The Effects of Dietary Nutrients On Serum Liquids in Chinese Dietary: A Community-Based Population Study

Background To examine the association between dietary nutrients (protein, carbohydrates, total fats, saturated fatty acids [SFA], monounsaturated fatty acids [MUFA], polyunsaturated fatty acids [PUFA], and cholesterol) and various serum liquids in Chinese adults. from various using the Table Blood were collected to test total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C).


Abstract Background
To examine the association between dietary nutrients (protein, carbohydrates, total fats, saturated fatty acids [SFA], monounsaturated fatty acids [MUFA], polyunsaturated fatty acids [PUFA], and cholesterol) and various serum liquids in Chinese adults.

Methods
As a part of Prospective Urban and Rural Epidemiology (PURE) study, 46,285 Chinese participants were recruited. Dietary nutrients from various foods were computed using the Chinese Food Composition Table databases. Blood were collected to test total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and highdensity lipoprotein cholesterol (HDL-C). 42,054 participants in PURE-China study were included. Mean TC, LDL-C, HDL-C, and TG were 4.68 ± 0.99 mmol/L, 2.63 ± 0.76 mmol/L, 1.36 ± 0.33 mmol/L, and 1.55 ± 1.09 mmol/L in our study, respectively. Subjects consuming higher proteins were more likely to have higher TC, TG, LDL-C (all P trend <0.001). High carbohydrate intake may play a role on reduction of TC (P trend <0.001), TG (P trend =0.031), LDL-C (P trend <0.001) and HDL-C (P trend <0.001). Higher fat intake was associated with increase of TC, LDL-C, and HDL-C (all P trend <0.001). Unlike SFAs and MUFAs, those consuming higher PUFAs are likely to have higher TG (P trend <0.001), but have stable LDL-C (P trend =0.136).

Conclusions
High carbohydrate intake could decrease signi cantly various liquids, while high-fat and high-protein intake might have harmful impacts on pro le of serum liquids. Subjects consuming higher SFAs and MUFAs were more likely to have higher levels of TC and LDL-C, but the intake of PUFAs has no in uence on LDL-C level.

Background
The atherosclerotic cardiovascular disease (ASCVD) is the leading cause of death and disability in the world [1,2]. Increased LDL cholesterol level is the major atherosclerotic factor, which have been identi ed by considerable amounts of observational and intervention studies [3,4]. In addition, decreased high-density lipoprotein cholesterol (HDL-C) can contribute to the development of ASCVD [5]. However, the ndings are still inconsistent in other studies [6,7]. Thereby, correcting lipid pro le was recommended as one of the most important strategies to prevent and control ASCVD risk in many countries [4,8,9].
Low-carbohydrate (LC) diet is recommended as a popular weight-loss approach compared with the conventional low-fat (LF) diet [10,11]. Several previous studies using systematic review and meta-analysis provided supportive evidence that LC diet produced greater weight loss, greater reduction of total cholesterol (TC) and triglycerides (TG), as well as greater increase of HDL-C compared with LF diet, but signi cant increase or lack of reduction in LDL-C was observed after consuming LC diet [12]. As we know, LDL-C plays an important role for ASCVD morbidity and mortality, so the bene ts of LC diet must be balanced with possible detrimental effect of increase LDL-C. Major types of dietary fat have been evaluated for preventing from ASCVD in moderate numbers of observational studies and clinical trials, suggesting that saturated fatty acid (SFA) intake should be replaced with polyunsaturated fatty acids (PUFAs) intake rather than monounsaturated fatty acids (MUFAs) or carbohydrate intake [13,14], though a recent meta-analysis did not support high consumption of polyunsaturated fatty acids and low consumption of total saturated fats [15]. However, most relevant studies were conducted in Europe and North America. It is not known whether their ndings could be extrapolated to other countries, especially developing countries, where dietary patterns vary more than developed countries.
Along with the development of economy and improvement of living conditions, dyslipidemia prevalence has shown a gradual increasing trend in China [16,17]. However, few studies have assessed dietary nutrient effects in Chinese populations. A cross-sectional study in a nationally representative sample of 46,239 adults aged at 20 years or older demonstrated that age-standardized levels of total cholesterol (TC), HDL-C, and LDL-C and TG were 4.72, 1.30, 2.68, and 1.57 mmol/L, respectively, but the effects of dietary intake on serum liquids were not evaluated [17]. In current study, we aim to examine the levels of lipids in China and evaluated the effects of three macro-nutrients (total fats, carbohydrates, protein) and three types of fats (SFA, MUFAs, PUFAs) on blood lipid measures (TC, TG, LDL-C, HDL-C, TC to HDL-C ratio, TG to HDL-C ratio) using data involving twelve administrative regions in China with 1:1 urban-torural recruitment ratio, based on an international prospective urban and rural epidemiology (PURE) study [18,19].

