Adherence to the Dietary Approaches to Stop Hypertension Diet and Non-alcoholic Fatty Liver Disease

Yuanyuan Sun; Shuohua Chen; Xinyu Zhao; Yanhong Wang; Yanqi Lan; Xiaozhong Jiang; Xiang Gao; Shouling Wu; Li Wang


Liver International. 2022;42(4):809-819. 

In This Article

Materials and Methods

Study Population

Our study was based on the Kailuan I and II, two large prospective cohorts designed to investigate risk factors for common non-communicable diseases, as detailed elsewhere.[9,10] The Kailuan I study comprised 101 510 participants (aged 18–98 years) from Tangshan City in China at the time of enrollment between June 2006 and October 2007. The Kailuan II study was established between June 2008 and October 2010, when 35 856 Kailuan residents (aged ≥18 years) who were not part of the Kailuan I study were enrolled to complete the first survey. All participants underwent a comprehensive survey, including a questionnaire (eg demographic characteristics, medical comorbidities, medication use and lifestyle factors), physical examination and laboratory tests, which was updated during follow-up every 2 years.[9,10]

To explore the association between DASH adherence and NAFLD, we included a total of 97 413 participants who received the abdominal ultrasound examination in 2014 from the above two cohorts because the dietary intake was collected using food frequency questionnaires (FFQ) in 2014. We excluded participants who reported to have chronic hepatitis C infection history, with hepatitis B surface antigen-positive or with hepatitis B/C virus infection status unavailable (n = 26 274), having alcohol consumption[11] or with missing alcohol information (n = 35 419), with cardiovascular disease history who may have changed their dietary status (n = 2249), having implausible dietary data (extremely high or low energy intake: <800 or >4200 kcal/day for men and <500 or >3500 kcal/day for women) or with more than half of the FFQ items missing (n = 5677). We also excluded the participants who were diagnosed with fatty liver (n = 15 454) or had no ultrasound examination before 2014 (n = 452) to ascertain non-NAFLD or incident NAFLD. Finally, 11 888 participants were used for analysis (Figure 1). NAFLD was diagnosed by abdominal ultrasonography (HD-15; Philips, Netherlands), which was based on the presence of at least two of the following abnormal ultrasonographic liver features, including diffusely increased echogenicity of liver relative to kidney, ultrasound beam attenuation and poor visualization of intrahepatic structures. Abdominal ultrasonography was routinely performed by experienced radiologists, who were blinded to both clinical presentation and laboratory findings.

Figure 1.

Flowchart of the study participants

The study was approved by the Ethics Committee of Kailuan General Hospital and Institute of Basic Medical Sciences Chinese Academy of Medical Sciences. Written informed consent was obtained from all participants.

Assessment of Dietary Exposure

Self-reported semi-quantitative FFQ with 33 food items and seven condiments, which was validated in the Chinese population and used for the Chinese national nutrition survey in 2002,[12] was used to assess dietary exposure. For each food item, participants were asked how frequently (never, daily, weekly or monthly) they consumed, followed by a question of the consumption amount each time in the unit of liangs (50 g/liang). The average daily intake of each food item was calculated by multiplying the intake frequency of each food per day by the amount consumed at each time. For condiments, the total family consumption (g) per month and the total number of family members eating at home were asked. The average daily intake for each condiment was calculated by total family consumption divided by total family members and 30 days. Nutrients from each food item were calculated based on the Chinese Food Composition Table by multiplying the nutrient density (nutrient per 100 g) for each food item by an average amount of each food item per day.[13]

We constructed the DASH score based on food and nutrients emphasized or minimized in the DASH diet, focusing on nine components: vegetables(refresh vegetables), fruit (refresh fruit and strawberry/blueberry), dairy (milk, milk powder and yogurt), beans (Tofu, Tofu skin, soybean and dried beans), whole grains (millet/sorghum/corn), meat (pork, beef/mutton, poultry, visceral and aquatic product), fat (vegetable oil and animal oil), sodium (salt) and beverage (juice and other sweetened beverages).[14,15] First, the quantified average daily intake for each of the nine components was energy-adjusted using the residual method.[16] Next, all the component scores were calculated separately for men and women. Then, participants were classified into quintiles by sex according to the ranking of their energy-adjusted average daily intakes.[17] We assigned the lowest quintile one point and the highest quintile five points for vegetables, fruits, dairy, beans and whole grains by quintile ranking. The lowest quintile was given five points and the highest quintile one point for sodium, meats, fat, sodium and beverage by the reversed quintile ranking. We then summed up the component scores to obtain an overall DASH score ranging from 9 (worst) to 45 (best) to evaluate the participant's adherence to the DASH diet.

Assessment of Covariates

Demographic characteristics, physical activity, alcohol drinking and smoking status were collected via questionnaires in 2014. Physical activity was categorized into low, moderate and vigorous levels according to the tertiles of the total energy expenditure (MET-minutes/week), which was calculated according to the Guidelines for Data Processing and Analysis of the International Physical Activity Questionnaire.[18] Body mass index (BMI) was calculated as weight (kg) divided by square of height (m2) and classified as <23 and ≥23 kg/m2 according to World Health Organization's recommendation for Asian populations.[19] Systolic and diastolic blood pressures (SBP and DBP) were measured, and high blood pressure was defined by SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg according to the guideline of high blood pressure in China.[20] Fasting blood samples were collected, and serum fasting blood glucose (FBG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), triglyceride (TG), C reactive protein (CRP) and alanine aminotransferase (ALT) were tested using an autoanalyser (Hitachi 747; Hitachi, Tokyo, Japan) at the central laboratory of the Kailuan General Hospital. FBG was classified as <6.1 and ≥6.1 mmol/L according to the World Health Organization's definition of impaired fasting glycaemia.[21] TG (<1.7 and ≥1.7 mmol/L), HDL-c (<1.0 and ≥1.0 mmol/L) and LDL-c (<3.4 and ≥3.4 mmol/L) were classified according to the guidelines for the prevention and control of dyslipidemia in China.[22] CRP was classified as ≤2 and >2 mg/dL according to the new expert consensus of metabolic risk factors for NAFLD.[23] ALT was classified as <40 or ≥40 U/L according to the guidelines of NAFLD in China.[24]

Statistical Analysis

The participants' characteristics were described as mean ± standard deviation (SD) for continuous variables and number (percentages) for categorical variables. We used logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs) to test the differences in the risk of NAFLD across DASH score quintiles using the lowest quintiles of the score as the reference. We fit three multivariate models: model 1 adjusted for age, sex and energy; model 2 further adjusted for BMI, physical activity, education, current smoking status; model 3 further adjusted for FBG, TG, SBP, HDL-c, LDL-c, CRP and ALT. Also, we explored the association of each component of the DASH diet with NAFLD with adjustment for confounders in model 3. Stratified analysis was used to test the effect of BMI, blood pressure, FBG, LDL-c, HDL-c, CRP and ALT on the association between DASH adherence and NAFLD.

To address the great difference in the water contents, we calculated the total intake of the dry weight of foods for each of the DASH components by multiplying the amount of each food consumed by 1 − percentage water content of the food item and then summing over all relevant foods.[25] Then, we conducted a sensitivity analysis to estimate the association between DASH adherence calculated based on dry weight and NAFLD.

Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA). Two-tailed tests were used in all statistical analyses, and P values below 0.05 were considered statistically significant.