The intervention is called “zhiji management”; zhiji is a Chinese word meaning “bosom friends.” Zhiji management is a 3-month lifestyle intervention program designed to quantify and balance dietary energy intake and energy consumption from physical activity and to provide practical suggestions on positive and safe lifestyle changes. All participants kept their medical treatment as usual.

Each participant was supplied with an electronic accelerometer-like energy consumption monitor (Zhiji Energy Monitor, YHKY Ltd, Beijing, China) to wear all day long except when sleeping or showering. The monitor recorded physical activity in terms of duration in minutes and intensity measured by metabolic equivalent (MET) of the activities (11). Effective physical activity was defined as any activity between 3 and 6 METs and lasting at least 6 minutes. Because most study participants were chronically ill older people, we adopted a shorter duration than the 10 minutes recommended in the US guideline (12). Participants were also instructed to keep a dietary diary at least 2 days a week, a workday (Monday through Friday) and a weekend day.

Health-related individualized consultations were provided during the 3-month intervention. Trained community physicians in each clinic, aided with customized computer software designed for the zhiji intervention, met study participants individually for weekly consultation sessions in the first month and biweekly sessions in the following 2 months, for a total of 8 sessions (Figure 1). During each 35- to 40-minute session, the community physicians downloaded physical activity data from the accelerometer to the computer in the clinic and converted the information from the dietary diary to computerized data. The software produced a detailed physical activity and dietary intake status, illustrated by charts and tables and individualized “prescriptions” for lifestyle modification. Community physicians explained these results to each participant and encouraged the participant to engage in appropriate, safe, and effective physical activity and to follow dietary recommendations in line with the Chinese dietary guidelines (13). The physicians also suggested ways to balance energy intake and consumption. The computer printouts with measurable targets and the prescription were given to the participants to take home as visual aids and reminders.

At the time of recruitment, marital status, educational level, employment status, and self and family medical history were collected through a standardized questionnaire administered by community physicians, and body weight and height were measured. During each of the 8 interventional sessions, body weight, waist circumference, SBP and DBP were measured by community physicians. Body weight was measured by an electronic scale (Xihen, Wuxi Scale Co, Wuxi City, Jiangsu Province, China) with subjects wearing light indoor clothing and no shoes. Waist circumference was measured at a level midway between the lowest rib and the iliac crest by a measuring tape. After a 5-minute rest, blood pressure was measured twice with an electronic blood pressure meter (Omron HEM-7200, Omron (China) Co, Ltd, Dalian City, Liaoning Province, China) and the average values were used in all analyses. In addition to these measurements, the early group had 2 additional measurements in the second and third month during the 3-month postintervention follow-up period and the late group had 2 measurements in the first and second month. For participants with diabetes only, blood tests of fasting glucose, 2-hour postprandial glucose, and glycated hemoglobin (HbA1c) were performed during the intervention and follow-up on a voluntary basis.

We compared selected characteristics of the early and late groups (percentages for categorical variables; mean and standard deviation for continuous variables). Within-group (preintervention and postintervention) and between-group differences in categorical variables were tested by χ² tests and continuous variables with paired t tests and independent t tests. Intention-to-treat analysis using the method of “last observation carried forward” was applied for all measures. Multivariable analysis was conducted to test the effect of demographic characteristics on the outcomes. All data analyses were performed by using SPSS version 15 for Windows (SPSS Inc, Chicago, Illinois).

The study followed the principles of the Declaration of Helsinki. This study was approved by the Project Examination Committee of the Tianjin Department of Labor and Social Security. All participants gave their written informed consent before participating in the study.

After the intervention, daily effective physical activity increased significantly (P < .001) in total energy expenditure (54.6 kcal) and duration of exercise (9.6 min) (Table 2). Daily total energy intake decreased significantly (P < .001) (328.5 kcal) with a small increase in total energy from protein and a small decrease from fat. Results were generally consistent between the early and the late group for the 3 months of the intervention phase.

During phase I, body weight, waist circumference, SBP, and DBP decreased significantly (P < .001) in the early group that received the 3-month intervention, and most indicators worsened in the late group that served as controls during the same phase (Table 3). The net between-group reductions were substantial and significant for weight (−2.6 kg), waist circumference (−2.9 cm), SBP (−10.9 mm Hg), and DBP (−4.0 mm Hg); all P values < .001. Results were similar for the subgroup of patients with diabetes for whom we had glucose and HbA1c data.

Multivariable analysis was conducted to control for other potential dependence-associated factors. The outcome variables were the 7 biophysical indicators, and the independent variables included group (intervention vs control), the outcome variable’s corresponding baseline level, sex, marital status, education, and occupational status. The multivariable results confirmed that the net differences between the early group (intervention phase) and late group (control phase) were highly significant for all the following biophysical indicators assessed: weight, waist circumference, SBP, DBP (all P values < .001). Among patients with diabetes, the intervention significantly improved 2-hour postprandial glucose and HbA1C results (P < .01), but the effect on fasting glucose levels was borderline significant (P = .08).