Study design and population selection
PURE study is a worldwide epidemiological study that recruited 153,996 adults from 628 urban and rural communities in 17 different economic-income levels countries and details of the study population and methods used in the PURE study have been published elsewhere [3,18,20]. Our analyses used baseline data from PURE-China [19,[21][22][23]. A total of 46,285 individuals were enrolled from 45 urban and 70 rural communities, of which 45,587 completed Food Frequency Questionnaires (FFQ) and had energy intake between 500 to 5000 kcal per day.

Data Collection
Individual data including socio-demographic data, self-reported medical history, lifestyle behaviors (tobacco use, alcohol consumption, and physical activities), antihypertensive and anti-diabetic medication were collected via a structured questionnaire. Physical examination was conducted to collect weight, height, and systolic/diastolic blood pressure for each participant by trained physicians. Blood samples were drawn by venipuncture after overnight fasting and processed at the eld center and shipped to the study center lab in Beijing to test serum liquids, including TC, LDL-C, HDL-C, and TG. Dyslipidemia was de ned to meet any one criteria as below: 1) TC ≥ 6.22 mmol/L; 2) LDL-C ≥ 4.14 mmol/L; 3) HDL-C < 1.04 mmol/L; 4) TG ≥ 2.26 mmol/L; 5) use of lipid-lowering medications [24].
A semi-quantitative Chinese FFQ was used to obtain food consumption with satisfactory outcomes of reproducibility and validity [25]. Total fats, carbohydrates, protein, SFA, MUFAs, and PUFAs from various foods were computed using the Chinese Food Composition Table database [25,26].

