Latent profile analysis of lifestyle characteristics and health risk behaviors among Koreans who have completed industrial accident care.

This study investigated the characteristics and health behavior profiles of 1,803 workers who had experienced industrial accidents. Average weekly exercise days, average number of cigarettes smoked per day, average daily sleep duration, and number of days of alcohol consumption were selected to investigate health behavior profiles. Specifically, latent profile analysis was applied to identify the health behavior profiles of people who had completed industrial accident care; the latent classes were the health-conscious type (n=240), the potential-risk type (n=850), and the high-risk type (n=713). Comparison of the health-conscious and potential-risk types indicated that younger subjects, the employed, and those with lower social status and life satisfaction were more likely to be the potential-risk type. Comparison of the health-conscious and high-risk types revealed that males, younger subjects, the employed, those without chronic illnesses, and those with lower social status and life satisfaction were more likely to be the high-risk type. The results suggest that industrial accident victims who have completed accident care have different health behaviors and it is necessary to improve health promotion based on health type characteristics.

Introduction

Over the last 10 yr, more than 90,000 Korean workers have been injured. Industrial welfare services make it possible for workers to return to stable lives after industrial accidents, and for employers to preserve their scarce labor forces, through a variety of rehabilitation services, and this rehabilitation is important as it restores workers and fulfills corporate social responsibilities. However, in spite of the importance of workers’ return to the workplace, South Korea’s worker retirement system is not actively operated, in contrast with advanced countries that have mandatory retirement systems. The interest in industrial accidents and demands for rational changes in rehabilitation services for industrial accident victims, including earlier return to work, are driving changes in policies and institutions. However, there are no statistics that can provide in-depth and dynamic information on workers who have suffered industrial accidents1).

To resolve these problems, the Korea Workers’ Compensation and Welfare Service (KWCWS) has conducted the Panel Study of Workers’ Compensation Insurance (PSWCI)1) since 2013 to provide accurate statistics on industrial accident victims including their overall socioeconomic activity, physical characteristics and abilities, environmental factors, and other factors following accidents. Investigation of persistent health behaviors of those who have completed industrial accident care is important for future national health care policies and costs; analysis of this population will determine the length and quality of care and health education programs.

As known from researchers, behaviors such as physical inactivity, alcohol consumption, and smoking are high-cost2), major contributors to chronic disease and death3). Regular engagement in preventive health behaviors such as physical activity and avoiding harmful health behaviors such as tobacco smoking reduce the risk of premature death, acute injury and illness, and some chronic diseases4). People who engage in healthy behaviors have a low incidence of functional disability, and the risk of disability decreases as the number of healthy behaviors increases5). As in the general population, factors that affect the health of those who have completed industrial accident care include average weekly exercise days, average daily sleep duration, average number of cigarettes smoked per day, and number of days of usual alcohol consumption. Variables such as exercise, sleep, smoking, and alcohol consumption are important determinants of health conditions6). Health behaviors including exercise, smoking, and alcohol consumption are greatly interrelated, and exercise7, 8), smoking9, 10), and alcohol consumption11, 12) are potential influencing factors for osteoporotic fractures. In addition, alcohol consumption, lack of exercise, and smoking are associated with headache, and poor sleep and stress are often cited as the cause of headache13, 14). In high-risk patients with cardiovascular disease, primary prevention and risk stratification have been advocated, such as moderate alcohol consumption, increased physical activity, smoking cessation, and weight management15). In a study of male factory workers, sleeping less than seven hours a day, working more than nine hours a day, smoking, alcohol consumption, and not performing exercise were not related to depression16).

In addition, recent policy studies have tended to focus on one side of a given behavior. For example, alcohol policy research is relatively isolated compared with tobacco policy research17, 18). However, lifestyles consist of multiple behaviors, and adults can fail to meet more than one government recommendation. In the UK, for example, 64% of 16-year-olds do not reach the recommended level of physical activity, 35% consume more than the recommended amount of alcohol on binge days, and 20% are regular smokers19). This suggests that young people are exposed to more than one health risk, and there is evidence of multiplicative rather than additive effects of health risk20).

