Change in Lifestyle through Health Promotion Program without Face-to-Face Intervention in a Large-scale Japanese Enterprise

Abstract

Objective: The aim of this study was to assess the effect of a health promotion program with 2–3-month Internet or printed material support without a face-to-face intervention at a large-scale worksite. Methods: An uncontrolled intervention was conducted. In 2005, 22,429 employees underwent a health check-up and a lifestyle assessment. In 2006, 2,096 employees were recruited to participate in an intervention group (IntG), and the remaining 20,228 employees were allocated to a control group (ConG). After one-year follow-up, 1,888 employees in IntG and 15,474 employees in ConG were analyzed by Breslow's lifestyle index, overall prevalence of cardiovascular risk (OPCR) and 10% 10-year risk based on cardiovascular risk factors in 2007. Results: Breslow's lifestyle index (over 5 points) in IntG showed a higher OR (1.13 with a 95% CI of 1.01–1.26) than that in ConG. Compared with those with a poor lifestyle, the subjects who had maintained or improved their lifestyle showed a lower OPCR (OR 0.88, 95% CI 0.79–0.97; OR 0.84, 95% CI 0.74–0.95 [p for trend=0.003]) and 10% 10-year risk trend (OR 0.38, 95% CI 0.31–0.47; OR: 0.29, 0.22–0.37 [p for trend <0.001]). Conclusion: The present study showed that there was a modest lifestyle behavior change using a program without face-to-face contact. The employees who maintained and improved their lifestyle had a lower OPCR and 10% 10-year risk. Future study should increase the number of employees actively attempting to improve their lifestyles at a worksite.

(J Occup Health 2013; 55: 74–83)

Introduction

Cardiovascular diseases are now the leading public health problem in industrialized countries^{1, 2}). In Japan, cardiovascular diseases account for approximately 15% of all deaths, and they remain the second highest cause of death, although the level has been declining slowly for years²). It is well known that cardiovascular diseases develop from biological risk factors such as being overweight, hypertension, dyslipidemia and diabetes^3–7). These biological risk factors are mainly caused by unhealthy lifestyle-related factors, such as limited physical exercise, poor dietary habits, alcohol intake, and smoking. Therefore, lifestyle behavior changes are considered to be important to improve these biological risk factors and to prevent cardiovascular diseases^8–11).

In worksites, many intervention trials have shown that workplace health promotion programs (WHPPs) effectively improved cardiovascular risk factors for employees^12–15). These programs mostly included individual consultations with one-on-one and group education and exercise with face-to-face contact for subjects, according to the systematic review of Groeneveld et al.¹⁶). However, it is difficult to deliver these high-intensity programs with one-on-one or face-to-face contact to each one of several thousand employees during an intervention period in large-scale enterprises under real-life conditions. Therefore, not only low-intensity intervention but also high benefits for reductions in cardiovascular risk factors need to be verified in WHPPs. Against this background, some studies that only used Internet-based^{17, 18}) and printed material-based^{19, 20}) programs in worksites have been reported. These studies showed that programs without face-to-face contact were simple and effective, but no strong evidence was obtained owing to small sample sizes and lack of compliance in participant^{19, 21}). Hence, the effectiveness of a WHPP without face-to-face contact in a large-scale company in a real-life setting is unclear.

The purpose of the present study was to assess the effect of a health promotion program with 2–3-month Internet or printed material support without face-to-face contact at a large-scale worksite.

Methods

Study design and subjects

An uncontrolled intervention was conducted at large-scale group companies including travel business companies, commercial business companies, publishing and advertising companies, etc. with approximately 30,000 employees in Japan. As shown in Fig. 1, in 2005, at baseline, 22,429 employees from 20 to 69 years old underwent annual health check-ups and a lifestyle assessment. In 2006, 105 employees taking medication for cardio-cerebral-vascular diseases were excluded, and 22,324 were recruited for the WHPP, which was named “Good Shape”. The staff of the health center in this company designed a leaflet that included an introduction of the contents of the “Good Shape” WHPP and an application form. Then the leaflets were distributed to every employee. Of these 22,324 employees, 2,096 employees (9.4%) were recruited to participate in the WHPP as an intervention group (IntG), and the remaining 20,228 (90.6%) employees were allocated to a control group (ConG). In 2006, the 2- or 3-month intervention program without face-to-face contact and a lifestyle assessment were performed. After 1-year follow-up, excluding dropouts, those who left or retired and those with missing data on health check-up records in IntG (n=208, 9.9%) and ConG (n=4,754, 23.5%), changes in lifestyle and outcomes of health check-ups in IntG (n=1,888) and ConG (n=15,474) were analyzed in 2007.

