mental health conditions: a cross-sectional study. BMC public health. 2019;19(1):1759.

19. As for the FE logistic regression analyses vs RE logistic analyses none of them account for the time each individual contributes with in the time series (person-years), and thus less accurate than the traditional methods used for longitudinal studies.

Response:

Thank you for the comment. As we wrote above, the FE logistic regression analyses and the RE logistic analyses are conventional statistical methods in economics, but these have not been very common in public health yet. But Ocean et al applied fixed effect model about a longitudinal UK study on the relationship between fruit and vegetable consumption and well-being.1)

The FE logistic regression accounts for the within-person changes while the RE logistic regression does for within-person changes as well as between-person differences, so they should be more accurate than the traditional logistic regressions.

Reference:

1) Ocean N, Howley P, Ensor J. Lettuce be happy: A longitudinal UK study on the relationship between fruit and vegetable consumption and well-being. Social science & medicine. 2019;222:335-45.

Authors’ response to reviewers

PLOS ONE

PONE-D-20-10688

Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan

Reviewer #2:

In general, this study is well constructed to gain a rational outcome. I have some minor comments to be addressed for better understanding of the study.

1. In the introduction section, the author mentioned that Japan was shown to be one of the countries with the highest suicide rates in the world. It is partially acceptable, but a little over-represented. Some Eastern European nations as well as Russia has higher suicide rate. Also, in latest statistics, those of the US and Sweden are not so different from Japan's. I am afraid that outdated articles the author referred can be misleading.

Response:

Thank you for your comment. We deleted the references cited and added new data, which was provided by the OECD. The sentence was also rewritten as follows:

‘According to data released by the OECD, the suicide rate was the top sixth in 34 countries, although the number of suicides committed by Japanese people has been declining gradually for 10 consecutive years (10).’

Reference: OECD (2020), Suicide rates (indicator). doi: 10.1787/a82f3459-en (Accessed on 03 June 2020)

2. In this study, the author classified some items the participants answered into some groups. 1) How the author decide the thresholds of each group? 2) For example, are there any reasons that people taking 6-7 hours of sleep in a day is different to those taking >7 hours sleep, not >8? In my understanding, there are rich evidence suggesting sleeping under 6 hours in a day is harmful. But how many hours you should sleep is controversial. 3) Also, the author referred a Glozier's work. But its subjects were limited to young people recommended to take 8-9 hours sleep. The author should consider to show better preceding studies.

Response:

Thank you for the comment. For the comments 1) and 2), we have some reasons to categorise the variables into some groups. As stated in our study objectives, examining the prevalence of poor mental health conditions and to analyse the models by gender is important. We thought that analysis across gender would be more important than that using all samples simultaneously as a mental health study. Regarding the categorical variables, we divided the items into three groups to preserve sufficient samples in each group across gender as many as possible. For example, the variable about sleeping hours, this needed to be divided into 3 groups (about 30% each) for analysing the models as Table 1 shows the results. If we categorised 4 groups (≥ 8 h, 7-8 h, 6-7 h, and < 6 h), the sample size of over 8 sleeping hours was 7.2% of female subjects (n=535) and 8.7% of male subjects (n=619). This was not enough to analyse the models by gender unfortunately.

About the average of sleeping hours among Japanese population, the Sleep Guidelines for Health Promotion 2014 published by the Ministry of Health, Labour and Welfare (MHLW) reported that about 60% of Japanese adults sleep between 6 hours and less than 8 hours and not over 8 hours; this is considered standard sleep duration. Our data showed similar trend with the report. As you recommended, analysing the association between long time sleep and mental health would be valuable. If we can publish a paper about the relationship between mental health and long sleeping hours, we will consider the sample size.

About the comment 3), we cited one more reference and added a sentences as follows:

‘For example, Golzier et al. analysed a cohort study and reported that shorter sleep duration is linearly associated with psychological distress in young adults (14). Lallukka et al. also reported that the combination of sleep duration and quality was associated with physical, emotional, and social functioning among Australian adults (15). ’

Table 1. Results of sleeping hours by gender.

