Introduction

The prevalence of treated end-stage renal disease (ESRD) has steadily increased from 2001 to 2014 in all countries, and become a burden to every nation and healthcare system [1]. In 2014, the prevalence of ESRD patients undergoing dialysis in Taiwan was 3093 patients per million population, about 90% of them receiving in-center hemodialysis treatment [1]. It was summarized that nutritional factor was implicated as a risk factor for the development of metabolic in chronic kidney disease, especially in ESRD patients [2].

Nutritional therapy is recognized as an effective approach to prevent metabolic abnormalities and unfavorable outcomes in people with chronic conditions [3,4,5,6,7,8]. Increased dietary energy intake is mentioned in the National Kidney Foundation-Kidney Disease Outcomes Quality Initiative (K/DOQI) guidelines [9]. It is recommended that consuming enough energy daily guarantees the nitrogen balance and prevents protein catabolism and tissue destruction, which could optimize the nutritional status and hemodialysis outcomes [9]. However, the daily intake of macro-nutrients and micro-nutrients are largely inadequate in hemodialysis patients [10]. More than a half of hemodialysis patients had problems to follow the healthy diet guidelines (related to energy and nutrients intakes) which related behaviors, technical difficulties, physical conditions, time, and food preparation [11]. Inadequate dietary intake is also a possible result of a significant lifestyle change while receiving dialysis treatment. On the other hand, adherence to a complicated and restrictive dietary intake further exacerbates nutrient deficits in this group of patients [9, 12,13,14].

The prevalence of metabolic syndrome was high in the ESRD patients undergoing hemodialysis [15]. The MetS has been implicated as a risk factor for the development of diabetes, cardiovascular disease, cancer, and all-cause mortality [16,17,18,19]. The prevalence of metabolic syndrome varied by different assessment criteria, e.g. 51%, 66.3%, and 75.3% according to National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III), International Diabetes Federation (IDF), and Harmonizing the Metabolic Syndrome (HMetS) criteria, respectively [20]. This indicated that there was not yet a single definition that could reflect the real spectrum of the epidemiology of MetS. Therefore, in the current study, two definitions were used with different focuses to assess the MetS: The American Association of Clinical Endocrinologists (AACE) definition, focused on hyperglycosemia, was glucocentric [21]; and Harmonizing Metabolic Syndrome definition was agreed by Joint statement from the IDF, American Heart Association (AHA) and the National Heart, Lung, and Blood Institute (NHLBI), the World Heart Federation, the International Atherosclerosis Society, and the International Association for the Study of Obesity, which relayed on collection of abdominal obesity and related CVD risk factors [22].

There were few studies investigated dietary intake among hemodialysis patients. One study compared the dietary intake status between 54 HD patients, and 47 non-HD patients, and between dialysis day and non-dialysis day among elderly people in Brazil [23]. The other study in the United States only examined the association between dietary energy intake and body composition changes in 13 HD patients [24]. In addition, the dietary approach was found as an effective therapy to decrease most of the risks for MetS in a randomized controlled trial [25]. However, hemodialysis patients were with high metabolic syndrome prevalence, and generally have difficulties achieving recommended energy intakes. In our knowledge, the role of dietary energy intake on metabolic disorders among hemodialysis patients remains to be investigated.

This study was to examine the association of inadequate dietary energy intake with metabolic abnormalities and metabolic syndrome among patients who receiving hemodialysis treatment from seven hemodialysis centers. It was hypothesized that hemodialysis patients with reported inadequate dietary energy intake (IDEI) more likely had metabolic abnormalities or metabolic syndrome.

Methods

Study design and setting

A cross-sectional study design was carried out from September 2013 to April 2017. We collected data from 492 patients from hemodialysis centers in seven hospitals. The study sample consisted 165 from Taipei Medical University Hospital, 91 from Taipei Medical University – Wan Fang Hospital, 39 from Taipei Medical University – Shuang Ho Hospital, 41 from Cathay General Hospital, 57 from Taipei Tzu-Chi Hospital, 49 from Wei-Gong Memorial Hospital, and 50 from Lotung Poh-Ai Hospital.

