Assessing quality of life using WHOQOL-BREF: a cross-sectional study on the association between quality of life and neighborhood environmental satisfaction, and the mediating effect of health-related behaviors

Hong Kong consists of three main territories including Hong Kong Island, Kowloon Peninsula and New Territories. There are 127 district council constituency areas in the Kowloon Peninsula. Nine district council constituency areas near Hung Hom and Tsim Sha Tsui located in the southwest of Kowloon Peninsula were selected as the target study sites. These areas cover mixed-use commercial and residential districts, old urban and more affluent districts with private and public housing, and people with low and high SES [20]. According to the 2011 Hong Kong Population Census, there were 158,103 people living in these areas [20]. Among this population, 88.0% were Chinese and 53.8% were female. Age distribution was comparable to the whole population in Hong Kong. Residents aged 18 years or above and had been living in the nine areas for at least 90 days were the target population. Those who were cognitively impaired or unable to communicate effectively in Cantonese, Mandarin and English were excluded. Assuming that 50% of the population was satisfied with the environment, a sample size of 266 would provide a precision of 6% from the true values at 95% confidence level.

The validated WHOQOL-BREF (Hong Kong version) was used to measure QOL [1, 21]. It consists of 24 items to assess perception of quality of life in four domains, including physical health, psychological, social relationships and environment, and two items on overall QOL and general health. The domain scores were transformed into a linear scale between 0 and 100 following the scoring guidelines [1]. A higher score indicated a better QOL. Physical activities were assessed by the International Physical Activity Questionnaire short form (IPAQ-SF) which is adequately reliable and valid in a Chinese population [22]. Walking, moderate-intensity and vigorous-intensity activities were assessed. The total MET-minutes/week was calculated for each participant. Metabolic Equivalent of Task (MET) is the energy cost of physical activities [23]. Physical activity levels were categorized as low, moderate and high, based on the criteria listed on the IPAQ guidelines [22]. People have to answer all questions on number of days in a week and daily time performing walking, moderate and vigorous activities in order to calculate their total MET and identify their physical activity levels. Those who were unable or refused to answer all these questions were removed from the analysis according to the IPAQ guidelines [22]. Smoking, alcohol drinking, and vegetable and fruit intakes were also assessed. For practicing low fat, low salt and low sugar diets, respondents could choose “never (1)”, “seldom (2)”, “sometimes (3)” and “always (4)”. Seven questions on satisfaction with the neighborhood environment with a five-point Likert scale (1: very unsatisfied; 5: very satisfied) were used to evaluate air quality, ventilation, drinking water quality, noise condition, lighting condition, environmental hygiene, and environmental protection and recycling. Six questions on satisfaction with open spaces including greening, parks and gardens, recreation and sports facilities, promenade, rest areas and pedestalization were also assessed using the same Likert scale. Socio-demographic profiles of the residents were also assessed.

Test-retest reliability of the questionnaire had been tested with 64 subjects before implementation of the main study. All question items showed a Cohen’s kappa value ranged from 0.50–1.00 or an intraclass correlation coefficient value ranged from 0.55–1.00, which indicated that the reliability of the questionnaire was fair to excellent.

This was a cross-sectional study using a convenience sampling method to recruit potential participants. Trained interviewers approached people at parks, resting areas and outside food markets and shopping centers in the nine district areas from 9 am to 7 pm. In addition to weekdays, the questionnaire interviews were also conducted on weekends to reach residents who spent most of their time in other districts on weekdays. After confirming the eligibility of participation, the purpose of the questionnaire survey was explained. If the potential participants were in a group, only one family member from each household was allowed to answer the questionnaire. A verbal consent was obtained from each of the participants. The answered questionnaire and the signing for receiving the token of appreciation were also the implied consent to participate in the study. A face-to-face interview was conducted on the recruitment site. A HKD 50 (approximately USD 6.45) supermarket shopping voucher was given as a token of appreciation to those who completed the questionnaire.

Data analysis was performed using SPSS version 21. Descriptive statistics were reported by mean ± standard deviation or proportion, as appropriate. Satisfaction levels with physical environment and open spaces across different districts were compared using chi-square tests. Relationships between QOL in the four domains, overall QOL and general health were assessed using Spearman’s correlations.