Smoking increases blood pressure in adults, although blood pressure among smokers varies according to age, weight, alcohol intake, and physical activity21). Sleep is also influenced by complex factors such as physical and psychological state, social environment, age, and lifestyle22, 23, 24, 25). In particular, restful sleep is important for disease prevention and recovery; however, little attention has been paid to this area26, 27, 28, 29). Both excessive and insufficient sleep has been recognized to significantly increase the incidence of cerebrovascular disease, heart disease, malignant neoplasia, and suicide30). In a study assessing the relationship between sleep duration and health status, both men and women most likely slept 7–8 h per night31). Additionally, a study investigated the relationship between health-related variables (including sleep duration) and biological indicators (chromosome aberrations and NK cells)32) has addressed, and other studies have considered sleeping more than nine hours or less than six hours as a poor health habit33, 34).

Exercise and rest have been reported to affect disease development and recovery35, 36, 37), and alcohol consumption and physical activity were the most important lifestyle predictors of sleep disorders38). Smokers are less likely to exercise, and people who like to perform exercise tend to eat more fresh food39). Moreover, sleep experts and lay literature recommend regular exercise for better sleep40, 41, 42); surveys have consistently suggested a relationship between exercise and sleep43, 44, 45), and epidemiologic studies have also indicated that exercise improves sleep26, 46, 47). In addition, positive changes in lifestyle behaviors such as losing weight and increasing exercise levels significantly reduce sleep disturbances and stress48).

Alcohol consumption and smoking are positively correlated with each other, and smoking is associated with a variety of risk factors that require intervention and evaluation. Increased mortality and morbidity that characterize smokers can potentially be reduced by improving sleep and physical activity49). Smoking is also related to binge alcohol consumption50) and decreased physical activity51, 52), and it is associated with a lack of intake of major vitamins, minerals, and fiber53, 54, 55). In particular, smokers’ poor diet and decreased physical activity increase the risk of chronic diseases56).

Meanwhile, many studies have assumed single homogeneous populations and have analyzed the correlations between related variables such as exercise, smoking, drinking, and sleep8, 9, 11, 46). Therefore, these studies have limitations because they do not reflect heterogeneous populations; they have been unable to provide comprehensive views without identifying qualitative differences within groups.

Although there is growing interest in industrial accidents, in-depth research on the subject is still lacking. In particular, little is known about the health behaviors of those who have completed industrial accident care after the accidents. Therefore, by analyzing patterns of the health behavior of those who complete industrial accident care after accidents, other researchers will be empowered to implement psychological, mental, social, and physical intervention for each type of health behavior based on the valuable data collected. Additionally, it will be possible to suggest a customized intervention method, thereby reducing national and individual medical costs and improving the quality of life of victims of industrial accidents. Thus, in the present study, exercise, smoking, alcohol consumption and sleep were used, which have been discussed in related studies, to distinguish between different health behaviors among people who have completed accident care, and it investigated their characteristics. Identifying health behaviors among a heterogeneous group of profiles rather than a single identical population may contribute to suggesting useful and appropriate practical interventions for individuals who have completed industrial accident care.

Subjects and Methods

Subjects

The following is an excerpt from the PSWCI data survey kit1), which describes the subjects and methods of the survey. The scope of the PSWCI can be broadly divided into industrial accident victims (persons who received disability grades due to industrial accidents), accident victims who have completed industrial accident care (persons who, after having completed medical care after an industrial accident, will be given a disability grade), and industrial accident workers (all workers injured while working in the industry). Separately, the rate of accidental injuries among foreign workers in Korea is increasing, although foreign workers are not suitable for the panel survey because of the difficulty in continuously investigating their accidents. Therefore, Korean industrial accident victims, victims who have completed medical care, and industrial accident workers, were considered as the subjects of the accident investigation in the PSWCI data survey.

Because the PSWCI considers industrial accident workers as a population, it is most appropriate to determine the entire worker group that is recognized to have experienced an industrial accident as the subject of the investigation1). However, from the implementation of the industrial accident insurance in 1964 to the end of June 2012, the number of disasters caused by industrial accidents reached 4,290,935, and the number of industrial accident workers has remained at around 90,000. In the South Korea KWCWS data, 35,000 injured workers have a disability rate of 38.9% among industrial accident workers according to the year and grade. In addition, 35,000 injured workers receive regular medical care, and furthermore, 56.0% of the patients treated for industrial accidents complete treatment within six months; therefore, it is considered that many of the workers who suffer an industrial accident and are treated within six months do not have any disabilities. Taking these factors into consideration, the focus of industrial accident insurance policies should be on workers who have received a work injury disability grade or who have been receiving long-term care for six months or more. However, building a panel of all workers was not appropriate for the purpose of the survey because there are comparatively few employees in these two groups. Instead, the panel consisted of employees who had completed industrial accident care.

