Scientific Papers

Analysis of different plant- and animal-based dietary patterns and their relationship with serum uric acid levels in Chinese adults | Nutrition Journal


Demographic characteristics

A total of 7,806 participants (4,180 female, 53.54% and 3,626 male, 46.45%) aged 18–94 years with average age 50 (15) years were included in the study. Among all participants, 70.52% were from rural areas, and 88.39% had an education level of high school or below. Participants were from communities with an average urban index of 66.83 (19.40). Participants’ average BMI and physical activity range were 23.36 (3.47) kg/m2 and 0–31,830.00, respectively.

The correlation of participants’ serum uric acid levels with urban index and BMI were all positive, with correlation coefficients r = 0.092 and r = 0.185 (P < 0.001), respectively. The correlation of participants’ serum uric acid levels with the physical activity was negative, with correlation coefficient r = − 0.023 (P = 0.040). Participants who were older, male, separated, from urban areas, ever smokers, and those with higher education levels had significantly higher serum uric acid levels (P < 0.01). See Table 1 for details.

Table 1 Participants’ demographic characteristics and the distribution of serum uric acid levels (n = 7806)

Latent profiles of dietary patterns

We examined latent profiles of participants’ plant-based dietary patterns. The LPA model fitting parameters are listed in Table 2. Model fit information for the five different models is listed, ranging from Profile 2 to Profile 5. In terms of the LMRT and BLRT, P-values for the Profile 2 and Profile 3 models were both < 0.05 (statistically significant). Profile 3 had lower AIC and BIC values compared with Profile 2. The entropy in Profile 3 was 0.937 > 0.800. We found that the accuracy of classification was greater than 90.00% [29]. The Profile 3 model was better than the other models in this study.

Table 2 Fit indices for Profile 2 through 5 models

The latent profiles of participants’ animal-based dietary patterns were examined. The LPA model fitting parameters are listed in Table 3. The model fit information for the five different models is listed, ranging from Profile 2 to Profile 5. In terms of the LMRT and BLRT, P-values for the Profile 2 and Profile 3 models were both < 0.05 (statistically significant). Profile 3 had lower AIC and BIC values compared with Profile 2. The entropy in Profile 3 was 0.968 > 0.800. The accuracy of classification was greater than 90.00% [28]. The Profile 3 model was better than the other models in this study.

Table 3 Fit indices for Profile 2 through 5 models

Food and nutrient intake in different latent profiles of dietary patterns

The food and nutrient intakes for each identified profile of plant-based dietary patterns are shown in Tables 4 and 5.

We found that participants with a Profile 2 dietary pattern had higher intake of cereals and cereal products, tuber starches and products, and vegetables and vegetable products than those with Profiles 1 and 3 (P < 0.001). Participants with a Profile 3 dietary pattern had higher intakes of dried legumes and legume products, and fruit and fruit products than those with Profiles 1 and 2 (P < 0.001). Participants with a Profile 1 dietary pattern had lower intakes of tuber starches and products, and vegetables and vegetable products than those with Profiles 2 and 3 (P < 0.001). See Table 4 for details.

The characteristics of the Profile 1 dietary pattern comprised the lowest intakes of tuber starches and products and of vegetables and vegetable products. The characteristics of the dietary pattern in Profile 2 showed higher intakes of cereals and cereal products, tubes starches and products, and vegetables and vegetable products. The characteristics of the dietary patterns in Profile 3 included the highest intake of dried legumes and legume products and fruit and fruit products. Based on the characteristics of these plant-based dietary pattern profiles, we denoted Profile 1–3 dietary patterns as the low tuber starches and vegetable plant-based diet (LTVP), high cereal, tuber starches, and vegetable plant-based diet (HCTVP), and high legume and fruit plant-based diet (HLFP), respectively. See Table 4 for details.

Participants with a LTVP dietary pattern had the highest intakes of vitamin A and calcium (P < 0.01). We also found that participants with an HCTVP dietary pattern had the highest intake of energy, lipids, carbohydrate, protein, dietary fiber, thiamine, riboflavin, niacin, vitamin C, vitamin E, phosphorus, potassium, sodium, magnesium, iron, zinc, selenium, copper, and manganese (P < 0.01). Participants with an HCTVP dietary pattern had higher intakes of protein, lipids, thiamine, riboflavin, niacin, vitamin C, vitamin E, phosphorus, potassium, sodium, iron, zinc, selenium, copper, and manganese than participants with an LTVP dietary pattern (P < 0.01). See Table 5 for details.