Gender Total

Female Male

Sleeping hours ≥7 hours n 2740 2710 5450

% 37.0% 38.2% 37.6%

6-7 hours n 2826 2769 5595

% 38.2% 39.1% 38.6%

< 6 hours n 1841 1607 3448

% 24.9% 22.7% 23.8%

Total n 7407 7086 14493

% 100.0% 100.0% 100.0%

3. In this study, only 7.9% of the participants lived alone. According to recent official statistics in Japan, one-fourth of the household is composed of one person. In my calculation, single person household was 10.7% in 2014. Did the author compare the demographic data with those of contemporary official statistics? If there is a large discrepancy between them, the representativeness of the panel is doubtful. The same thing is also adapted to employment status, but it seems consistent with the official statistics, as far as my checking.

Response:

Thank you for your comment. We compared the demographic data of JHPS with those of the national census taken on Oct. 2015. As you highlighted, the result of JHPS seems to have considerably fewer participants living alone than the results of the national census. As you noticed, the participants of JHPS were selected at the individual level, and those at the national census were at the household level. We also calculated the composition of single-person living at the individual level as 14.4% (the number of single person household is 18,420, which amounts to 14.4% of the Japanese population of 128,000 thousands). Therefore, there is about a double difference between population (14.4%) and our sample (7.9%).

To account for the possible bias due to the smaller sample of single-person, we conducted a weighted logit regression using the inverse of the number of households as weight. As shown in the table below, although some of the statistical significance are different, the sign (more than 1 or not) and magnitude of the odds ratio seem not to be so different between the estimates with and without weight. Thus, we regard that although there is a sampling bias in our data in the sense that less persons living alone were selected, it does not largely affect the estimation results.

Table Results of RE and estimation with weight

Results of RE　 Results of estimation with weight

odds ratios odds ratios

Sex Men Refe Refe

Women 1.419*** 1.236***

(1.107 - 1.819) (1.122 - 1.362)

Age Under 39 Refe Refe

40-49 0.971 0.951

(0.791 - 1.190) (0.866 - 1.045)

50-59 0.836 0.861***

(0.666 - 1.049) (0.782 - 0.948)

Number of Persons in the household ≥ 2 people Refe Refe

One person 1.286 1.248***

(0.946 - 1.749) (1.091 - 1.428)

Employment status unemployed 1.501*** 1.165**

(1.122 - 2.008) (1.022 - 1.328)

self-employed 0.832 0.832***

(0.626 - 1.105) (0.737 - 0.940)

regular employee Refe Refe

non-regular employee 0.983 0.987

(0.774 - 1.247) (0.887 - 1.098)

Disposable income per household ≥ 6,000 K Refe Refe

2,000 K-< 6,000 K 1.254*** 1.198***

(1.075 - 1.464) (1.106 - 1.299)

< 2,000 K 1.977*** 1.770***

(1.475 - 2.650) (1.517 - 2.064)

Sleep duration of the respondent ≥ 7 h Refe Refe

6-7 h 1.100 1.028

(0.951 - 1.272) (0.945 - 1.119)

< 6 h 1.644*** 1.418***

(1.355 - 1.993) (1.288 - 1.561)

Physical exercise ≥ 3 days/week Refe Refe

≤ 2 days/week 1.175 1.228***

(0.897 - 1.538) (1.054 - 1.429)

No exercise 1.637*** 1.449***

(1.284 - 2.088) (1.270 - 1.652)

Smoking habit Never Refe Refe

Quit 1.135 1.076

(0.909 - 1.419) (0.979 - 1.183)

Sometimes + everyday 1.217 1.126**

(0.950 - 1.560) (1.020 - 1.243)

Drinking alcohol Never Refe Refe

≤ 2 times/week 0.893 0.879***

(0.745 - 1.069) (0.806 - 0.959)

≥ 3 times/week 0.785** 0.894**

(0.629 - 0.979) (0.812 - 0.985)

Constant 0.189*** 0.360***

(0.129 - 0.278) (0.301 - 0.431)

Observations 12,681 12,681

Number of respondents 3,359

*** p<0.01, ** p<0.05, * p<0.1

To explain the above points, we added the following sentences:

‘First, for the proportion of single person household, we compared the demographic data of this study with those of the national census taken on Oct. 2015. The result of this study seems to have considerably fewer participants lived alone than the results of the national census. Because the participants of this study were selected at the individual level, that of the national census were at the household level. We also calculated the composition of single-person living at the individual level as 14.4% (the number of single person household is 18,420, which amounts to 14.4% of the Japanese population of 128,000 thousands). Therefore, there is about a double difference between population (14.4%) in the national census and our sample (7.9%). To account for the possible bias due to the smaller sample of single-person, we conducted a weighted logit regression using the inverse of the number of households as weight. Although there are some statistical significance between RE-CLR and RE-CLR with weight, the sign (more than 1 or not) and magnitude of the odds ratios seem not to be so different between the estimates with and without weight. Thus, we regard that although there is a sampling bias in our data in the sense that less persons living alone were selected, it does not largely affect the estimation results. ’

4. Table 2 is difficult to read at a glance because each number and the mean share the same column. The author should rearrange it.

