Conception of health has been changed with the development of medicine and medical sciences since 1970's. Health is defined as a dynamic state of human wellbeing characterized by a physical, mental, and social potential which satisfies the demands of a life corresponding to age, culture, and personal responsibility, and not merely the absence of disease or infirmity. Health related quality of life (HRQL) is an individual's satisfaction or happiness with the dimensions of life insofar as they affect or are affected by "health" as defined above. HRQL has been introduced to assess people's health status. To date, a number of questionnaires have been developed to evaluate HRQL, and the 36-item Short Form Health Survey (SF-36) is the most commonly used one.

SF-36 was developed from the Medical Outcomes Study or RAND Health Insurance Experiment [1]. It is a short-form derived from a larger 149-item instrument and is more precise than its predecessor, SF-20 [2]. SF-36 has been proven useful in monitoring population health, estimating the burdens of different diseases, monitoring outcome in clinical practice, and evaluating medical treatment effects. It has been translated into many languages with its content examined cross cultures [3–6]. In mainland China, the Mandarin SF-36 has been used in some surveys to assess the quality of life of the population with special chronic diseases [7, 8]. However, the surveys on general populations were conducted only in Sichuan and Hangzhou [9, 10].

In this study, we aimed at (1) testing the reliability and validity of the Mandarin version of SF-36; (2) assessing health related quality of life in the population of Shanghai, China; and (3) evaluating risk factors that may significantly influence HRQL. The study was approved by the Second Military Medical University Ethics Committee.

Quality of life is a study area that has attracted increasing interests over the past two decades. SF-36 has been used as an instrument for assessing quality of life world-wide. Normative data have also been obtained in many countries [25, 26]. In mainland China, the Mandarin SF-36 has been used in a few surveys to assess the quality of life of general population and the population with special chronic diseases [7–10]. Our survey evaluated the feasibility of using the Mandarin SF-36 for investigating health related quality of life in the population of Shanghai, China.

Statistical analyses used in this study included split-half reliability coefficient, ICC, paired-sample t test for the difference between the test and retest scores, and Cronbach's α. The results indicated that SF-36 was quite stable for the purposes of the study with a good internal consistency.

In particular, the SF dimension had the lowest Cronbach's α coefficient in this study, which was consistent with other surveys using the Mandarin version of SF-36 [10, 21, 27, 28]. The SF dimension also had the lowest ICC and split-half reliability, indicating there might be some problems in the conceptualization of social function. Traditionally, Chinese people don't think much about social function, and commonly say little or nothing about how the physical health or emotional problems would interfere with their social activities. In addition, the SF dimension included two questions as follows: (1) "during the past 4 weeks, to what extent has your physical health or emotional problems interfered with your normal social activities with family, friends, neighbors, or groups?", and (2) "during the past 4 weeks, how much of the time has your physical health or emotional problems interfered with your social activities (like visiting with friends, relatives, etc.)?" It appeared that the answers of the two questions had reverse orders, which may lead to the low reliability. The other reason may be the cultural diversity. In China, "social activities", translated as "she hui huo dong", refer to not only the everyday life within a family or one's circle of friends, but also the formal activities with other people such as going to a ballroom dancing event or attending a conference. The misunderstanding may result in the low reliability.

MH and VT also had relatively low reliabilities [[10, 21, 27], and [28]]. The Cronbach's α coefficients for the VT and MH dimensions were 0.66 and 0.75 in the survey of Hangzhou [10], 0.72 and 0.71 in Sichuan [27], 0.74 and 0.77 in Hong Kong [28], 0.73 and 0.74 in American Chinese [21], and 0.78 and 0.69 in our study, respectively. This may be due to the characteristics of Chinese people since they are not used to talking about their feelings and emotions in public.

Our results indicated a credible construction validity of SF-36 that was consistent with the outcomes of other surveys [10, 29]. Factor analysis proved that our results were basically in accordance with the theoretical construction of SF-36. Correlation analysis indicated that each of the 36 items was highly correlated within the hypothesized dimensions, while relatively low correlations were observed between the items and other dimensions.

Therefore, we concluded that SF-36 was acceptable and applicable for evaluating the quality of life in the general population of Shanghai, China. Compared our survey with other studies, American Chinese had the worst quality of life among different Chinese populations. Shanghai population had the best quality of life, even better than American and Canadian [9, 10, 21–25]. It should be noticed that the other studies in the comparison were undertook much earlier in time than our survey, and China has made impressive progresses in living standard during recent years. Especially, Shanghai is the financial and commercial center of China with the best medical and sanitation conditions. For example, the average life expectancy of the Shanghai population was 81.08 years old in 2007, which is slightly lower than the average life expectancy of Andorra, Macau, Japan, Singapore, San Marino, Hong Kong, and Canada. The infant mortality rate was 3.0‰ and maternal mortality rate was 6.68 deaths per 100,000 live births [30, 31]. All these factors may lead to high HRQL in the Shanghai population.

In addition, we found a very interesting fact about the normative values of the SF-36 dimension scores. Although the female had worse HRQL than the male in most subgroups, in some subgroups female did report a better mental health. The same results were found in other Chinese populations [9, 10, 21, 22], a fact which is not usual in non-Chinese population. The outcome indicated the serious mental problems in Chinese men, which might be due to the huge stress in both work and life.

We found that region, gender, current job, current marital status, the highest level of education, total income of family per month, frequency of activities, BMI, and chronic diseases had influences on at least one SF-36 dimension. But drinking and smoking did not significantly affect HRQL. When the interaction effects among these factors were excluded in multivariate regression, some risk factors such as resident region, chronic diseases, current job, frequency of activities, and age had strong influences on three or more SF-36 dimensions, while current marital status, the highest level of education, and total income of family per month affected only one or two of the SF-36 dimensions. These results were analogous to the previous study in Sichuan in which chronic condition, personal income, inhabitant places, age, and educational level were found to be the significant risk factors influencing quality of life, while marital status had impacts on few SF-36 dimensions [32]. All of the SF-36 dimensions were remarkably impaired by chronic diseases. People with chronic diseases had a worse quality of life than those without. It had been considered as the main risk factor impairing quality of life [7, 8, 32]. The PF, MH, GH, and VT dimensions were highly correlated with frequency of activities. Everyday activities, such as housework and walking, may help to stay healthy. The impact of age on quality of life was also notable. It was evident that the health problems became more and more serious with increasing age.

There are some limitations in this study. Detailed information on non-responders were not collected, we were not sure whether there were differences between responders and non-responders. Although the interviewers received uniform training, there still might be influence of the interviewers' explanation on the results, and it was difficult to evaluate, which was also the limitation of this survey. Migrant workers, who make up a significant portion of the Shanghai population, were unable to be sampled because they remain officially registered in their place of origin. In addition, the sampling in the suburbs should be considered more carefully. Since the 1990s, urban population increased rapidly due to economic development and suburbanization in Shanghai. More and more people settled down in suburban regions, especially the group of white collars [33]. It made the distribution of suburban population more complicated.

In summary, the Mandarin SF-36 is a valid and reliable questionnaire for evaluating both physical and mental health status. The quality of life in the Shanghai population is quite good compared with those in other Chinese populations. The primary influencing factors are region, chronic diseases, age, and frequency of activities. The key to improving quality of life includes the prevention and control of chronic diseases, and participating in moderate and regular activities. In addition, the elderly people should pay more attention on quality of life.