Sample size

The sample size in a cross-sectional design is calculated using the formula: \( n=\frac{Z^2P\left(1-P\right)}{d^2} \) Where n (sample size), Z (level of confidence), P (expected prevalence), and d (precision, corresponding to effect size) [26]. The sample of 92 was calculated with Z = 1.96 for type I error of 5%, P = 0.745 as the prevalence of MetS was 74.5% in hemodialysis patients [27], and d = 0.1 as suggested for a medical study [28]. In the current study, the final sample of 228 patients is adequate for analysis and depicted in Fig. 1.

Fig. 1
figure 1

Flow chart of patients sampling and study procedure. ESRD: End-stage renal disease

Full size image

Patient recruitment criteria

The study patients in the current study fulfilled the recruitment criteria as mentioned elsewhere [29,30,31].

Data collection procedure

The physicians and nurses in each hospital screened for qualified patients who satisfied the recruitment criteria. The interviewers (Registered Dietitians) then contacted the eligible patients and asked for their voluntary participation.

The eligible patients signed the informed consent form before participating in the face-to-face or telephone interviews which conducted by registered dietitians (three-day dietary intake, physical activity). The medical charts were reviewed after the interviews. Anthropometric, and energy expenditure values were also measured. Licensed nurses collected blood samples at the first dialysis session during the study week, biochemical data was then analyzed using available laboratory test kits, the procedure was described in details elsewhere [32].

Measures

Patients’ characteristics

The information regarding age, gender, hemodialysis vintage, comorbidities calculated using the Charlson comorbidity index [33], history of hypertension, cardiovascular diseases, and type 2 diabetes mellitus (T2DM), body mass index, BMI (kg/m2), pre-dialysis systolic (SBP) and diastolic (DBP) blood pressure were also assessed using medical records. The waist circumference (WC), body fat mass (FM) were assessed using the bio-electrical impedance analysis device (InBody S10, Biospace, Seoul, Korea), the detailed procedure was described elsewhere [34]. Elevated body fat mass was defined as FM ≥ 25% for men, FM ≥ 30% for women, respectively [35]. Interdialytic weight gains (IDWG) was also calculated. Higher IDWG linked with higher BP in hemodialysis patients [36].

Physical activity

The short version of the International Physical Activity Questionnaire was used to evaluate physical activity level. Patients were asked about their time spent (days per week, and minutes per day) on different levels of physical intensity (vigorous, moderate, walking, and sitting), questionnaire took 4 to 15 min to complete [37]. The overall physical activity score was calculated as the sum of minutes spent on activities at different levels of vigorous, moderate, walking, and sitting over last seven days multiplied by 8.0, 4.0, and 3.3, 1.0, respectively [38]. The common method using metabolic equivalent task scored in minute per week (named as MET- min/wk) was used to represent the physical activity [39].

Dietary energy intake

We used three-day dietary intake record to assess patient’s intake, and confirmed data by a 24-h dietary recall, the details were mentioned elsewhere [32, 40]. In brief, the information related to names of food, brand, ingredients, cooking methods, portion or weight, meal location and time were collected. The e-Kitchen software, a nutrient analysis software (Nutritionist Edition, Enhancement plus 3, version 2009, Taichung, Taiwan) was used for analyzing nutrients.

The recommended daily dietary energy intake was ≥35 kcal/kg for patients younger than 60 years old, and ≥ 30 kcal/kg for those who 60 years old or older, respectively [9]. Inadequate dietary energy intake was defined as patients consumed less than the recommended levels. In order to enhance the reliability of measures and analysis, the under-reported dietary energy intake (EI) data were excluded from the final analysis if the ratio of EI:REE < 1.27 [41]. The results of the analysis were not affected by excluding the under-reporters in the study [42]. The resting energy expenditure (REE) was assessed using a hand-held indirect calorimeter, named MedGem (Microlife USA, Dunedin, FL). A modified Weir equation together with a fixed respiratory exchange ratio of 0.85 were used to estimate carbon dioxide production. Patients wore a nose clip and a mouthpiece, then breathe normally for about 7–10 min, or until the volume of oxygen is stable. The MedGem has been validated against several metabolic calorimeters such as Douglas Bag method [43], and metabolic cart systems [44, 45]. This device has the similar accuracy of commonly used prediction equations such as the WHO/FAU/UNU, Mifflin, or Harris–Benedict equations [46], and used in hemodialysis patients [47].