Fig. 1.

Flow chart of the program.

All employees were informed by a leaflet about the WHPP, and written informed consent to participate in the WHPP was obtained. The identities of participants remained anonymous, in compliance with the Ethical Guidelines for Epidemiological Research²²), and the study protocol was approved by the Dokkyo University School of Medicine Ethics Committee.

Intervention programs

The WHPP consisted of four courses, named “Health-up Navi”, “Health Challenge”, “No Smoking Marathon” and “Walking 0.8 & 1 Million Steps.” All courses created behavior change without face-to-face interventions. In IntG, 2,096 participants could freely choose one or more of the 4 courses in 2006. Among the 2,096 participants in IntG, 240 (11.5%) persons chose “Health-up Navi,” 293 (14.0%) persons chose “Health Challenge”, 45 (2.2%) persons chose “No Smoking Marathon”, 1,259 (60.2%) persons chose “Walking 0.8 & 1 Million Steps” and 255 (12.2%) persons chose two of more of the 4 courses.

“Health-up Navi” was supported by a computer-tailored comprehensive lifestyle modification support tool for three months. This course included the following basic contents: first, the participants answered a questionnaire, and received the results of lifestyle assessment and then set some goals and planned how to improve their lifestyle themselves; the goals included exercise, nutrition and food habits. Finally, they put their planned changes in behavior into action and recorded their progress and results on the web. Sometimes, they received advice concerning knowledge about health and encouragement for behavior changes via the web.

“Health Challenge” was also a comprehensive lifestyle modification course, but it was supported by the printed educational materials for two months. In this course, the participants read some health-related messages and answered a questionnaire about lifestyle. According to the results of their lifestyle assessment, the participants could identify their own poor lifestyle habits and receive feedback concerning how to improve them. Then they set some goals; finally, they put their planned changes in behavior into action and recorded their progress and results in the printed educational materials.

“No Smoking Marathon” was supported by a computer-tailored quitting smoking tool for three months. “Walking 0.8 & 1 Million Steps” was supported by printed educational materials for three months. After reading some health-related messages on smoking or exercise and receiving advice concerning how to quit smoking and improve their level of exercise, the participants attempted to quit smoking and started walking for exercise and recorded their result in the printed educational materials and on the web, respectively.

The control group was not provided with these activities or resources.

Measurements

1) Cardiovascular risk: The medical records in the 2005 and 2007 annual health check-ups were used. Variables and criteria of abnormality were as follows: age, sex, height, weight, systolic/diastolic blood pressure (SBP/DBP, ≥130/85 mmHg), triglycerides (TG, ≥150 mg/dl), high-density lipoprotein cholesterol (HDL, <40 mg/dl), low-density lipoprotein cholesterol (LDL, ≥120 mg/dl), total cholesterol (TC, ≥200 mg/dl, female over 50 years, ≥220 mg/dl), fasting blood glucose (FBG, ≥110 mg/dl), HbA1c (≥5.6%, JDS) and uric acid (UA, ≥7 mg/dl). Body mass index (BMI, ≥25 kg/m²) was calculated as body weight (kg) divided by the square of height (m²). The criteria of the abnormality were judged using the Guidelines for Diabetes and Cardiovascular Diseases in Health Care of the Aged²³). Net changes in the overall prevalence of cardiovascular risk (OPCR) were determined based on more than one abnormality in BMI, SBP/DBP, TG, HDL, LDL, TC, FBG, HbA1c and UA. According to Appendix 1, the Framingham Point Score (FPS) was calculated with sex, age, smoking status, TC, HDL, SBP and hypertension treatment using the Adult Treatment Panel III criteria of the National Cholesterol Education Program^{24, 25}) (NCEP-ATP III, 2002). The 10% 10-year risk was defined as an FPS of more than 10% in males (at 12 points of total FPS) and 11% in females (at 20 points of total FPS).