Response:

Thank you for your suggestions. The number of the subject and the mean are rearranged in Table 2.

5. Is the mean income of Japanese only 542.8 JPY?

Response:

Thank you very much for highlighting this. It was my mistake. I corrected the numbers in Table 2.

6. What does "0 point - 12 points" mean in the GHQ score section?

Response:

This meant that the range of GHQ scores was 0 point to 12 points. However, I deleted it.

7. Figure 1. is hardly understandable and space-killing. What does the vertical axis mean? I reckon it shows the percentage. Anyway, it seems better to choose another type of graph (polygonal line graph or a chart maybe).

Response:

Thank you for your suggestions. We tried to make another graph, but the results were not suitable to be presented as a graph. Therefore, we deleted the figure and made a table labelled as Table 2.

Table 3. Differences in rates of GHQ scores of 4 points or more from 2014 to 2018.

Years Females Males P value 1)

2014 41.9% 35.8% ***

2015 41.9% 35.7% ***

2016 41.8% 37.9% *

2017 43.6% 36.9% ***

2018 40.7% 34.3% **

Pooled Data 42.0% 36.1% n.s.

1) ***: p < .001, **: p < .005, *: p < .05.

2) The levels of the General Health Questionnaire 12-item scores: Poor = ≥ 4 points, Fine = ≤ 3 points.

Authors’ response to reviewers

PLOS ONE

PONE-D-20-10688

Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan

Reviewer #3:

This epidemiological survey is important for mental health field, worth reading. The authors however should 1) propose the hypothesis of the study clearly, and 2) emphasize the new findings in the present study, and 3) the differences from the previous reports throughout the manuscript.

Response:

Thank you very much for reviewing our manuscript. We have made the suggested changes, added sentences and rewritten many parts of the manuscript.

1) We have stated our hypothesis of this study as follows:

‘We hypothesised that better SES and healthy lifestyle-related factors may associate with better mental health conditions positively.’

2 &3) We have rewritten many parts of the abstract and added the aims clearly. The results of this study also added into the abstract:

‘This study aims to reveal the prevalence of poor mental health conditions among Japanese individuals and to identify the SES- and lifestyle-related risk factors that might lead to these conditions.’

‘The prevalence of poor mental health conditions, represented by a GHQ-12 score of 4 or more, was 36.1% and 42.0% of men and women, respectively.’

‘Various factors, such as unemployment, low household income, short nightly sleeping duration, and lack of exercise, showed significant longitudinal (within) associations with mental health conditions estimated by the FE-CLR models.’

Introduction is rewritten as follows:

‘The fixed-effects regression models are used to adjust for all time-constant unobserved confounders and decrease the risk of omitted variable bias, as well as adjust for identified time-varying confounders. ’

Discussion is also rewritten as follows:

‘The results of GHQ-12 indicated that 36.1% of men and 42.0% of women were shown to have poor mental health conditions (≥ 4 GHQ-12 score). The results indicated statistically significant differences between men and women for five years and pooled data of all waves. Hori et al. reported that 24.2% of men and 32.0% of women were shown to have GHQ score with 4 points or more (36). The prevalence of poor mental health conditions in this study was higher than those of Hori’s study. However, Matsuba et al. reported that 43.2% of Japanese people in Thailand indicated poor mental health conditions (37). According to the results of previous studies, scores for the female respondents tended to be higher than those for the male respondents. Further research is needed to determine the cause of the high prevalence rate.’

Submitted filename: 200910.Response_to_Reviewers.26.docx

Click here for additional data file.

23 Sep 2020

Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan.

PONE-D-20-10688R1

Dear Dr. Nagasu,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Kenji Hashimoto, PhD

Section Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Akihiro Shiina

25 Sep 2020

PONE-D-20-10688R1

Impact of socioeconomic- and lifestyle-related risk factors on poor mental health conditions: A nationwide longitudinal 5-wave panel study in Japan

Dear Dr. Nagasu:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Kenji Hashimoto

Section Editor

PLOS ONE