The biochemical values

Fasting blood glucose (FPG), fasting plasma insulin (FPI), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), high sensitive C-reactive protein (hs-CRP), Creatinine, Albumin, intact parathyroid hormone (iPTH), the normalized protein nitrogen appearance (nPNA) was estimated using the formula: nPNA = Pre-BUN/[25.8 + 1.15*(eKt/V) + 56.4/(eKt/V)] + 0.168, where pre-BUN is pre-dialysis blood urea nitrogen (mg/dL), post-BUN is post-dialysis blood urea nitrogen (mg/dl), and equilibrated Kt/V is dialysis quality [48].

Diagnosis of metabolic syndrome (MetS)

The MetS was classified by American Association of Clinical Endocrinologists (AACE), hereafter referred as AACE-MetS [21]. Patients were identified as MetS if they had (1) and any of the criteria (2), or (3), or (4). (1) Impaired fasting glucose (IFG) which patients had FPG ≥ 100 mg/dL, or previously diagnosed T2DM [49]. (2) Overweight or obese (BMI ≥ 24.0 kg/m2 for Taiwanese) [50]. (3) TG ≥150 mg/dL, HDL-C < 40 mg/dL for men or HDL-C < 50 mg/dL for women. (4) SBP ≥ 130 mmHg or DBP ≥ 85 mmHg.

To affirm the non-spurious association, the Harmonizing Metabolic Syndrome definition (HMetS) was also used to evaluate MetS. Patients were classified as MetS if they have three or more abnormalities (WC ≥ 90 cm for men, WC ≥ 80 cm for women, TG ≥150 mg/dL, low HDL-C, high BP, or IFG) [22].

Other biochemical value classifications

The lipid profile (LDL-C ≥ 100 mg/dL, and TC ≥ 200 mg/dL) [51], inflammation maker (high sensitive-CRP > 0.5 mg/dL) [52], elevated insulin (FPI ≥ 12 mU/L) [53, 54], iPTH ≥300 pg/mL [55]. In addition, the poor nutritional status including nPNA < 1.0 g/kg, serum albumin (Alb) ≤ 3.5 mg/dL, and serum creatinine (Cr) ≤ 7.5 mg/dL [56].

Statistical analysis

The study sample was described using mean ± standard deviation (SD), or median (interquartile range), or frequency (percentage). The continuous variables were tested for normality by using a Shapiro-Wilk’s test [57, 58], and histograms, box plots, and normal Q-Q plots were examined. The ANOVA, Mann-Whitney U test, or Chi-Square test were recruited in order to compare characteristics and metabolic parameters of the adequate and inadequate EI groups. The bivariate logistic regression models were recruited for examining associations of patients’ characteristics, dietary intake with metabolic abnormalities and MetS. The multivariate logistic regression analyses were then utilized for examining the association of inadequate dietary intake of nutrients with metabolic abnormalities and MetS. The sub-group analyses were performed in different groups of diabetes mellitus, hypertension, and cardiovascular diseases. Patients’ gender, age, physical activity, hemodialysis vintage, Charlson comorbidity index (CCI), hs-CRP, and IDWG were controlled in the multivariate analyses as they showed the associations with metabolic syndrome [59,60,61,62,63]. The analyses were performed for both diagnosed criteria of MetS (AACE-MetS and H-MetS) to affirm the non-spurious association. The IBM SPSS software version 20.0 for Windows (IBM Corp., New York, USA) was used for all analyses. The statistically significant level was set at P value < 0.05.

Results

The mean ± SD of age, hemodialysis vintage, physical activity, CCI, and interdialytic weight gains were 59.4 ± 11.3, 5.5 ± 5.0, 4831.3 ± 1893.1, 4.6 ± 1.5, and 3.0 ± 1.7, respectively. Of study sample, there were 64.9% men, 38.2% diabetes, 48.2% hypertension, and 29.8% cardiovascular diseases, 28.5% with an elevated level of hs-CRP, 54.5% elevated body fat mass. The REE was lower in patients with inadequate EI (1014.5 ± 280.4) than those with adequate EI (1100.9 ± 274.7), with p = 0.023. Regarding metabolic abnormalities, the prevalence of IFG, overweight or obese, elevated WC, high BP, high TG, and low HDL-C were 64.9%, 36.4%, 26.3%, 81.6%, 39.0%, and 61.0%, respectively. The prevalence of metabolic syndrome was 63.2% as diagnosed by AACE criteria, and 53.9% as diagnosed by HMetS criteria. The prevalence of the metabolic abnormalities (not hypertension) and syndromes were statistically significantly higher in hemodialysis patients with inadequate EI than those who with adequate EI (Table 1). Out of patients, 60.5% reported less than the recommendation level of dietary energy intake. Patients with inadequate EI more likely consumed inadequate protein and fat, but consumed less mineral, water, and vitamin than those with adequate EI (Table 2).