2) Lifestyle assessment: All subjects were asked to answer a questionnaire including medication conditions and Breslow's lifestyle index^{26, 27}) at the annual health check-ups in 2005, 2006 and 2007. Since this was a questionnaire for assessing employee's lifestyle at health check-ups, the questions and possible answer categories included the following: 1) how long do you sleep everyday? (less than 4 hr/4–6 hr/6–8 hr/more than 8 hours); 2) are you a smoker now? (smoker/nonsmoker included ex-smoker); 3) do you have breakfast everyday? (yes/no); 4) do you have a snack between meals everyday? (yes/no); 5) do you often get physical exercise? (no/1–2 times per month/1–2 times per week/3–5 times per week/everyday); 6) do you often drink alcohol? (no/1–2 days per month/1–2 days per week/3–5 days per week/everyday), and 7) How much alcohol do you drink in one? (less than 1 go/1 go/2 go/3 go/more than 4 go) (1 go is about 20 g ethanol). Maintenance of proper weight was determined by the proper weight ±10%. The proper weight (kg) was calculated as 22 (kg/m²) times height squared (m²). With regard to lifestyle assessment, Breslow's lifestyle index was determined through 7 items, sleeping patterns (sleeping 6–8 hours per day=1, other=0), smoking status (nonsmoker=1, smoker=0), eating breakfast (yes=1, no=0), alcohol consumption (less than 5 days per week and less than 20 g per day=1, other=0), physical exercise (3 times per week or more=1, other=0), having a snack between meals (yes=0, no=1) and maintaining a proper weight (yes=1, no=0). In the present study, the high- and low-score lifestyle groups were defined with the threshold (medium) of a total of 5 points. According to the high- or low-score lifestyle groups, improved lifestyle was determined to be moving from low-score group at 2005 to the high-score group in 2006 and 2007, and maintained lifestyle was determined to be remaining in the high-score group in 2005, 2006 and 2007, other cases were considered to be poor lifestyle.

Statistical analysis

All variables were classified into binomial categories. The chi-square test or McNemar's test was used for comparison between intervention and control groups and before (pre) and after (post) the program. The odds ratio (OR) with 95% confidence interval (95%CI) of each cardiovascular risk and lifestyle factor between IntG and ConG was analyzed after adjusting for sex, age and baseline data of a specific dependent variable. The association of OPCR and 10% 10-year risk with change of lifestyle and presence of intervention was estimated using a multiple logistic regression model.

Probability values <0.05 were considered statistically significant. IBM SPSS Statistics 19 for Windows was used for all statistical analyses.

Results

Baseline characteristics, cardiovascular risk and lifestyle

As shown in Table 1, there were significant differences in gender and age group between IntG and ConG at baseline. The abnormality of cardiovascular risk and 10% 10-year risk together or individually in IntG was greater than in ConG. As for lifestyle, the proportions of participants who exercised more than 3 times per week and did not have a snack between meals in IntG were higher than those in ConG, and there were no differences in other items and mean high score (over 5 points) of Breslow's lifestyle index between ConG and IntG at baseline. Compared with males, females had lower proportions of cardiovascular risk and poor lifestyle at baseline (date not shown)

Table 1. Comparison of the characteristics between the intervention and control groups at baseline

	Intervention group		Control group		p-valuea
	n=2,096		n=20,228
	n	%	n	%
Gender
Female	1,069	51.0	12,036	59.5	<0.001
Male	1,027	49.0	8,192	40.5
Age groups
20–29 yr	634	30.2	7,927	39.2
30–39 yr	668	31.9	6,561	32.4
40–49 yr	388	18.5	2,920	14.4	<0.001
50–59 yr	393	18.8	2,575	12.7
60–69 yr	13	0.6	245	1.2
Cardiovascular risk
BMI≥25 kg/m2	439	20.8	2,940	14.5	<0.001
SBP≥130 mmHg	532	25.4	4,310	21.3	<0.001
DBP≥85 mmHg	331	15.8	2,296	11.4	<0.001
TG≥150 mg/dl	339	16.2	2,401	11.9	<0.001
HDL < 40 mg/dl	69	3.3	480	2.4	0.01
TC≥200 mg/dl or≥220 mg/dlb	475	22.7	3,823	18.9	<0.001
LDL≥120 mg/dl	716	34.2	6,194	30.6	0.001
FBG≥110 mg/dl	146	7.1	1,082	5.3	0.002
HbA1c≥5.6%	120	5.7	945	4.7	0.031
UA≥7 mg/dl	256	12.2	1,901	9.4	<0.001
OPCRc≥one item	1,227	58.5	10,532	52.1	<0.001
10-Year riskd≥10%	221	10.8	1,830	9.0	0.024
Breslow's lifestyle index, each item
Smoking status, nonsmoker	1,464	69.8	13,729	67.9	0.065
Physical exercise,≥3 times per week	140	6.7	985	4.9	<0.001
Sleeping patterns, 6–8 hours per day	893	42.6	8,693	43.0	0.745
Alcohol consumption, <20 g per day and <5 days per week	1,062	50.7	10,559	52.2	0.181
Eating breakfast every day, yes	1,688	80.5	15,945	78.8	0.068
Not having a snack between meals, yes	1,895	90.5	17,996	89.0	0.043
Maintaining proper weight, yes	1,549	73.9	15,313	75.7	0.068
Breslow's lifestyle index score, ≥5 point	842	40.2	7,976	39.4	0.509

^a  p-values were calculated using the chi-square test.