Statistical Analyses
All statistical analyses were performed with SAS 9.4 (SAS Institute Inc, Cary, North Carolina, USA). The data are presented as means ± standard deviation (SD) for continuous variables or percent for categorical variables.
Participants were categorized into quartiles of each nutrient intake (total fats, carbohydrates, protein, SFAs, MUFAs, PUFAs, and cholesterol intake), based on percentage of energy intake provided by speci c nutrients. The primary outcomes of this study are concentrations of blood lipids (TC, TG, LDL-C, HDL-C, TC to HDL-C ratio, and TG to HDL-C ratio). Multilevel linear regression with random-effect models to account for center was used to assess the effects of dietary nutrients with blood lipid concentrations. Potential confounding variables included age, sex, education levels, urban or rural locations, smoking status, and taking liquid-lowering medications, as in our PURE global paper [27]. In order to compare the strength of associations between the various nutrients and serum liquids using common units, standardized coe cients were computed, representing standard deviance (SD) change in serum liquids per one SD increase in nutrient intake.
After adjustment for covariates, we found that higher protein and fat intake was associated with higher TC (both P trend <0.001), but higher carbohydrates related with lower TC (P trend <0.001), which were illustrated in Fig. 2A-2C. Figure 2D-2F indicated that TG were associated positively with protein intake and negatively with carbohydrates (both P trend <0.001), while no signi cance with fat intake (P trend =0.140). Similar trends were observed for LDL-C and TC levels (all P trend <0.001; Fig. 2G-2I). HDL-C levels can increase along with fat intake increase, but decrease with carbohydrate intake increase (both P trend <0.001), but not related with protein intake (P trend =0.123; Fig. 2J-2L).
Regarding TC to HDL-C ratio illustrated in Fig. 2M-2O, positive and negative associations were shown for protein intake (P trend <0.001) and fat intake (P trend <0.001), respectively, but stable trend for carbohydrate intake (P trend =0.249). Figure 2P-2R presented that higher fat intake had lower TG to HDL-C ratio (P trend =0.013), was marginally associated with protein intake (P trend =0.050), and no related with carbohydrate intake (P trend =0.300). and MUFAs were more likely to have higher LDL-C levels (both P trend <0.001, Fig. 3G-3H), but stable LDL-C were observed in those with high and low PUFA intake (P trend =0.136, Fig. 3I). Nice dose-response trends of HDL-C levels were presented in all three fatty acids (P trend <0.001, Fig. 3J-3L). Negative relationship between TC to HDL-C ratio and three fatty acids, such as SFAs (P trend =0.010, Fig. 3M), MUFAs (P trend =0.003, Fig. 3N), and PUFAs (P trend =0.015, Discussions Surprisingly, high carbohydrate intake seemed to play a bene cial role on serum liquid pro le in Chinese dietary structure, though HDL-C levels also decrease slightly. The higher total fat intake and intake of two types of fats including SFAs and MUFAs, the higher most serum liquids including TC, LDL-C, and HDL-C, except for TG. However, those consuming higher PUFA intake had higher TG levels, while stable LDL-C levels were observed in various intake levels of PUFAs.  [17]. Though there are potential sampling bias and urban-rural gap, the rapid increase in serum liquids among Chinese population is inevitable. Moreover, dyslipidemia prevalence difference in urban (female 30.4% and male 38.1%) and rural residents (female 28.6% and male 30.5%) was concerned. After age-and sex-standardized using 2010 population census [30], our urban residents had the higher levels of dyslipidemia than rural residents, which consisted with the previous studies [31].
Hence, increasing burden resulted from high serum liquids levels in the Chinese general population should be obtained great attention from public health institutions and governments, especially for urban areas.
Ideal dietary regimen to reduce ASCVD risk may decrease LDL-C and raise HDL-C levels simultaneously. Previous studies have shown that higher fat intake could increase TC and LDL-C levels, but also increase HDL-C levels, thereby reduce the ratio of TC to HDL-C, including PURE-global data [27,32]. Our study had consistent results with previous studies. Additionally, low-carbohydrate diets may have more favorable changes in HDL-C and TG levels, but less favorable changes in LDL-C levels, compared with those consuming a low-fat diet [12]. Inversely, those consuming lower carbohydrate intakes were more likely to have higher TC, TG, LDL-C and HDL-C, but had no effects on TC to HDL-C ratio, and TG to HDL-C ratio in our PURE-China population. As we know, staple food is the main food in Chinese dietary, such as rice and noodle, and wheaten food, which are high in carbohydrate and low in fat.
Currently, randomized controlled trials of comparing low-carbohydrate diets to low-fat diets were conducted mainly in USA and Australia, where cooking methods might not be applicable to China. Hence, results may be much different from previous RCTs [33][34][35], if similar RCTs will be performed in Chinese population.
Saturated fats are main content in red meat, but not in white meat, which has been suggested to be associated with higher mortality of cardiovascular diseases [36]. Current dietary guidelines have largely focused on reducing saturated fatty acid intake and replacing it with carbohydrate and unsaturated fats [27,37], and therefore could reduce the risk of cardiovascular diseases [14]. In our study, as one type of unsaturated fats, MUFAs have similar impacts with SFAs on serum liquid pro le in Chinese dietary especially for TC and LDL-C, while another unsaturated fats (i.e. PUFAs) did not affect the levels of LDL-C. On the other hand, higher intake of PUFAs in daily dietary may increase TG levels, but SFAs and MUFAs not. In addition, higher intake of any types of fatty acids can raise the levels of HDL-C, which might neutralize adverse impacts of fatty acids, especially for SFAs and MUFAs. Current guidelines emphasize on the effects of diet on LDL-C, though TC to HDL-C ratio was recognized as a better predictor of cardiovascular disease risk [38]. In our PURE-China study, whatever saturated or unsaturated fatty acid intake was associated inversely with TC to HDL-C ratio and TG to HDL-C ratio. Hence, increasing intake of unsaturated fatty acids might be ineffective to improve serum liquids pro les in Chinese adults. Chinese cooking styles might be an important in uence factor. Fussy process steps and additive agents may produce trans-fats and nullify the possible bene ts of polyunsaturated fatty acids on blood lipids [27,39]. Additionally, our study observed that high level of dietary cholesterol intake was more likely to have higher TC, LDL-C level, which consist with the PURE-global and other results [27,40]. Recently, extensive research did not show evidence to support a role of dietary cholesterol in the development of CVD [41,42]. As a result, the 2015-2020 Dietary Guidelines for Americans removed the recommendations of restricting dietary cholesterol [43]. One of the most reasons is that most foods are rich in cholesterol and SFAs, which may increase the risk of CVD.
Our PURE-China study had the same limitations with global PURE [18, 44,45]. Firstly, FFQ questionnaires were specialized for Chinese people [25], while which did not differentiate between raw and cooked vegetables. As we know, Chinese special cooking pattern were much different from the Western countries, which might introduce confounding for the evaluation between nutrients and serum liquids, especially for cooked vegetables using oil and sauce. Secondly, though our PURE-China adopted cluster sampling with 1:1 urban-to-rural recruitment ratio and involved in 12 administrative regions, our population was not representative to mean levels in China. Specially, achieving good follow-up retention is the primary goal for PURE study. Thirdly, only baseline data were analyzed, which has potential methodological limitations to detect the causality between exposure and outcome. Data from follow-up period will be analyzed further to con rm the association of dietary intake and lipid pro les, especially for

Conclusions
Despite these limitations, our study found that higher carbohydrate intake was unexpected to reduce the levels of serum liquids, though HDL-C also decreased, but high protein and fat intake could raise serum liquid concentrations.
Both SFAs and MUFAs were associated positively with TC, LDL-C and HDL-C levels. However, the intake of PUFAs has not in uence on LDL-C level. Hence, enhancing intake of unsaturated fatty acids might be useless to improve serum liquids pro les in Chinese adults due to Chinese-speci c cooking styles.     carbohydrate intake, (O) TC to HDL-C ratio by fat intake, (P) TG to HDL-C ratio by protein intake, (Q) TG to HDL-C ratio by carbohydrate intake, (R) TG to HDL-C ratio by fat intake Note: TC, total cholesterol; TG, triglycerides; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol.