The term diabetes mellitus (DM) embraces a group of metabolic diseases characterized by chronic hyperglycemia and alterations in the metabolism of carbohydrates. According to the International Diabetes Federation (IDF),1 there are 59.8 million known cases of diabetes in Europe, while the number of undiagnosed cases is reckoned to be 23 million. According to the Di@bet.es study conducted in Spain,2 it is estimated that 13.8% of the adult Spanish population suffer from diabetes, and that up to 30% present some kind of alteration in the metabolism of carbohydrates. In hospitalized patients, the prevalence of diabetes is also very high (more than 30% in patients aged over 65 years).3 DM represents a serious health problem, on account of its prevalence, morbidity-mortality, and expense.1,4

Malnutrition is frequently found in association with diabetes.5 EN (enteral nutrition) is a nutritional support that enables nutrients to be supplied directly to the digestive tube. The indications for EN are similar in patients with DM and those without it, except in the case of severe diabetic gastroparesis, where a post-pyloric approach is preferred.6 The hyperglycemia in these patients (caused by DM or metabolic stress) increases morbidity-mortality7,8 and health costs, especially when hyperglycemia is combined with malnutrition.9 HEN treatment in patients with DM seeks to achieve or maintain an adequate nutritional state through suitable metabolic glycemic and lipid control, as well as reducing morbidity–mortality and the associated health costs.6,10

Specific formulas for diabetes or stress hyperglycemia with a high monounsaturated fat content have been proposed as a good alternative for HEN patients, as they permit a short or medium-term improvement the metabolic control,11–15 while also possibly reducing the associated expense.9

Despite the clinical importance and repercussions of this topic, very few studies have evaluated the prevalence of diabetes in patients receiving EN in routine clinical practice, especially in ambulatory and home settings.8,16–18 Similarly, few studies, apart from controlled studies and clinical trials, have examined programmed nutritional and antidiabetic treatment,6,18,19 or the complications encountered, or the degree of metabolic control achieved.9

The objectives of the study were: (1) to analyze the prevalence of DM in patients receiving HEN as part of the routine clinical practice of a tertiary hospital; (2) to compare the sociodemographic characteristics of patients with and without diabetes and the data related to EN (reasons for indication and suspension, the program and formulas prescribed, and complications); (3) to describe the antidiabetic treatment prescribed and the degree of metabolic control attained in DM patients.

During the period under study, 198 patients started receiving HEN, and of these none were excluded from the analysis. The median of the follow-up was 104 days. The prevalence of DM was 31.8% (63 subjects) (Table 1). The DM had been known in 87.3% of the cases and had been unknown in 12.7%. Most of the DM was type 2 (85.7%) (Table 2).

Effect of diabetes on mortality

Out of the total sample, 132 patients (66.7%) died; 44 had DM (69.8% of the DM group) and 88 did not have DM (65.2% of the non-DM group).

The presence of diabetes did not increase the risk of death during the follow-up (after adjustment for age, gender, and diagnosis) (HR 0.90; 0.60–1.33; p=0.581) (Fig. 1). In contrast, the presence of neoplasia was associated with a higher risk of death (HR 1.57; 1.05–2.35; p=0.043).

Figure 1.

Cox regression to analyze survival according to the presence or absence of diabetes mellitus.

(0.05MB).

More than 85% of the DM patients received a specific formula for diabetes, whereas those who did not have DM mainly received standard formulas (p<0.0001) (Table 1).

The accumulated incidence of complications, whether mechanical (dislodgement of feeding tubes) or infectious, was similar in both groups (Table 1).

There are very few studies in the literature that have evaluated the prevalence of diabetes in patients receiving HEN via tube feeding, the characteristics of the nutritional support used, and the pharmacological treatment prescribed in these patients.

In our study the prevalence of diabetes was very high: over one in three (31.8%) of the patients who started HEN via tube feeding.

This prevalence is lower than that found by Arinzon et al. in elderly patients admitted to long-term care service, where 47% of the patients receiving EN had DM.17 The age of the patients (higher in the study of Arinzon et al.: 77.1 vs. 66.5 years in our series), the method used for the diagnosis of diabetes, the higher prevalence of severe neurodegenerative pathology, and various comorbid factors could explain these differences.

In contrast, Pih et al. found, in a sample of 401 patients receiving enteral nutrition by percutaneous endoscopic gastrostomy (PEG), a prevalence of DM of 22.4%. In this case, the mean age (68 years) was similar to that of our sample and the most frequent diagnosis was neurological disease. In Spain, Villar-Taibo et al. reported a similar prevalence (19.7% of patients had DM), but in this case, 69.2% of the sample used oral supplements.25 Finally, the prevalence that we observed was notably higher than that published by Vallumsetla et al. in patients with a median age of 66 years; they found a mere 3.7% of subjects with diabetes,18 although their recording method (via data collected from the history) possibly underestimated the true prevalence.

Furthermore, the data collected from our sample are similar to those of the population study Di@bet.es2 conducted in Spain, where 29.8% of the women and 42% of the men aged from 61 to 75 years suffered from diabetes.