^b  Females over 50 yr.

^c  OPCR means an overall prevalence of cardiovascular risk including more than one abnormality in BMI, SBP/DBP, TG, HDL, LDL, TC, FBG, HbA1c and UA.

^d  10-Year risk means a Framingham point score of more than 10%.

Lifestyle change

Table 2 shows that, for lifestyle, compared with the levels at baseline, the proportions of participants who were nonsmokers, slept 6–8 hours per day and had a high score (over 5 points) on Breslow's lifestyle index significantly increased, and the proportion of those with alcohol consumption less than <20 g per time and <5 days per week significantly decreased in the ConG and IntG after two years, respectively. In the ConG, the proportion of participants getting physical exercise ≥3 times per week significantly increased, and those not having a snack between meals and maintaining proper weight significantly decreased. After adjusting for age, sex and baseline, the proportions of participants who were nonsmokers, got physical exercise ≥3 times per week, ate breakfast every day and had a high score (over 5 points) on Breslow's lifestyle index in IntG were higher than those in ConG (OR 1.27, 95% CI 1.08–1.50; OR 1.26 95% CI 1.03–1.53; OR 1.17, 95% CI 1.01–1.36; and OR 1.13 95% CI 1.01–1.26, respectively [Table 2]).

Table 2. Comparison of the lifestyle between the intervention and control groups

		Pre		Post		Post minus pre			IntG vs. ConG
	Groupa	n (1)	% (1)	n (2)	% (2)	n (2)—n (1)	% (2)—% (1)	p-valueb	ORc	95% CI
Breslow's lifestyle index, each item
Smoking status,	ConG	10,359	66.9	11,108	71.8	749	4.9	<0.001	1.00
nonsmoker	IntG	1,317	69.8	1,425	75.5	108	5.7	<0.001	1.27	1.08–1.50
Physical exercise,	ConG	744	4.8	826	5.3	82	0.5	0.011	1.00
≥3 times per week	IntG	128	6.8	146	7.7	18	0.9	0.173	1.26	1.03–1.53
Sleeping patterns,	ConG	6,572	42.5	7,077	45.7	505	3.2	<0.001	1.00
6–8 hours per day	IntG	812	43.0	863	45.7	51	2.7	0.033	0.97	0.87–1.08
Alcohol consumption,	ConG	7,966	51.5	7,724	49.9	−242	−1.6	<0.001	1.00
<20 g per day and <5 days per week	IntG	957	50.7	887	47.9	−70	−2.8	0.002	0.94	0.84–1.05
Eating breakfast every day,	ConG	12,234	79.1	12,146	78.5	−88	−0.6	0.064	1.00
yes	IntG	1,527	80.9	1,541	81.6	14	0.7	0.405	1.17	1.01–1.36
Not having a snack between meals,	ConG	13,884	89.7	13,741	88.8	–143	–0.9	<0.001	1.00
yes	IntG	1,710	90.6	1,700	90.0	−10	−0.6	0.516	1.06	0.88–0.26
Maintaining proper weight,	ConG	11,733	75.8	11,637	75.2	−96	−0.6	0.017	1.00
yes	IntG	1,394	73.8	1,389	73.6	−5	−0.2	0.780	1.01	0.86–1.18
Breslow's lifestyle index, score ≥5 point	ConG	6,034	39.0	6,297	40.7	263	1.7	<0.001	1.00
	IntG	766	40.6	823	43.6	57	3.0	0.016	1.13	1.01–1.26

^a  IntG, intervention group; ConG, control group.

^b  p-values were calculated using McNemar's test.

^c  Adjusted for age, sex, baseline and course.

Cardiovascular risk change

As shown in Table 3, the abnormality rates significantly increased for DBP, TC, OPCR and 10% 10-year risk in both groups and for BMI, TG and UA in ConG after 1-year follow-up. IntG showed a tendency for a decrease in TC, HDL, LDL and HbA1c. ConG showed a significant improvement in SBP. Compared with ConG, IntC showed a higher odds ratio (OR 1.25, 95% CI 1.10–1.42) in SBP and a lower odds ratio (OR 0.84, 95% CI: 0.74–0.95) in LDL.