Table 1 Characteristics, and metabolic parameters, and other biochemical values in hemodialysis patientsa
Full size table
Table 2 Dietary intake among hemodialysis patientsa
Full size table

The results of bivariate logistic regression analyses presented that higher age associated with higher prevalence of IFG and AACE-MetS with odd ratio, OR = 1.03, 95% confidence interval, 95%CI, 1.00–1.05, p <  0.05, and OR = 1.03, 95%CI, 1.01–1.06, p <  0.05, respectively. Men experienced higher prevalence of overweight or obesity (OR = 1.85, 95%CI, 1.03–3.33, p <  0.05), but lower prevalence of elevated waist circumference (OR = 0.32, 95%CI, 0.17–0.59, p <  0.001) than women. Hemodialysis vintage was negatively associated with IFG (OR = 0.91, 95%CI, 0.86–0.97, p <  0.001), Overweight/obese (OR = 0.92, 95%CI, 0.86–0.98, p <  0.05), high TG (OR = 0.94, 95%CI, 0.89–0.99, p <  0.05), low HDL-C (OR = 0.95, 95%CI, 0.90–0.99, p <  0.05), AACE-MetS (OR = 0.90, 95%CI, 0.85–0.96, p <  0.001), and HMetS (OR = 0.94, 95%CI, 0.89–0.99, p <  0.05), respectively. Charlson comorbidity index was positively associated with IFG (OR = 1.38, 95%CI, 1.14–1.67, p <  0.001), Overweight/obese (OR = 1.21, 95%CI, 1.01–1.45, p <  0.05), AACE-MetS (OR = 1.44, 95%CI, 1.19–1.74, p <  0.001), and HMetS (OR = 1.28, 95%CI, 1.07–1.53, p <  0.01), respectively. Interdialytic weight gains was positively linked with IFG (OR = 1.21, 95%CI, 1.03–1.43, p <  0.05), AACE-MetS (OR = 1.22, 95%CI, 1.04–1.43, p <  0.05), and HMetS (OR = 1.24, 95%CI, 1.06–1.46, p <  0.01), respectively (Table 3).

Table 3 Bivariate analysis the effects of personal factors and dietary intake on metabolic abnormalities and metabolic syndrome
Full size table

Reported inadequate dietary energy intake associated with 1.83–6.20 folds of metabolic abnormalities or metabolic syndrome. It was significantly linked to higher prevalence of IFG (OR = 2.50, 95%CI, 1.43–4.37, p <  0.001), overweight/obese (OR = 6.10, 95%CI, 3.10–11.99, p <  0.001), elevated waist circumference (OR = 6.20, 2.78–13.84, p <  0.001), high triglyceride (OR = 1.90, 95%CI, 1.09–3.34, p <  0.05), low HDL-C (OR = 1.83, 95%CI, 1.06–3.15, p <  0.05), AACE -MetS (OR = 2.34, 95%CI, 1.35–4.06, p <  0.01), and HMetS (OR = 3.24, 95%CI, 1.86–5.63, p <  0.001), respectively. The sodium and fluid intake were not associated with metabolic abnormalities or MetS (Table 3).

The associations of inadequate energy intake with metabolic abnormalities, AACE-MetS, and HMetS were stronger by 2.26 to 8.17 folds after adjusted for gender, age, physical activity, hemodialysis vintage, Charlson comorbidity index (CCI), hs-CRP, and IDWG in multivariate analyses. Inadequate energy intake did not show the significant association with high TG, low HDL-C or high blood pressure (Table 4). On the other hand, the consumption of MUFA greater or equal to 20% of EI is associated with higher likelihood of having IFG (OR = 2.85, 95%CI, 1.39–5.87, p <  0.01), and AACE-MetS (OR = 3.01, 95%CI, 1.45–6.26, p <  0.01, Table 4).

