The incidence and prevalence of diabetes continue to rise unprecedentedly, and this epidemic has accelerated in low-income and middle-income countries (LMICs) (1). China has the highest number of people with diabetes in the world (114.4 million in 2020), with type 2 diabetes (T2D) accounting for more than 90% of patients with diabetes and leading to a huge burden on healthcare systems (2). Increased understanding of risk factors for T2D can help reduce morbidity and mortality.

Transition in food consumption among Chinese residents has played an important role in the fast increase of T2D (3) and diet has been identified as a modifiable factor for occurrence and progression in diabetes by many researchers (4, 5). The Guideline for Prevention and Treatment of T2D in China (2020 Edition) also pointed out the necessity to set reasonable treatment targets for medical nutrition (6). Nowadays, dietary pattern analysis has emerged as an alternative and complementary approach to examine the relationship between diet and diseases. It considers the interaction and joint effects of different food items and examines the effects of the overall diet. Several studies have suggested that dietary patterns derived from factor analysis were valued in the nutritional epidemiology of diabetes but different patterns with positive or negative health effects such as the anti-inflammatory dietary pattern were applied (4, 7–9). Such disparities in dietary patterns may be partly due to the diversity in available food, food cooking, and eating habits over study populations in the context of various times, race, and culture, which meant that it should benefit the prevention and control of diabetes to understand particular recommended diet patterns for the local people.

The “refined grains and meat” dietary pattern and the “oil and salt” dietary pattern were reported to increase T2D risk in the Xinjiang Region (10), and individuals with the Western dietary pattern had an elevated risk of T2D in Zhejiang Province, China (11). Even in mainland China, dietary patterns varied over provinces from Northeast China to Southwest China. So far, there were no empirical data on the association between dietary patterns and diabetes in Southwest China. Several specialized dietary patterns were characterized by using factor analysis and their associations with T2D incidence were explored in this community population cohort from Southwest China.

Studies regarding the association of empirically derived dietary patterns and the risk of T2D in Southwest China were sparse so far. This prospective study identified two major dietary patterns by factor analysis, and both the junk food pattern and vegetable-grain pattern were significantly associated with T2D among the study population in Southwest China. Those findings remained robust in the sensitivity analysis among those followed up for 2 years or longer and with a BMI of 18.5 or greater at baseline. Moreover, significant positive linear trends were observed for associations between junk food pattern, and averages of follow-up FPG and its changes, and inverse linear trends existed in the associations between vegetable-grain pattern, and averages of follow-up FPG and its changes. Those results were in agreement with previous studies reporting that various dietary patterns were related to incident T2D (18, 19). They suggested that a healthy diet pattern may benefit the prevention and control of T2D and some interventions should be developed to improve the diet behaviors.

In some researches, fruit juices, sweetened beverages, sweets, and desserts had been proved to be less healthy plant foods for diabetes (19, 20). A meta-analysis based on 5 studies suggested that “unhealthy” dietary pattern like consuming higher amounts of high-fat dairy and fried products was associated with a 44% increased risk of diabetes (4). In this study, low intake of junk food pattern, characterized by high consumption of fried food, soft drinks, and desserts resulted in a 28% reduction in risk of T2D compared with medium intake. Dietary components in such a pattern that may affect T2D risk included higher saturated fat and glycemic load from high intakes of snacks and desserts (21, 22), as part of the western pattern's unhealthy constituents. The high glycemic load had been documented to adversely affect insulin sensitivity, a well-established detrimental factor for diabetes (23, 24). On the one hand, a high level of added sugar and calorie content strongly increases the risk for diabetes with weight gain, worse metabolism of glucose, and lower insulin sensitivity (25, 26); on the other hand, advanced glycation of high-fat products can contribute to enhanced oxidative stress and higher inflammatory markers that may be accompanied with insulin resistance (27). But the increased risk was not observed significantly in the high level of such a pattern. One of the possible reasons was that the intake did not reach the dangerous thresholds in the study population. Another was that the sample size was not enough for such association in this study.

Also, the vegetable-grain pattern was observed a protective effect on incident T2D in this study, with a 20% lower T2D risk, which was similar to the results from a previous systematic review and meta-analysis (28). Jordi et al. provided evidence on the inverse relationship between adherence to a plant-based dietary pattern and the risk for developing T2D, too (5). Several potential mechanisms may explain the observed favorable association. Firstly, whole grains and vegetables may bring a rich array of dietary fiber, which has been proved in several prospective studies to reduce levels of inflammatory markers (29, 30). Secondly, one previous study in Chinese patients with diabetes has recognized increasing fiber and rich antioxidant intake from vegetables as an effective approach to lower HbA1c levels (31) and further reduce T2D risk. Thirdly, whole grains, tubers, and vegetables in this pattern had a low glycemic index and may help to improve insulin sensitivity, and thereby enhance glycemic control (19). Of course, it is most likely a combination of dietary factors responsible for the robust associations. However, the findings remain controversial. Contrary to our results, Erber et al. showed that the vegetable pattern did not confer a lower risk for diabetes in women (32). Another previous research also reported refined grains and meat patterns to be a major risk factor for diabetes (10). Although the vegetable-grain pattern in this study shared some components with other patterns, no meat was contained and the concrete effect partly depended on the proportion and category of food items, like the whole or refined grains (23), and dark or light vegetables. The complex difference in components of different patterns may be due to the diversity in race, age, culture, and lifestyle over the different study populations.

To the best of our knowledge, there were few studies on dietary patterns derived from factor analysis in Southwest China. This study highlights the various effects on T2D of different dietary patterns which may be targeted for local people and emphasizes the importance of considering the quality of foods consumed for diabetes. This study had other strengths. On the one hand, this prospective cohort study filled the gap of the association between dietary patterns and T2D risk in Southwest China among diverse minority groups with different food preferences, and those findings offered further insights into the dietary structure for T2D prevention among the adult Chinese population. On the other hand, dietary pattern beyond a single nutrient or food is easier to make recommendations for the local population and improve dietary behaviors for local residents. The limitations of this study should also be noted. First, the baseline covariates information used in the analysis might lead to residual confounding if those covariates may be time-varying. Second, dietary pattern analysis requires subjective decisions in naming the food pattern although some previous studies can be referred to. Third, the effects of different foods may be related to cooking ways, which were not considered in this study. In addition, diet information was self-reported by subjects, and measurement errors were inevitable which may bias the findings from this study.

Higher adherence to the vegetable-grain pattern and lower adherence to the junk food pattern significantly lowered T2D incidence among the Chinese population in Southwest China. Those findings added evidence on the recommendation of healthy dietary patterns to prevent incident T2D for the Chinese adult population in China, especially in Southwest China. Some interventions and policies should be developed and implemented to improve the diet behaviors to prevent and control T2D better among the Chinese population in the future.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by the Institutional Review Board of Guizhou Province Center for Disease Control and Prevention (No. S2017-02). The patients/participants provided their written informed consent to participate in this study.

LX and YW conceptualized, designed and conducted the study with the help of NW, LZ, FZ, KX, TL, and CF. LX and YW wrote the first draft of the manuscript. NW, LZ, FZ, KX, TL, and CF contributed significantly to the revision of the manuscript. All authors read and approved the final manuscript, had full access to all the data in the study, and accept responsibility for the decision to submit for publication.

This work was supported by the Guizhou Province Science and Technology Support Program (Qiankehe [2018]2819).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.