Relationship between cardiovascular risk and changed lifestyle

As shown in Table 4, compared with poor lifestyle, the subjects who had maintained or improved their lifestyle showed a lower OPCR (OR 0.88, 95% CI 0.79–0.97; OR 0.84, 95% CI 0.74–0.95 [p for trend=0.003]) and 10% 10-year risk trend (OR 0.38, 95% CI 0.31–0.47; OR: 0.29, 0.22–0.37 [p for trend<0.001]).

Discussion

The present study showed that a WHPP without face-to-face contact could maintain or improve some lifestyle factors, such as smoking, exercise and having breakfast, and the proportion of participants who had high scores on Breslow's lifestyle index in IntG. To the best of our knowledge, this is the first study to suggest that a program provided via the Internet or printed materials and without face-to-face contact is effective for changing the lifestyles of workers at a large-scale worksite.

Our results are consistent with previous studies showing that smoking cessation was effectively achieved using Internet-based programs^{28, 29}) or printed material³⁰), but the subjects of these previous studies were employees in small- and medium-sized companies. In some prior studies, programs without face-to-face contact were reported to increase body weight loss, exercise and improve dietary habits^{17, 21, 31–33}). In the present study, these items could not be shown to improve, but a relative improvement or maintenance was achieved in IntG. As a result, a composite lifestyle change score, Breslow's lifestyle index (over 5 points), in IntG showed significant improvement, consistent with the findings of Jacobs et al.³⁴). Compared with some high-intensity interventions such as individual consultation and group education with face-to-face contact with an expert¹⁶), the WHPP in this study used only a 2–3-month intervention, and the lack of face-to-face contact could be considered as reflecting a minimal-intensity intervention. Therefore, the present findings suggest that modest lifestyle behavior change using a low-intensity intervention is possible.

Table 3. Comparison of the cardiovascular risk between intervention and control groups

	Pre			Post		Post minus pre			IntG vs. ConG
	Groupa	n (1)	% (1)	n (2)	% (2)	n (2)—n (1)	% (2)—% (1)	p-valueb	ORc	95% CI
Cardiovascular risk
BMI≥25 kg/m2	ConG	2,434	15.7	2,699	17.4	265	1.7	<0.001	1.00
	IntG	408	21.6	428	22.7	20	1.1	0.085	1.03	0.84–1.25
SBP≥130 mmHg	ConG	3,468	22.4	3,318	21.4	−150	−1.0	0.005	1.00
	IntG	487	25.8	522	27.6	35	1.8	0.083	1.25	1.10–1.42
DBP≥85 mmHg	ConG	1,863	12.0	2,337	15.1	474	3.1	<0.001	1.00
	IntG	306	16.2	337	20.0	31	3.8	<0.001	1.14	0.99–1.32
TGɥ150 mg/dl	ConG	2,013	13.0	2,118	13.7	105	0.7	0.01	1.00
	IntG	314	16.6	304	16.1	−10	−0.5	0.556	0.95	0.81–1.12
HDL<40 mg/dl	ConG	406	2.6	400	2.6	−6	0.0	0.803	1.00
	IntG	65	3.4	58	3.1	−7	−0.3	0.443	0.97	0.70–1.34
TC≥200 mg/dl or≥220 mg/dld	ConG	3,084	19.9	3,539	22.9	455	3.0	<0.001	1.00
	IntG	428	22.7	473	25.1	45	2.4	0.01	0.96	0.84–1.10
LDL≥120 mg/dl	ConG	4,964	32.1	5,207	33.6	243	1.5	<0.001	1.00
	IntG	654	34.6	636	33.7	−18	−0.9	0.399	0.84	0.74–0.95
FBG ≥ mg/dl	ConG	869	5.6	962	6.2	93	0.6	<0.001	1.00
	IntG	133	7.0	138	7.3	5	0.3	0.657	0.91	0.71–1.17
HbA1c≥5.6%	ConG	756	4.9	785	5.1	29	0.2	0.167	1.00
	IntG	105	5.6	102	5.4	−3	−0.2	0.801	0.89	0.66–1.21
UA≥7 mg/dl	ConG	1,615	10.4	1,755	11.3	140	0.9	<0.001	1.00
	IntG	240	12.7	246	13.0	6	0.3	0.693	0.97	0.81–1.16
OPCRe	ConG	8,457	54.7	8,707	56.3	250	1.6	<0.001	1.00
	IntG	1,123	59.5	1,173	62.1	50	2.6	0.008	1.10	0.96–1.26
10% 10-Year riskf	ConG	1,466	9.5	1,567	10.1	101	0.6	0.001	1.00
	IntG	202	10.7	228	12.1	26	1.4	0.03	0.99	0.80–1.22

^a  IntG, intervention group; ConG, control group.