The indications for HEN were similar in the patients with and without diabetes, and the main diagnoses were neoplasia (60%), followed by neurological disease (20%). These indications were different from those recorded in adult patients from the HEN register in Spain by the NADYA group,26 where the main diagnosis was neurological disease (59.0%). These discrepancies could be explained by the fact that our study took place in a tertiary hospital that treats patients with cancers of great complexity.

Moreover, we did not find any differences between patients with and without DM with respect to the type of tube used (gastrostomy in 54.0 and 45.9% of the patients, respectively) or in the form of administration (this was mainly intermittent, via gravity assist). These findings are similar to those published by the NADYA register,26 although they show a greater use of a gastrostomy tube than other more regional studies.25,27 Other series18 have presented a higher number of jejunostomies, probably on account of a higher prevalence of surgical pathology.

The patients with DM also had a similar rate of gastrointestinal, infectious, and mechanical complications (1 in every 5 experienced accidental slippages of the tube in both groups). Other series25 have presented a higher prevalence of gastrointestinal complications, primarily on account of mild digestive symptoms.

Poorly controlled hyperglycemia increases the risk of infection in diabetes patients, and diabetes can be an independent risk factor for long-term complications of HEN, such as infection of the wound or leaks through the tube in patients receiving PEG (percutaneous endoscopic gastrostomy).8 In our sample, however, the presence of diabetes did not lead to a higher rate of infectious complications. The effective control of glycemia in our patients could have had an effect on these results.

The mortality during the follow-up was very high in our series (over 65% in both groups), possibly because the patients who started HEN presented serious diseases that influenced their short- and medium-term prognosis. Our mortality rates are slightly higher than those of the NADYA register for 2016–2017, where 51.1% died and only 17.0% reverted to the oral route, although this difference could probably be explained by the greater length of our follow-up period.

Notable differences were evident, however, in the use of specific formulas. Most of the formulas used in the patients with diabetes were specific (85.7%), whereas they were mainly standard in the group without diabetes (88.6%). These findings contrast sharply with those of Vallumsetla et al, where no specific formulas were used,18 although they do concur with findings from Andalucía, where, back in 2007, 24% of all the specific formulas used were for diabetes or stress hyperglycemia.19

Although the use of specific enteral formulas for DM is not recommended in some guidelines, as there is considered to be a lack of sufficient evidence in their favor,28 other guidelines endorse them as a good alternative for patients receiving HEN,6,15,29 as they allow a short- to long-term improvement in the metabolic control (baseline, mean, and/or postprandial glycemia11 and, in some cases, HbA1c11 and insulin requirements13,14) in comparison with standard formulas (usually containing fiber). The benefits derived from specific formulas could also help reduce the associated health costs.9

Most of the patients with diabetes in our series received some type of antidiabetic treatment (84.1%), most often exclusively in oral form (47.6%), while 36.5% used insulin, either alone or in combination with oral treatment. Our results are somewhat different compared to those published by Vallumsetla et al., who reported 39.7% of their patients using oral antidiabetics exclusively and 34.5% using insulin. These discrepancies can be explained by the fact that their patients had higher HbA1c readings than those in our series.18

Metformin was the most widely used oral antidiabetic (54.0%). There are no specific guides recommending which antidiabetics, apart from insulin, should be used in patients receiving HEN, although the iSGLT2 and aGLP1 do not seem indicated, due to their shared effect on weight, as well as the risk of urinary infections in the former case and gastrointestinal complications in the latter. Moreover, the iDPP4 are not recommended, as insufficient data are available on their absorption after being crushed for tube feeding.

The evolution of the anthropometric and analytical parameters during the follow-up of the DM patients showed no significant differences as regards the BMI or HbA1c, although there was a significant reduction in venous glycemia (153.5±77.5mg/dL at the start vs. 119.7±48.6mg/dL at the end; p=0.015) (Table 3). In contrast, Vallumsetla et al. observed a significant reduction of HbA1c in their series, probably because they started with higher initial readings.18

Our study evaluates the prevalence of DM in a large sample of patients receiving HEN who were followed up in standard clinical practice. The study also analyzes the type of nutritional treatment prescribed, the complications, and the metabolic control, and as such it provides data on a topic that has scarcely been investigated in the literature. Our study is not exempt from limitations, however: on the one hand, it is a single-center study, and it is therefore possible that its results cannot be extrapolated to other centers with different characteristics; on the other hand, as the study was retrospective and based on digital clinical histories, some data from the metabolic follow-up could not be recorded for all the patients in our sample.

In conclusion, there was a high prevalence of DM in the patients receiving HEN via tube feeding in our sample, and most of these DM patients were treated with specific nutritional formulas. However, the presence of diabetes was not associated with greater morbidity-mortality or any differences in the schedules or indications for HEN.

G. Olveira, J. Abuín and V. Doulatram equally contributed to the conception and design of the research; F. Sánchez-Torralvo and V. Contreras contributed to the design of the research; S. Padín and M. Gonzalo contributed to the acquisition and analysis of the data; J. Abuín and V. Doulatram contributed to the interpretation of the data; and J. Abuín and G. Olveira drafted the manuscript. All authors critically revised the manuscript, agree to be fully accountable for ensuring the integrity and accuracy of the work, and read and approved the final manuscript.

None declared.

G. Olveira has had occasional interventions of consultancy or speaker fees from Abbott, Nutricia, Vegenat, Fresenius, Nestle, Persan Farma and Adventia.