Enteral Nutrition in Anorexia Nervosa GlOVANNI BUFANO, M.D., CARLOTTA BELLINI, M.D., GIANFRANCO CERVELLIN, M.D., CARLO COSCELLI, M.D. From the
Department of Medicine
C.
Cattani,
I Division
AND
of Medicine, Ospedale Maggiore, Parma, Italy
ABSTRACT. Biochemical and anthropometric data were collected in 9 subjects with anorexia nervosa before and after enteral nutrition by the nasogastric route. All subjects but one accepted the treatment which was stopped as soon as an adequate spontaneous food intake was resumed. Enteral nutrition caused a significant increase of mid-arm muscle circumference and of tricipital skinfold. Body weight gain was remarkable, averaging a rate of 8.22 kg/month. This value was not different from the levels attained by anorectic subjects put on total
parenteral nutrition. Furthermore, a significant rise of some biochemical indexes, namely prealbumin and total iron binding capacity, was achieved. The most impressive drawbacks of enteral nutrition were occasional hypophosphatemia and mod-
The great variety of therapeutic efforts reported in the medical literature on anorexia nervosa (AN) reflects the present lack of knowledge of the pathogenesis of the disease, which seems to be increased in frequency’ and severity’ during the last decades. The poor response to treatment is emphasized by some recent research4.5 showing persistent personal difficulties and anomalous feeding behavior in spite of a resumption of normal body weight and regular menses. The same authors state that about 25% of these patients continue in amenorrhoea in spite of regained original weight, and one-third of the patients do not regain the original weight at all. A critical point, which seems to have been poorly investigated, regards the role of artificial nutrition in the treatment of AN. This approach aims not only at facing actual malnutrition of the anorectic patients, but also breaks off the vicious circle resulting from the manutrition itself, which could enforce more and more the anorectic behavior.’ In 32 cases7-10 total parenteral nutrition (TPN) was employed. The patients responded to the treatment with a very high rate of weight increase (about 10 kg/month), but they developed many complications (46%), half of which were clinically relevant and one, fatal. There are very few reports about the use of enteral nutrition (EN) in AN. This kind of nutritional support was administered to 3 patients by the intrajejunal route, with an energetic supply ranging from 2000 to 2400 Kcal/day.l A mean weight increase rate of 5.9 kg/month was obtained, the only complication reported being the occurrence of abdominal pain and diarrhea in 2 subjects. These symptoms disappeared after dilution of the nutrient and the use of an antispastic medication. Three other subjects with AN, part of a group of malnourished individuals with different disease states, received EN by the nasogastric route. 12 The caloric support was not specified and a mean weight increase of 5.1 kg/month was obtained. without any side
effect. This and the preceding reportll seem to point out that the slower weight increase with EN, in comparison to TPN, is balanced by the absence of significant com-
Reprint requests: Dr. Giovanni Bufano, Via Don Sturzo, Parma. Italy.
no.
6, 43100
erate rise of serum
aspartate and alanine aminotransferases. Both of these abnormalities were transient and apparently not associated with clinical disturbances. ( Journal of Parenteral and Enteral Nutrition 14:404-407, 1990)
plications. METHODS
Over the course of 5 years, nine young women were referred to us because of an extreme reduction of spontaneous food intake and prominent body weight loss. All patients met, the DSM III criteria for AN, except two, who did not reach the loss of 25% or more of the original body weight. None of the patients reported bulimia, vomiting, purging or diuretic abuse. All but one reported amenorrhea. In one case amenorrhea preceded loss of weight. The length of time elapsed between the onset of anorexia and amenorrhea and, respectively, our examination is reported in Table I. At admission, every patient underwent anthropometric measurements such as actual body weight, left tricipital skinfold, mid-arm muscle circumference, and blood chemical tests including total serum protein and serum albumin (both SMAC II, Technicon), serum t’tansferrin expressed by total iron
binding capacity of serum (TIBC), serum prealbumin (immunoprecipitation and nephelometric determination: Laser nephelometer, Hyland), serum total cholesterol (SMAC II, Technicon), serum potassium (SMAC II, Technicon), magnesium (Spectrophotometer Video 12), and inorganic phosphorus (SMAC II, Technicon), serum alanine aminotransferase and aspartate aminotransferase (SMAC II, Technicon), and hemoglobin concentration. A careful 24-hr urine collection allowed the determination of daily urea excretion which was employed for the estimate of apparent nitrogen balance (NB) according to the equation: NB (g/24 hr) protein intake (g/24 hr)/6.25 - (urinary urea nitrogen (g/24 hr) + 2.5). During the first 3 days after admission the patients were allowed free spontaneous feeding, with simultaneous recording by a dietist of caloric and proteic intake. Daily caloric intake of each subject in this phase averaged a value less than basal metabolic rate, estimated