The PSWCI was conducted as a 1:1 interview with survey panel members; the interviewers visited the panelists directly and received their responses. Because it is a panel survey, the questionnaire’s structure and content are complex, and there are so many questions that it is practically impossible for even skilled investigators to carry out investigations. In order to solve this problem, PSWCI used computer-assisted personal interviewing (CAPI), a computer-based interpersonal interview method. The interviewer was able to directly input responses into the CAPI program on a laptop rather than on paper and to immediately check the error response according to the logic of the program. This method improved the reliability, accuracy, and convenience of the survey.

The PSWCI has collected stratified systematic sampling data every year since 2013 on 2,000 employees who have completed industrial accident care (including non-disabled workers) from January 2012 to December 2012. The actual data identified a target population of 89,921, but there were 73 unknown addressees and 7,350 foreigners and Jeju Island residents in that population; after excluding these individuals, 82,498 persons comprised the survey population. Population stratification was analyzed based on gender, age (divided into five categories), region (nine areas), disability grade (six categories), and rehabilitation service availability. As noted earlier, the PSWCI provides information on economic activity, health, quality of life, household status, income, and disbursement after industrial accidents. In the PSWCI survey data, 2,000 people were selected as the sample considering the limited research cost and statistical utilization1). Finally, after excluding 197 non-respondents, the second-year (2014) PSWCI data for 1,803 of the 2,000 subjects were analyzed.

Data

Four health behavior variables were used to create the health behavior types: average weekly exercise days, average sleep duration per day, average number of cigarettes smoked per day, and number of days of alcohol consumption. The details of the questions used in the PSWCI survey1) are as follows: For average weekly exercise days, the PSWCI included exercise for rehabilitation and counted it as one day even if it occurred multiple times a day; if the response was ambiguous, the interviewer used the data for the previous week and rated it on a scale of 0–7. The survey did not include naps in the average sleep duration per day; however, if sleep was not possible at all due to nighttime pain, the interviewer included sleep during the day. Because respondents could have had insomnia, responses were based on the actual sleep duration, and not on the amount of time spent lying in bed waiting to fall asleep. When responses were ambiguous, respondents answered based on the previous week, and if they responded to the minute, their responses were rounded up; thus, the possible response range was from 0 to 24. Number of days of alcohol consumption was divided into the following five categories: less than once a month, two to three times a month, one to two times a week, three to four times a week, and almost every day.

After analyzing the health behavior types, the following variables were used to examine the characteristics of each type: gender, age, education background, marital status (none=0), employment status (unemployed=0), socioeconomic status, presence of chronic diseases (none=0), and life satisfaction. The response options for education level were no education, elementary school, middle school, high school, and more than college graduation. Employment status was based on being employed at the time of responding to the survey. Life satisfaction involved the following six questionnaire items: family income, leisure life, residential environment, family relationships, relative relationships, and social friendships. Respondents ranked each item on a 5-point Likert scale (1=very satisfied, 5=very dissatisfied), and the responses were reverse coded; i.e., the higher the score, the greater the life satisfaction based on the average values of the six sub-dimensions. The characteristic variables are presented in Table 1.

Table 1.

Characteristic analysis of the variables

Variables	Division	Persons	%	Variables	Division	Persons	%
Sex	Male	1,514	84.0	Education Level	No education	69	3.8
	Female	289	16.0		Elementary school	300	16.6
Presence of spouse	No spouse	520	28.8		Middle school	345	19.1
	Having spouse	1,283	71.2		High school	814	45.1
Employment condition	Unemployment	358	19.9		More than college graduates	275	15.3
	Employment	1,445	80.1	Socioeconomic status	Low	711	39.4
Presence of chronic disease	Disease not present	1,413	78.4		Middle	965	53.5
	Disease present	390	21.6		High	127	7.1
Age		Minimum=18, Maximum=76, M=50.54, sd=11.113
Life satisfaction		Minimum=1, Maximum=5, M=3.27, sd=0.519

Data analyses

Mplus 5.2157) and SPSS version 20.058) were used to analyze the data. Specifically, to investigate the health behavior type profiles of survey respondents who had completed industrial accident care, the four aforementioned behavior categories were inputted into the model and latent profile analysis (LPA) was conducted. Variance analysis and multinomial logistic regression analysis were also used to examine the characteristics of each group in each profile.