^b  p-values were calculated using McNemar's test.

^c  Adjusted for age, sex, baseline and course.

^d  Females over 50 years.

^e  OPCR means an overall prevalence of cardiovascular risk including more than one abnormality in BMI, SBP/DBP, TG, HDL, LDL, TC, FBG, HbA1c and UA.

^f  10-Year risk means a Framingham point score of more than 10%.

In the present study, we did not find any difference in lifestyle improvement between sex and among courses (data not shown). Considering the potential biases that there were differences at baseline, we used a multiple logistic regression model adjusted for sex and courses to limit the bias from them, but it could not be completely avoided. Regarding the signification decrease in prevalence of SBP≥130 mmHg in ConG, we considered that this occurred because the people in ConG and IntG were same from the company, and so those in the control group could get the same information about the contents of the intervention program as those in IntG. Therefore, we cannot deny the possibility that some people in ConG improved their lifestyle.

With regard to deterioration of cardiovascular risk factors, Fukuda et al.³⁵) reported that the main cause was aging in a follow-up lasting 5 years at a worksite, and other previous studies also reported that blood pressure, serum lipids, and BMI in the general Japanese population were related to aging³⁶–³⁹). Therefore, the progress of the OPCR and 10% 10-year risk in this study could not be obstructed because of aging. Another important reason could be that the modest lifestyle behavior change in this study could not impact the total change in biological risk factors. However, when the change in lifestyle was assessed using three points, one each in 2005, 2006 and 2007, for a two-year period among all employees, we found that the employees who maintained or improved their lifestyle had a lower OPCR and 10% 10-year risk, although these cardiovascular risk indicators deteriorated during the two-year study period in both IntG and ConG. At the same time, we found a trend relationship among poor, maintained and improved lifestyle behavior, so it is an important indication to produce remarkable effect of the intervention for cardiovascular risk, if the number of maintained and improved lifestyle could be raised in the WHPPs. Therefore, future study should increase the number of employees actively attempting to improve their lifestyle at a worksite.

Table 4. The cardiovascular risk related to some factors after two-year follow-up

		OPCRa
	n	No	%	ORc	%95 CI	p for trend
Presence of intervention
Control group	15,474	8,707	56.3	1.00
Intervention group	1,888	1,173	62.1	1.10	0.97–1.26
Change of lifestyle
Poor lifestyle	11,084	6,663	60.1	1.00
Maintained lifestyle	3,895	1,955	50.2	0.88	0.79–0.97	0.003
Improved lifestyle	2,383	1,262	53.0	0.84	0.74–0.95
		10-Year Riskb
	n	No	%	ORc	%95 CI	p for trend
Presence of intervention
Control group	15,474	1,567	10.1	1.00
Intervention group	1,888	228	12.1	1.01	0.82–1.26
Change of lifestyle
Poor lifestyle	11,084	1,410	12.7	1.00
Maintained lifestyle	3,895	214	5.5	0.38	0.31–0.47	<0.001
Improved lifestyle	2,383	171	7.2	0.29	0.22–0.37

^a  OPCR means an overall prevalence of cardiovascular risk including more than one abnormality in BMI, SBP/DBP, TG, HDL, LDL, TC, FBG, HbA1c and UA.

^b  10-Year risk means a Framingham point score of more than 10%.

^c  Adjusted for sex, age, baseline and course

The main strength of this study was its large sample size of employees in one workplace setting. In addition, the two-year study period included one-year follow-up after intervention. Finally, a general lifestyle index, Breslow's 7 items, and the change in lifestyle based on three time points were assessed.

The limitations of this study included that its was not a randomized controlled trial. Due to this design, the characteristics of the participants in ConG were different from those in IntG, who had higher cardiovascular risks, higher levels of some lifestyle and higher motivation for healthy behavior than those in ConG. Although there was no difference in Breslow's lifestyle index (over 5 points) at baseline between ConG and IntG, selection bias might not have been eliminated. Another potential limitation was that Breslow's 7-item lifestyle index does not represent a person's whole lifestyle. In addition, 4,962 (22.2%) employees in the control group had missing health check-up data because of leaving their job, retiring or not undergoing a health check-up after two years. Finally, only one company was used, so we should focus on other worksites to show that the findings are generalizable.

Conclusion

The present study showed that a WHPP without face-to-face contact can improve some lifestyle factors and the proportion of participants with high scores of Breslow's lifestyle index. The employees who maintained and improved their lifestyle had a lower OPCR and 10% 10-year risk. Future study should increase the number of employees actively attempting to improve their lifestyle at a worksite.