Mediation analysis was conducted to study the effects of overall levels of satisfaction with the neighborhood environment (independent variable) on QOL in the four domains, overall QOL and general health (criterion variables), as well as the potential mediating roles played by various health-related behaviors (Fig. 1). To obtain the overall levels of satisfaction with the neighborhood enevironment, the mean scores of those environmental and open spaces variables which showed significant differences across the nine districts were summed to calculate the range of the overall environmental satisfactory scores (1: very unsatisfied; 5: very satisfied). Scores in the lowest 1/3 were categorized as “low satisfaction”; the second 1/3 were categorized as “moderate satisfaction”, while the highest 1/3 were categorized as “high satisfaction”. To conduct the mediation analysis, the existence of relationships between the independent variable and criterion variables were assessed using one-way ANOVA and linear regression in the first place, followed by assessing the existence of relationships between the independent variable and the potential health-related behavior mediators with regression tests. Mediated models with the combined effects of overall levels of satisfaction with the neighborhood environment and each of the health-related behaviors in predicting QOL were evaluated using multiple linear regression tests. Regression coefficients were used to assess the effect of mediation and further confirmed by both the Aroian and Goodman tests. A P-value of < 0.05 was considered as statistically significant.

A total of 317 residents were successfully interviewed during July – September 2015. The majority were Chinese (99%), 44% were male, 60% were married and the mean age was 45.12 years (SD = 17.28) (Table 1). Nearly 46% had been living in the nine district areas for more than 10 years and 46% were living in a self-owned private permanent housing. About 33% had attained a university level. Around 25% claimed that they had no income while 31% received an individual monthly income more than HKD 14800 (USD 1922), the median monthly income of the Hong Kong population in 2014 [24].

A total of 10.5% of the residents smoked every day, almost every day or occasionally (Table 1). About 18.3% claimed that they consumed alcohol at least 2–4 times/month. Regarding fruit and vegetable intakes, 13.2% and 7.6% had achieved the recommended intake of at least 2 servings of fruits and 3 servings of vegetables a day, respectively. More than 60% claimed that they sometimes or always practiced low fat, low salt or low sugar eating. For physical activity, those who were unable or refused to estimate their number of days or daily time performing walking, moderate and/or vigorous activities were removed from the analysis. 240 out of 317 respondents completed the IPAQ assessment. Most of them were categorized as having moderate physical activity level (62.9%), and 13.3% were having low activity level or not performing physical activity.

The residents were most satisfied with water quality (55.4%) in their living districts, whereas they were most dissatisfied with air quality (29.7%), ventilation (27.1%) and noise level (25.9%). For open spaces, they were most satisfied with the promenade (36.9%) but another 22.7% were dissatisfied with the promenade. On the other hand, 25.2% and 22.1% were dissatisfied with the pedestrianization and greening in the open spaces, respectively. Crosstabulations of the living districts and satisfaction with physical environment and open spaces found that satisfaction levels with air quality (P = 0.029), noise condition (P = 0.041), parks and gardens (P = 0.015), recreation and sports facilities (P = 0.034) and promenade (P < 0.001) were significantly different across the nine districts (Table 2).

Table 3 shows the scores of QOL in the four domains, overall QOL and general health. Comparing the four domains of the residents, physical health domain was the highest with a mean score of 70.83 ± 12.69 while the environmental domain was the lowest with a mean score of 61.98 ± 13.76. Using one SD below the mean as the cut-off standards for low QOL [23], 18.9% of the respondents were considered to have poor QOL in the physical domain, followed by environmental domain (16.7%), psychological domain (14.5%) and social relationships domain (10.4%). The four domains, overall QOL and general health were significantly and positively interrelated with low to moderate relationships (r = 0.30–0.53, P < 0.01) (Table 3).

A convenience sampling method was used in this study and residents who participated were mainly those who stayed outdoors or used the facilities and amenities in the neighborhood. Those who spent most of their time indoors or at home were less likely to be approached. These limited our studying of the QOL of residents and their perspectives of the environment. A random sampling telephone interview could be an alternative in the future. The low specificity but high sensitivity of the IPAQ-SF and the potential of inaccuracy of self-reported data could also lead to the insignificant mediating effect of physical activity levels, walking activity and total physical activity observed on the relationship of satisfaction with neighborhood environment and QOL in the psychological domain. Individual in-depth interviews or focus group discussions together with questionnaire survey can help data triangulation to facilitate understanding of the reasons behind the ratings of the environment and open spaces. In future studies, objective data such as vegetation density, air pollutants and traffic noise should be considered to supplement the self-rated environmental data. Other factors like residential satisfaction, sense of community, and family and neighbor relationships which may potentially mediate the effect of neighborhood environment on QOL can be further investigated.