Table 4 Associations of dietary intake and metabolic abnormalities and metabolic syndrome via multivariate logistic regression analyses a
Full size table

In sub-group analyses, inadequate EI showed an significant association with higher prevalence of AACE-MetS in non-hypertension group (OR = 4.09, 95%CI, 1.55–10.77, p = 0.004), and non-cardiovascular disease group (OR = 2.59, 95%CI, 1.23–5.42, p = 0.012); and associated with HMetS in group of diabetes (OR = 8.33, 95%CI, 2.08–33.37, p = 0.003), non-hypertension (OR = 5.33, 95%CI, 1.97–14.40, p = 0.001), hypertension (OR = 2.59, 95%CI, 1.05–6.37, p = 0.038), and non-CVD (OR = 3.79, 95%CI, 1.80–7.97, p < 0.001, Table 5).

Table 5 Association between inadequate energy intake and metabolic syndrome in subgroups of medical historya
Full size table

Discussion

In the present study, results elucidated that reported inadequate dietary energy intake (IDEI) associated with more MetS abnormalities, and a higher proportion of MetS. The reported IDEI strongly determined 2.26 to 8.17 folds of metabolic abnormalities and MetS diagnosed either by AACE or HMetS criteria. In hemodialysis patients, IDEI disrupts the energy balance, and the nitrogen balance, increases the tissue destruction, and protein catabolism which cause the MetS and exacerbate the dialysis outcomes [64]. On the other hand, the MetS was found to be a high-risk for many chronic health problems such as obesity, T2DM, cardiovascular diseases, cancer, and all-cause of death [16,17,18,19]. Therefore, the early MetS identification and nutritional therapy were highly recommended to reduce above adverse health problems [25, 65]. In addition, patients who consumed adequate energy-rich-protein can improve the balance of body protein, body composition which further improve hemodialysis outcomes [66].

The current study showed that about 60% of hemodialysis patients consumed low dietary energy intake. This was in the line with a reliable previous publication in which patients had at most 75% of the energy and protein intake as recommended by K/DOQI guidelines [9]. MetS prevalence was high in the present study (63.2% AACE-MetS, 53.9% HMetS), and in previous studies in Southern Taiwan was 61.0% measured by NCEP-ATP III criteria [15]. In comparison with previous studies, the prevalence of MetS was lower in the current study than that in a study in Brazil (74.5%) using the HMetS criteria [27], and in United States (69.3%) using the NCEP-ATP III criteria [67].

The consumption of PUFA and SFA did not show the significant association with MetS and its abnormalities, while the consumption of MUFA equal or greater than 20% demonstrated the association with higher IFG and AACE-MetS in the current study. In a previous study, no association between PUFA, or SFA, and MetS was found [68]. Inconsistently, a number of previous studies suggested that the consumption of dietary MUFA improves insulin sensitivity. In addition, MUFA intake as a substitution for SFA demonstrated the benefit for reducing the metabolic syndrome [69, 70]. The discrepancy between the findings of this study and other studies could be explained by the cross-sectional design of the current study, the causal relationship is not generated. In addition, the 24-h dietary recall is subject to reporting bias from patients. In practical application, the MUFA was with high density in the Mediterranean dietary pattern (MDP). Strong evidence from several studies and trials proved that the MDP was inversely associated with the incidence of MetS, cardiovascular diseases [71,72,73,74]. Therefore, this MDP can be still encouraged and adopted in various population and cultures, with cost-effective serve for preventing the MetS and its components [75]. However, the application is with precaution and more studies are suggested to intensively investigate the MDP effect on the MetS in hemodialysis patients.

The current results illustrated that the inadequate dietary EI was associated with high prevalence of HMetS in different sub-groups. In a study conducted in Italy, the authors found that patients with MetS reported lower energy intake than those without MetS [76]. This suggested that MetS diagnosed by Harmonizing Metabolic Syndrome criteria is more sensitive than AACE-MetS in relation to energy intake. In practice, in order to improve the hemodialysis outcomes, the adequate dietary EI is recommended by the K/DOQI guidelines which can reduce the risks of MetS [9].