Acknowledgment

The authors wish to thank all the volunteers who participated and staff who worked on or cooperated with the present project. This study was supported by grants from the Secom Science and Technology Foundation, Japan.

References

• 1)  WHO, World health statistics 2009. 2009, World Health Organization: Geneva, Switzerland.
• 2)  Health, Labour and Welfare Statistics Association, Journal of health and welfare statistics. 2011/2012, Tokyo, Japan: Kousaidou Inc. (in Japanese).
• 3)   Castelli  WP,  Garrison  RJ,  Wilson  PW,  Abbott  RD,  Kalousdian  S,  Kannel  WB. Incidence of coronary heart disease and lipoprotein cholesterol levels. The framingham study. JAMA 1986; 256: 2835-8.
• 4)   Pechère-Bertschi  A,  Greminger  P,  Hess  L,  Philippe  J,  Ferrari  P. Swiss Hypertension and Risk Factor Program (SHARP): cardiovascular risk factors management in patients with type 2 diabetes in Switzerland. Blood Press 2005; 14: 337-44.
• 5)   Daousi  C,  Casson  IF,  Gill  GV,  MacFarlane  IA,  Wilding  JP,  Pinkney  JH. Prevalence of obesity in type 2 diabetes in secondary care: association with cardiovascular risk factors. Postgrad Med J 2006; 82: 280-4.
• 6)   Ridderstråle  M,  Gudbjörnsdottir  S,  Eliasson  B,  Nilsson  PM,  Cederholm  J. Steering Committee of the Swedish National Diabetes Register (NDR). Obesity and cardiovascular risk factors in type 2 diabetes: results from the Swedish National Diabetes Register. J Intern Med 2006; 259: 314-22.
• 7)   Wang  W,  Lee  ET,  Fabsitz  RR, et al. A longitudinal study of hypertension risk factors and their relation to cardiovascular disease: the Strong Heart Study. Hypertension 2006; 47: 403-9.
• 8)   Bautista-Castano  I,  Molina-Cabrillana  J,  Montoya-Alonso  JA,  Serra-Majem  L. Variables predictive of adherence to diet and physical activity recommendations in the treatment of obesity and overweight, in a group of Spanish subjects. Int J Obes Relat Metab Disord 2004; 28: 697-705.
• 9)   Hu  G,  Sarti  C,  Jousilahti  P,  Silventoinen  K,  Barengo  NC,  Tuomilehto  J. Leisure time, occupational, and commuting physical activity and the risk of stroke. Stroke 2005; 36: 1994-9.
• 10)   Arai  H,  Yamamoto  A,  Matsuzawa  Y, et al. Prevalence of metabolic syndrome in the general Japanese population in 2000. J Atheroscler Thromb 2006; 13: 202-8.
• 11)   Pitsavos  C,  Panagiotakos  DB,  Tzima  N, et al. Diet, exercise, and C-reactive protein levels in people with abdominal obesity: the ATTICA epidemiological study. Angiology 2007; 58: 225-33.
• 12)   Aldana  SG,  Greenlaw  RL,  Diehl  HA,  Salberg  A,  Merrill  RM,  Ohmine  S. The effects of a worksite chronic disease prevention program. J Occup Environ Med 2005; 47: 558-64.
• 13)   Muto  T,  Yamauchi  K. Evaluation of a multicomponent workplace health promotion program conducted in Japan for improving employees' cardiovascular disease risk factors. Prev Med 2001; 33: 571-7.
• 14)   Proper  KI,  Hildebrandt  VH,  Van der Beek  AJ,  Twisk  JW,  Van Mechelen  W. Effect of individual counseling on physical activity fitness and health: a randomized controlled trial in a workplace setting. Am J Prev Med 2003; 24: 218-26.
• 15)   Méndez-Hernández  P,  Flores  Y,  Siani  C, et al. Evaluation of a workplace health program to reduce cornary risk factors. Clin Govenace Int J 2008; 13: 95-105.
• 16)   Groeneveld  IF,  Proper  KI,  van der Beek  AJ,  Hildebrandt  VH,  van Mechelen  W. Lifestyle-focused interventions at the workplace to reduce the risk of cardiovascular disease—a systematic review. Scand J Work Environ Health 2010; 36: 202-15.
• 17)   Pressler  A,  Knebel  U,  Esch  S, et al. An internet-delivered exercise intervention for workplace health promotion in overweight sedentary employees: a randomized trial. Prev Med 2010; 51: 234-9.
• 18)   Burling  TA,  Seidner  AL,  Gaither  DE. A computer-directed program for smoking cessation treatment. J Subst Abuse 1994; 6: 427-31.
