Changes In Dietary Patterns And Their Associations With General And Central Obesity Among Adults In China (1991-2009)

BackgroundWith the rapid development of the economy, China has experienced nutrition transition. At the same time, the overweight and obesity rates of Chinese adults were rising rapidly in the most recent decade. The traditional dietary analyses focused merely on the relationship between individual nutrient or food and obesity, which has certain limitations. However, dietary pattern analysis examines the effects of overall diet. Conceptually, dietary patterns represent combined effects of foods and nutrients and thus may be more effective to shed light on the impact of diet on health outcomes.ObjectiveTo investigate characteristics and dynamics of dietary patterns among Chinese adults as well as the potential influencing factors of these dietary patterns; To examine the effects of different dietary patterns on overweight and obesity.MethodsThe study was based on the data from the China Health and Nutrition Survey (CHNS), an ongoing longitudinal cohort study. According to the aims of the present study, study subjects were divided into two groups. The first group included those adults aged18to80years who participated in at least one wave of surveys between1991and2009. Factor analysis was performed to establish dietary patterns of each survey year, and the nutrition composition and potential influencing factors for each dietary pattern were investigated further. Meanwhile, the18-year shifts in dietary patterns between1991and2009were also examined. The second group included those who were involved in both the1993and2009survey years and had completed dietary and anthropometric data. Multilevel mixed effect model was conducted to examine the effects of different dietary patterns on overweight and obesity risks of adults.Results1. The overweight and obesity rates of adults in China increased significantly between1991and2009:The rate of overweight increased by73.9%(16.1%vs.28.0%), the rate of general obesity increased by156.7%(3.0%vs.7.7%). Central obesity rate increased by85.5%(24.2%in1993vs.44.9%in2009).2. Rice and flour were frequently consumed. The intake of rice was declining and the intake of flour decreased a little. The intakes of pork, fish and shrimp were very high and on the rise. Meanwhile, the intake of tubers was also growing. However, the intakes of soy products declined slightly. Since2004, the intakes of fast food, desserts and cakes had grown rapidly. Male vegetable intake was about350g/d, and female was about330g/d. Both vegetable intakes were remained stable. The intakes of fruits and eggs were increased. Dairy consumption was not stable.3. Four dietary patterns were identified among men and women. A southern pattern was rich in rice, vegetables and pork. A northern pattern was rich in flour, other cereals and tubers. A snack pattern was rich in milk, fruits and eggs. A fast food pattern was rich in fast food, dairy and snacks.4. An alcohol pattern was identified specially among men, that was rich in meat, alcohol and nuts. Two dietary patterns were identified specially among women. A meat pattern was rich in livestock, poultry, fish and shrimp. A coarse cereal pattern was rich in tubers and other cereals.5. For men, the correlation coefficients for southern pattern, meat pattern and snack pattern between1993and2009were0.62,0.32and0.24, respectively(p<0.01). For women, the correlation coefficients for southern pattern, meat pattern and snack pattern between1993and2009were0.63(p<0.01),0.32(p<0.01) and0.05(p>0.05), respectively.6. Multilevel mixed effect linear regression model of dietary patterns indicated that community-level variation accounts for66%,17%and18%of total score variation of southern, meat and snack dietary patterns among men respectively, whereas community-level variation explained65%,17%and19%of each dietary pattern among women, respectively. Therefore, community-level factor was the major part for score variation of the southern dietary pattern. Individual-level factors in relation to dietary patterns included age, education, income level, physical activity, smoking, and alcohol consumption.7. Men in the top quintile of the southern dietary pattern had lower risk of overweight/general obesity and central obesity compared to men in the lowest quintile of the southern dietary pattern(OR,0.53and0.60). The southern dietary pattern in women was negatively associated with waist circumference (p<0.01).8. Men in the top quintile of the snack dietary pattern had higher risk of overweight/general obesity and central obesity compared to men in the lowest quintile of the snack dietary pattern(OR,1.66and1.88).ConclusionsFrom1991to2009, there were various dietary patterns among China adults. And those dietary patterns kept changing over time. The traditional dietary pattern remained relatively stable compared to other patterns. Dietary patterns were influenced by community-level factor, individual-level factor and time factor. Due to differences in food nutrients and components, the effects of dietary patterns on adult obesity were varied. The southern dietary pattern was associated with reduced general obesity and central obesity in men and was negatively associated with waist circumference in women, which may serve as a protective factor. The snack pattern in men was positively associated with both general and central obesity.