Potential profile analysis is excellent for estimating a unique profile of mean and variance59), and it has the advantage of optimizing the distinction between potential profiles and accurately estimating population parameters60). Compared with the traditional group classification method, there is no restriction according to the scale such as the equal variance assumption, and the statistical model for the population can be set based on the model61).

The following three criteria recommended in previous studies were used62, 63): First, the number of classes according to the potential profile was confirmed using the Bayesian information criterion (BIC) and the sample-size adjusted Bayesian information criterion (SSABIC), which is excellent in the information index. The BIC and SSABIC indicate better model fit with low indices64). Second, the bootstrap likelihood ratio test (BLRT) was used to test statistical significance between the alternative and null hypotheses; the BLRT supports the alternative hypothesis when p<0.05. Third, the minimum percentage of the classes was applied to the sample to assess the real usefulness of the latent profiles. For the minimum ratio, some researchers use at least 5% of the sample65), and other researchers use more than 1% of the sample66).

Results

Development of the latent profile model

To determine the health behavior type profiles, LPA was performed for the latent classes and characteristics of the latent class groups were identified; models with one to five classes were run (Table 2). The number of final classes was determined based on the results of the given conditions. Based on the four goodness-of-fit indices and latent class classification ratios (Table 2), neither the Lo-Mendell-Rubin (LMR) likelihood ratio test nor the BLRT supported the alternative hypothesis for the five-class model, and there was a class <1% in the latent class classification. Based on the five criteria—BIC, SSABIC, LMR, BLRT, and latent class classification ratios—four layers were the most appropriate. While selecting the final model, it is necessary to ensure that the differences between the groups are explained statistically and visually (graphically). In this study, four classes were statistically the most suitable based on the BIC and SSABIC. However, the three-class model was selected as the final model because the three classes in the graph represented the most obvious differences among the groups.

Table 2.

Applicable index and latent class rate

Model	BIC	SSABIC	LMR	BLRT	Latent class rate (%)
					1	2	3	4	5
1-class	26,358.645	26,333.230	na	na	100
2-class	24,580.051	24,538.750	0.0000	0.0000	60.5	39.5
3-class	24,362.456	24,305.271	0.0000	0.0000	13.3	47.1	39.6
4-class	23,380.760	23,307.690	0.0000	0.0000	42.5	13.0	27.6	16.9
5-class	23,709.919	23,620.964	0.9725	1.0000	0.4	14.5	45.5	32.8	6.7

n: 1,803; na: not applicable

Profile group characteristics

The quality of classifications can be determined through entropy and posterior probability. The posterior probability range of the three classes used was 0.90–1.0067), and the entropy value was 0.922, indicating high classification accuracy68). The profiles of the health behavior types based on the four indices are presented in Table 3. The group names were determined based on the characteristics of each subtype group in the health behavior type profiles; the names were simplified based on previous research to increase readability. The first group of respondents was designated as health conscious (n=240, 13.3%) because they reported the highest average number of weekly exercise days, the lowest average number of cigarettes smoked per day, and the lowest alcohol consumption frequency. Another group was designated as potential risk (n=850, 47.1%) because they had the lowest average number of weekly exercise days, low average number of cigarettes per day, and moderate alcohol consumption frequency. The remaining individuals were designated as high risk (n=713, 39.6%) because they reported the highest average number of cigarettes per day and the highest alcohol consumption frequency; they reported low average number of weekly exercise days, but they exercised more frequently than the potential-risk group. Each index had a statistically significant difference (F value) in the health behavior profiles (p<0.001; Table 3).

Table 3.

Differences in health behavior type profiles by index

Group	Health-conscioustype (n=240)	Potential-risktype (n=850)	High-risktype (n=713)	Total(n=1,803)	F
Profile index
Average weekly exercise days	5.90 (1.038)	0.88 (1.212)	1.46 (2.127)	1.77 (2.304)	921.922***
Average daily sleep duration	6.45 (1.116)	6.82 (1.162)	6.84 (1.106)	6.78 (1.141)	11.664***
Average number of cigarettes smoked per day	0.05 (0.227)	0.10 (0.297)	2.44 (0.530)	1.02 (1.22)	7,500.798***
Number of days of alcohol consumption	1.15 (1.422)	1.62 (1.502)	2.51 (1.496)	1.91 (1.574)	105.641***