• 19)   Kishchuk  N,  Anbar  F,  O'Loughlin  J,  Masson  P,  Sacks-Silver  G. Efficiency of printed materials in worksite health promotion. Am J Health Promot 1991; 5: 355-9.
• 20)   Fouad  MN,  Kiefe  CI,  Bartolucci  AA,  Burst  NM,  Ulene  V,  Harvey  MR. A hypertension control program tailored to unskilled and minority workers. Ethn Dis 1997; 7: 191-9.
• 21)   Faghri  PD,  Omokaro  C,  Parker  C,  Nichols  E,  Gustavesen  S,  Blozie  E. E-technology and pedometer walking program to increase physical activity at work. J Prim Prev 2008; 29: 73-91.
• 22)  Ministry of Education, Science and Technology, Ministry of Health, Labour and Welfare. Ethical Guidelines for Epidemiological Research. 2008.
• 23)   Tatara  K. Guideline for diabetes and cardiovascular diseases in health care of the aged. 2003, Tokyo, Japan, Japan public health association (in Japanese).
• 24)  Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA 2001; 285: 2486-97.
• 25)  National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report. Circulation 2002; 106: 3143-421.
• 26)   Belloc  NB,  Breslow  L,  Hochstim  JR. Measurement of physical health in a general population survey. Am J Epidemiol 1971; 93: 328-36.
• 27)   Belloc  NB,  Breslow  L. Relationship of physical health status and health practices. Prev Med 1972; 1: 409-21.
• 28)   Braverman  MT,  Aaro  LE,  Hetland  J. Changes in smoking among restaurant and bar employees following Norway's comprehensive smoking ban. Health Promot Int 2008; 23: 5-15.
• 29)   An  LC,  Betzner  A,  Schillo  B, et al. The comparative effectiveness of clinic, work-site, phone, and Web-based tobacco treatment programs. Nicotine Tob Res 2010; 12: 989-96.
• 30)   Kottke  TE,  Brekke  ML,  Solberg  LI,  Hughes  JR. A randomized trial to increase smoking intervention by physicians. Doctors Helping Smokers, Round I. JAMA 1989; 261: 2101-6.
• 31)   van Wier  MF,  Dekkers  JC,  Hendriksen  IJ, et al. Effectiveness of phone and e-mail lifestyle counseling for long term weight control among overweight employees. J Occup Environ Med 2011; 53: 680-6.
• 32)   Irvine  AB,  Philips  L,  Seeley  J,  Wyant  S,  Duncan  S,  Moore  RW. Get moving: a web site that increases physical activity of sedentary employees. Am J Health Promot 2011; 25: 199-206.
• 33)   Smith  DT,  Carr  LJ,  Dorozynski  C,  Gomashe  C. Internet-delivered lifestyle physical activity intervention: limited inflammation and antioxidant capacity efficacy in overweight adults. J Appl Physiol 2009; 106: 49-56.
• 34)   Jacobs  N,  Clays  E,  De Bacquer  D, et al. Effect of a tailored behavior change program on a composite lifestyle change score: a randomized controlled trial. Health Educ Res 2011; 26: 886-95.
• 35)   Fukuda  H,  Haruyama  Y,  Nakade  M,  Muto  T. Relationship between lifestyle and change of cardiovascular risk factors based on a five-year follow-up of employees in Japan. Ind Health 2007; 45: 56-61.
• 36)   Yamamoto  A,  Temba  H,  Horibe  H, et al. Life style and cardiovascular risk factors in Japanese population —from an epidemiological survey on serum lipid levels in Japan 1990— (Part 1) Influence of life style and excess body weight on HDL-cholesterol and other lipid parameters in men. J Thromb Atheroscler 2003; 10: 165-75.
• 37)   Yamamoto  A,  Temba  H,  Horibe  H, et al. Life style and cardiovascular risk factors in the Japanese population—from an epidemiological survey on serum lipid levels in Japan 1990 part 2: association of lipid parameters with hypertension. J Atheroscler Thromb 2003; 10: 176-85.
• 38)   Hasegawa  T,  Nakasato  Y,  Sasaki  M. Factor analysis of lifestyle-related factors in 12,525 urban Japanese subjects. J Atheroscler Thromb 2005; 12: 29-34.
• 39)   Arai  H,  Yamamoto  A,  Matsuzawa  Y, et al. Serum lipid survey and its recent trend in the general Japanese population in 2000. J Atheroscler Thromb 2005; 12: 98-106.