During the intervention, a generally consistent pattern of gradual improvement was observed for all markers and both intervention groups (Figure 2). No further improvement was observed during the follow-up in the early group for whom such data were available. Nevertheless, the effect of the intervention was maintained for all markers assessed. Additional analyses based on participants without any missing values for all sessions (early group = 135, late group = 78) revealed consistent results.

The intensive zhiji intervention led by community physicians was effective in encouraging the participants to adopt and maintain a healthier lifestyle in terms of physical activity and diet. This finding is consistent with the findings of most other studies of this kind, including the Diabetes Prevention Program (DPP), the study by Lindström et al, and other studies (2,8,14–17). Also, the Look AHEAD (Action for Health in Diabetes) research group found that lifestyle intervention was effective among diabetes patients (6). In addition to these findings, our study added further evidence showing that a community-based intensive lifestyle intervention is effective at getting patients with mild diabetes (including prediabetes) or hypertension (including prehypertension) to adopt and maintain a healthier lifestyle.

We found that total physical activity did not increase but total effective physical activity significantly increased through the intervention. In this intervention program, the concept of effective physical activity was emphasized, with patients instructed to engage in physical activities that reach a certain level of intensity (at least 3 METs) to achieve health benefits but are not too strenuous to be unsafe (more than 6 METs). The instantaneous quantification of activity intensity in MET was made possible by the energy consumption monitor, which had more sophisticated functions than pedometers.

The zhiji intervention program in this study shares some common features with other programs such as the DPP (2), including being individualized, being delivered by a case manager (in our study, by community physicians), and having frequent contact between physician and patient. However, unlike the DPP or other studies (2,4,18–20) that used goal-based behavioral intervention and a structured diet and physical activity plan for the core curriculum, our study adopted a “free choice” lifestyle modification model, operating under well-established guidelines. This model may have empowered patients’ self-management skills through enabling them to select the most comfortable and easy to adopt behavioral changes.

Significant reductions were observed in anthropometrics (weight and waist circumference), SBP, DBP, and glucose measures. These findings were consistent with those of previous studies in which intensive individual-based dietary and physical activity modification programs produced similar mean losses in weight (17,19,21,22), waist circumference (4,17,19,23), and blood pressure (4,16,24,25). Most previous studies were conducted among individuals at high risk of developing diabetes or hypertension. Our study demonstrated that intensive lifestyle modification in the community setting could also lead to improvements in these markers in patients with mild diabetes (including prediabetes) and hypertension (including prehypertension).

The chronic care model provides a framework for a systematic approach with 6 key elements that are critical to effective management of chronic diseases (26–29). The 6 elements are the community, the health system, self-management support, delivery system design, decision support, and clinical information systems. Although not explicitly designed on the basis of this model, this intervention program incorporated all 6 of those elements.

Our study has several limitations. First, this trial was not randomized. Nevertheless, the late group served as controls during phase I of the study, and provided a way to assess net changes in many measures instead of only pre–post comparison. In addition, both groups had similar demographic characteristics. Second, follow-up data were limited to the early group and lasted only 3 months. Therefore, the long-term effect of the 3-month intervention could not be assessed. Third, our study sample was derived from volunteer participants in 5 community clinics, and it was possible that these participants were not representative of all patients with hypertension or diabetes. Nevertheless, no systematic differences were detected between the early and the late group, making them generally comparable. Lastly, there may be contamination for the late group while serving as controls because they lived in the same community with the early group participants undergoing intervention. However, the outcomes improved in the early group while most of them worsened in the late group, suggesting that contamination, if any, was limited.

Our study also has several strengths. First, the intervention program was carefully designed to use, and modify when appropriate, established physical activity and dietary guidelines. Second, the program was well accepted by study participants, and completeness of data collection for behavioral and clinical measures was better than they were for most other studies. Third, our study provides important evidence regarding the effectiveness of these interventions at not only preventing diabetes and hypertension but also at managing these conditions. After the study was complete, the zhiji program was implemented in more communities in several cities in China such as Beijing, other areas of Tianjin, and Xi’an.

The burden of diabetes and hypertension has been rising fast in developing countries. Previous evidence on the effect of lifestyle modification focused on prevention of hypertension and diabetes among people who did not have these conditions but were at high risk of developing them. Highlights of the innovative features of our intervention included training and delivery of the intervention by community physicians, supplying participants with energy monitors, emphasizing effective physical activity, using computer software for electronic decision support and for printing a tailored intervention program for each participant, and empowering patients with self-management skills and community support. This community-based individualized lifestyle intervention program produced short-term beneficial changes in activity, diet, and clinical parameters in patients with diabetes or hypertension. Its cost-effectiveness or feasibility in scaling up into more communities is not yet clear. Future large-scale research with longer follow-up is needed to fully evaluate the cost-effectiveness and feasibility of such a lifestyle intervention among patients with diabetes, hypertension, or other chronic disease.