Mean (SD) ***p<0.001

Prediction of the health behavior type characteristics

Multinomial logistic regression analysis was conducted using the health-conscious and potential-risk types to predict the respondents’ characteristics based on their health behaviors (Table 4). On comparison of the health conscious and potential-risk types, the potential risk was associated with younger age (0.98 times), being employed (3.42 times), lower socioeconomic status (0.62 times), and lower life satisfaction (0.66 times). On comparison of the health conscious and high risk types, the high-risk type was associated with male (51.65 times), younger age (0.97 times), being employed (3.78 times), absence of chronic disease (0.46 times), lower socioeconomic status (0.48 times), and lower life satisfaction (0.54 times). On comparison of the high-risk and potential-risk types, the high-risk type was associated with being male (65.47 times) and younger (0.99 times), not having a chronic disease (0.64 times), and having lower socioeconomic status (0.78 times).

Table 4.

Comparison of characteristics of health behavior type by group

Reference group	Predictive variable	Comparison group
		Potential-risk				High-risk
		B	Wald	odd		B	Wald	Odd
Health-conscious	Sex (male)	−0.237	1.745	0.789		3.944	55.640	51.649***
	Age	−0.021	5.238	0.979*		−0.036	13.812	0.965***
	Educational Level	−0.098	1.202	0.906		−0.155	2.640	0.857
	Presence of spouse	0.291	2.376	1.338		0.037	0.034	1.037
	Employment condition (employment)	1.229	47.936	3.416***		1.329	45.475	3.776***
	Presence of chronic disease	−0.322	3.442	0.725		−0.768	15.945	0.464***
	Socioeconomic status	−0.487	11.504	0.615**		−0.740	22.944	0.477***
	Life satisfaction	−0.421	5.887	0.656*		−0.615	11.279	0.540**
Potential- risk	Sex (male)					4.182	67.380	65.470***
	Age					−0.015	5.040	0.985*
	Educational Level					−0.056	0.725	0.945
	Presence of spouse					−0.255	3.544	0.775
	Employment condition(employment)					0.100	0.400	1.105
	Presence of chronic disease					−0.446	9.477	0.640**
	Socioeconomic status					−0.253	5.562	0.777*
	Life satisfaction					−0.194	2.543	0.824

*p<0.05; **p<0.01; ***p<0.001

Reference variables: Sex (Female=0), Presence of spouse (No=0), Employment condition (unemployment=0), Presence of chronic disease (No=0)

Discussion

In this study, the health behavior profiles of individuals who had completed industrial accident care were explored based on the KWCWS’s two-year PSWCI survey data. The characteristics of the groups in each profile were examined, using average weekly exercise days, average sleep duration per day, average number of cigarettes per day, and alcohol consumption frequency as the profile indices. The resulting health behavior profile subtypes were health conscious (13%), potential risk (47%), and high risk (40%), and the indices differed significantly among the three groups. This means that the health behaviors of those who have completed industrial accident care differ, and it is necessary to apply different practical interventions for each type.

Multinomial logistic analysis used the health-conscious and potential-risk types as the reference groups to predict the characteristics of the groups in different health behavior profiles. Compared with the health-conscious type, the potential-risk type was likely to be younger and employed and to have lower socioeconomic status and less life satisfaction. The 2003 Health Survey for England (n=11,492) reported similar results in its clustering; risk factors such as smoking, heavy drinking, and lack of physical activity were more prevalent among younger respondents, men, lower social class households, and respondents who were economically inactive69).

In the present study, younger people were more likely to engage in risky health behaviors, and in a clustering study of 16,789 men and women aged 20–59 yr, alcohol and tobacco consumption were also higher among younger people70). In another study, “Knowledge, Attitude, and Practice of Health Promotion,” the protective behavior pattern was dominant than the risky behavior pattern among those aged 55 and over whereas the risky behavior pattern was dominant among those under 5571). In contrast, a cross-sectional study of 4,812 community-dwelling residents older than age 60 found a higher prevalence of multiple risk factors (cigarette smoking, heavy alcohol drinking, physical inactivity) among the older age groups and a lower prevalence among those who were employed72). Interestingly, in the 2003 Scottish Health Survey (n=6,574), the prevalence of risky health behaviors had a strong relationship with poorer socioeconomic circumstances73), and the 2002 wave of the English Longitudinal Study of Ageing (n=11,214) also found that risky health behaviors were rare among people with high socioeconomic status74). These results for large populations, including our study, suggest that health behavior policies need to address low socioeconomic groups.