In the study, the proportion of participants who reported having LTVP, HCTVP, and HLFP was 85.0%, 8.3%, and 6.6%, respectively. There were correlations between animal-based dietary pattern and sex (χ2 = 24.690, P < 0.001), education level (χ2 = 217.866, P < 0.001), age (F = 5.631, P = 0.004), and urban index (F = 146.107, P < 0.001).

Table 4 Daily dietary food intakes in latent profiles of plant-based dietary patterns
Table 5 Daily dietary nutrient intakes in latent profiles of plant-based dietary patterns

The intakes of foods and nutrients in each identified profile of animal-based dietary patterns are shown in Tables 6 and 7.

Participants with a Profile 1 dietary pattern had higher intakes of milk and milk products and eggs and egg products than those with Profile 2 and 3 dietary patterns (P < 0.001). Participants with a Profile 3 dietary pattern had higher intakes of meat and meat products and fish shellfish, and mollusks. Participants with a Profile 2 dietary pattern had lower intakes of eggs and egg products, fish shellfish, and mollusks than those with Profiles 1 and 3 (P < 0.001). See Table 6 for details.

The characteristics of the Profile 1 dietary pattern included highest intakes of milk and milk products and eggs and egg products. The characteristics of the dietary patterns in Profile 2 included the lowest intakes of eggs and egg products and fish, shellfish, and mollusks. The characteristics of the dietary patterns in Profile 3 included the highest intakes of meat and meat products and fish, shellfish, and mollusks. Based on the characteristics of these animal-based dietary pattern profiles, we denoted the Profile 1–3 dietary patterns as the high milk and egg animal-based diet (HMiEA), low egg and fish animal-based diet (LEFA), and high meat and fish animal-based diet (HMeFA), respectively. See Table 6 for details.

Participants with an HMiEA dietary pattern had the highest intakes of energy, carbohydrate, cholesterol, vitamin A, thiamine, riboflavin, vitamin C, calcium, phosphorus, potassium, sodium, and selenium (P < 0.01). Participants with an HMeFA dietary pattern had the highest intakes of protein, lipids, niacin, vitamin E, magnesium, iron, zinc, copper, and manganese (P < 0.01). See Table 7 for details.

In the study, the proportion of participants who reported having LTVP, HCTVP, and HLFP was 4.6%, 89.8%, and 5.6%, respectively. There were correlations between an animal-based dietary pattern and sex (χ2 = 25.038, P < 0.001), education level (χ2 = 269.214, P < 0.001), age (F = 11.564, P < 0.001), and urban index (F = 168.029, P < 0.001).

Table 6 Daily dietary food intakes in latent profiles of animal-based dietary patterns
Table 7 Daily dietary nutrient intakes in latent profiles of animal-based dietary patterns

Relationship between individual serum uric acid levels and dietary patterns

There was no significant difference in participants’ serum uric acid levels according to different types of plant-based dietary patterns (F = 1.176, P > 0.05). We found a significant difference in serum uric acid levels between different types of animal-based dietary patterns (F = 32.792, P < 0.001). Participants who followed an HMeFA diet (participants’ mean serum uric acid 5.82 mg/mL) had higher serum uric acid levels than those who had an HMiEA diet (participants’ mean serum uric acid 5.12) and LEFA diet (participants’ mean serum uric acid 5.12) (P < 0.01).

Table 8 shows the association between the plant/animal-based dietary pattern and participants’ serum uric acid levels. In the unadjusted model, significant coefficients for serum uric acid levels were observed for the HEMA diet (β = 0.027, P = 0.018) and the HMeFA diet (β = 0.089, P < 0.001). Furthermore, in the adjusted model, significant coefficients for participants’ serum uric acid levels were observed for the HCTVP diet (β = −0.022, P = 0.031) and HMeFA diet (β = 0.061, P < 0.001).

Table 8 Association between dietary patterns and serum uric acid levels



Source link