The present study demonstrated that the higher prevalence of IFG and AACE-MetS was observed in older patients. The association was also found in previous studies on the general population in Norway, which MetS was diagnosed by either NCEP- ATP III, or IDF criteria [60], and in individuals in the United States [61]. This emphasized that the old people are more likely to have metabolic abnormalities, risks for CVD, and type 2 diabetes. Therefore, the MetS definitions should be specifically classified for elderly people, as in need of comprehensive assessment for risk factors. On the other hand, men more likely experienced overweight/obesity, but less likely had elevated waist circumference in comparison with women. This could be explained that men have greater abdominal visceral adipose tissue (likely corresponding to BMI), but less abdominal subcutaneous adipose tissue (likely corresponding to waist circumference) than women [77].

The study conducted on 153 hemodialysis patients in three dialysis centers in Tehran demonstrated that the prevalence of MetS among women was higher than that among men [62]. However, in the current study, gender was significantly associated with metabolic abnormalities, but not with AACE MetS or HMetS. This suggests that gender should take into consideration when assessing or treating patients with MetS, the presence of metabolic disorders in men or women may depend on their specific lifestyles and behaviors.

The longer hemodialysis vintage has shown the protective impact on MetS among studied hemodialysis patients. This somehow expressed the quality of hemodialysis in dialysis centers in Taiwan which reflected the effectiveness of multi-disciplinary care program in hospitals since 2003 to combat chronic kidney disease and related comorbidities [78]. In addition, the full reimbursement of dialysis costs by National Health Insurance in Taiwan medical system could further optimize the quality of care [79], which in turn reduced the prevalence of metabolic disorders in this study.

Physical activity was not associated with metabolic abnormalities or MetS in the present study. However, a review of several randomized trials concluded that the physical activity decreased the likelihood of development of MetS; if there were no contraindications, more intensive physical exercise or resistance training should be considered to prevent and treat MetS [63]. In addition, patients who performed regular exercise had better dialysis outcomes and health benefits as reported in an international study on hemodialysis patients [80].

Finally, the elevated level of hs-CRP did not show the association with MetS and its components. Inconsistently, the association was existed in the previous study, that inflammatory biomarkers had a correlation with MetS in hemodialysis patients [62]. An elevated level of hs-CRP may be a key independent predictor of adverse outcomes in hemodialysis patients with MetS. Therefore, reducing serum hs-CRP level should be considered for preventing MetS, CVD, and finally mortality among hemodialysis patients.

There was some limitations in the current study. Firstly, the causality cannot be proved between dietary EI and metabolic abnormalities and MetS in a cross-sectional design. In addition, the application of adequate EI but less MUFA intake was not clearly addressed because of the nature of the cross-sectional study design, and unavoidable reporting bias. More in-depth longitudinal studies and trials are required. The self-reported dietary assessment using food records and recalls had impacts on energy underreporting, appropriate interpretations of the results are recommended [81]. In the current study, we excluded those patients underreported their energy intake in order to avoid the bias and improve the reliability of findings [42]. However, the sample size is relatively small for subgroup analysis. Further investigation should be conducted on larger sample, to enhance the reliability of finding. The present study demonstrated a number of strengths that patients’ body composition was measured precisely and directly using the BIA, while biochemical data were assessed by using standard laboratory tests. Two MetS definitions reflecting the glucocentric, obesity, and CVD risk factors were used to assure the non-spuriousness of the relationships. Future longitudinal studies or trials were recommended to confirm the relationship between dietary intake and MetS and impacts of nutritional interventions on dialysis outcomes.

Conclusions

This was the first study exploring the association of the reported dietary EI with metabolic abnormalities and MetS diagnosed by AACE and Harmonizing Metabolic Syndrome criteria in hemodialysis patients. We found that inadequate EI was high prevalence and associated with up to 2.26–8.17 folds of MetS and its components. Promoting adequate dietary energy intake following the K/DOQI guidelines could help to improve dialysis quality, prevent MetS, minimize the negative effects of metabolic disorders and their consequences, in turn, optimize the quality of care, and improve the quality of life of HD patients. Future studies are suggested for carefully exploring the mechanism, and evaluating the effect of dietary energy interventions.