In the present study, there were no significant differences in education level between the potential-risk and the health-conscious behavior types, and another researcher found that health-risk behaviors are more related to social status than to education74). There is still disagreement regarding whether education level is associated with less risky health behaviors among adults with higher education levels75) or with more risky behaviors among adults with lower education levels75, 76, 77).

In terms of life satisfaction and health behavior, there is one report which reveals that the two factors are independent of each other78). However, the 2005 Behavioral Risk Factor Surveillance System study represented that the lower the life satisfaction, the more risky the health behaviors such as alcohol consumption, smoking, and lack of exercise79). Similarly, a study of 102 teachers suggested that the greater the life satisfaction, the more the teachers engage in health-promoting behaviors80).

There are also researches on health behaviors in relation to employment status. Some findings reveal that employed people report a higher frequency of smoking and alcohol consumption81, 82) and less physical activity83), but other studies found that the prevalence of multiple health risk factors was lower among the employed72). Another study found that unemployed participants had higher risky drinking and physical inactivity, but employed participants reported higher tobacco consumption84).

The results of the comparison among the health behavior types are consistent with prior reports which suggest that low-income earners are more likely to engage in risky health behaviors69, 73, 85) and that men are more likely to drink alcohol than women86). In the National Longitudinal Study of Adolescent Health (n=18,911), adolescents and young adult males were more susceptible to risky health behaviors, and these behaviors increased in frequency over time87). In addition, men represent a higher prevalence of engaging in multiple risky health behaviors72), although cigarette consumption is steadily increasing in both men and women88).

In our analyses, people with chronic illnesses were more likely to engage in healthy behaviors; the following associations between chronic diseases and smoking and alcohol consumption are sufficient reasons for those with chronic diseases to avoid risky health behaviors such as smoking and consuming alcohol. Smoking is a major factor for developing or worsening cardiovascular disease89), coronary artery disease90), acute coronary thrombosis and stroke91, 92), diabetes93), and major vascular complications94, 95). Alcohol consumption is also associated with epilepsy96), hypertensive heart disease97), ischemic heart disease98), ischemic stroke99), colorectal cancer, rectal cancer, and liver cancer100).

Based on these results, the following action plan is suggested for managing industrial workers’ health and for preventing industrial accidents. First, after these accidents, customized counseling and guidance are required for each health behavior type. This means that health care and prevention education programs should vary according to workers’ health behavior types; in particular, to maintain workers’ health, these programs should consider their gender, age, employment status, chronic illness status, socioeconomic status, and life satisfaction, which can be improved with treatment and leisure activity programs. Second, it is specifically necessary to secure the manpower required to conduct these specialized programs; leisure activity and health programs for industrial workers who have experienced accidents at work are insufficient, partly because of the lack of an available workforce. Third, the probability of belonging to the potential-risk and high-risk behavior types increases for those of lower socioeconomic status, younger people, males, and people who do not have a chronic disease. These results suggest that young male workers are more exposed to risk factors that encourage exercise avoidance, smoking, and excessive alcohol consumption, and therefore, national institutional support is necessary for promoting more healthy behaviors among young male workers. In addition, public awareness campaigns are needed to decrease, and ultimately prevent, risky health behaviors.

The limitations of this study and suggestions for subsequent research are as follows: First, this study was based only on the second-year (2014) PSWCI data; therefore, follow-up studies are needed to analyze the data related to the health behavior types presented in this previous study. Second, in this study, the health behavior profiles of workers who had completed industrial accident care were analyzed based on cross-sectional data. In follow-up studies, it will be necessary to examine causality among the relationships between variables based on changes reflected in the longitudinal data. Third, it is necessary to re-investigate this topic because only self-reported data were used in this study. Fourth, all physical activities were counted in the exercise category in this study, and fifth, the PSWCI does not enquire about the severity of injuries among the patients on the panel, and thus, severity could not be discussed.

Despite these limitations, in this study, significant health behavior and lifestyle patterns were observed among people who had completed industrial accident care. These patterns, which connect factors such as physical activity, sleep, smoking, and alcohol consumption, enable tailored health promotion strategies for industrial accident victims who have completed industrial accident care and provide important insights into their health-related needs.