404 Downloaded from pen.sagepub.com at UNIV CALGARY LIBRARY on May 26, 2015
=
405
proteins 19.7%, carbohydrates 44%, lipids 36%) was started with a gradual induction (25% of caloric and osmotic load on the 1st day, 50% on the 2nd day, 75% on the 3rd, 100% afterwards). Caloric requirements were
The administration of nutrient did not induce side effects such as vomiting, nausea, diarrhea, or abdominal pain. Acceptance of treatment was encouraged by previous agreement about the upper limit of weight increase to be obtained with EN: this limit was fixed at intermediate level between actual weight and the weight preceding anorexia. Some subjects complained of intestinal constipation: this disturbance was corrected with occa-
calculated using the Harris-Benedict equation modified
sional 10% glycerol
factor of about 2.0. This was equal to a mean daily caloric load of 2311 ± 607 Kcal. Every day nutrient was administered by continuous drop infusion for 14 hr. Weekly, during EN and at the end of the treatment,
Arethropometric and Biochemical Measurements
according to the Harris-Benedict equation. Thus, our patients were trained to EN after nasogastric tubing with a silicone Ch 0.9, mercury-tipped tube (Vygon 23967). Enteral feeding with
by
a
polymeric formula (Nutrisond;
enemas.
a
blood chemistry tests were repeated. Body weight was checked daily, whereas other anthropometric data were controlled at the end of the treatment. This was considered accomplished when the patients asked to be returned to spontaneous feeding and showed a food intake exceeding the basal metabolic rate, calculated according to the Harris-Benedict equation, by at least 30% for 3 consecutive days. At this point the patients were discharged. The follow-up included outpatient visits every 2 weeks and psychiatric consultation if needed. Statistics The difference between the
anthropometric and blood
chemistry tests before and after the treatment were analyzed by means of Student’s test for paired data. RESULTS
Treatment Course and
± 14 days (range 9-51 days). The shift to spontaneous eating was the result of a surge of intense appetite which occurred after a variable length of treatment. Two patients showed a shift toward a very high spontaneous caloric intake (more than 3000 Kcal/day) which lasted about a month after the dis-
distinguished by
treatment
and
rose
to
Five subjects showed serum total cholesterol concentration higher than 200 mg/dl; these subjects decreased their cholesterol concentration after the treatment. In one subject, whose serum total cholesterol level was 116 mg/dl before treatment, the level rose to 300 mg/dl at the end of the treatment. Serum electrolytes, scarcely altered in the pretreatment phase, were modified by EN
different extent. Serum potassium levels were in 3 subjects and were normalized after EN. Serum magnesium was in the normal range in pretreatment phase and showed scanty and not significant variations after EN. Serum phosphorus pretreatment levels were normal in all of our subjects. Three of them, after 2 weeks of EN, showed a marked fall of serum phosphorus level, which in 2 cases fell below 0.2 mmol/ a
slightly subnormal
Complications
was
markedly negative before positive values after EN. was
to
One patient, after 2 weeks of EN, refused to continue the treatment and asked to be discharged from the hospital. The other subjects accepted EN surprisingly well, showing a cooperative behavior after the first days of treatment. The length of EN was variable, averaging 21
charge. Their eating behavior
EN induced a remarkable change in the anthropometric, and some of the biochemical, variables of our subjects (Table II). Body weight increased at an average of 8.22 ± 3.43 kg/month. Also, tricipital skintold and mid-arm muscle circumference rose to a significant extent. Among biochemical parameters, only the faster components of serum proteins (prealbumin and TIBC) significantly increased, whereas serum albumin and total serum proteins showed no significant change. Apparent nitrogen balance
TABLE II Parameters measured before and afthr F’N
an
increase of the number of meals (5-6 meals/day) and by a relative rise of carbohydrate in comparison to protein
intake. TABLE I
* BNN’, body weight: ~’I11~, left mid-arm muscle circum!’ left tricipital skiiifold, .~, serum albumin; PA. serum prealtai._ total iron binding capacity of serum; C, total serum cholestl.,......BL1’. serum alanine aminotransferase: AST. serum aspartate aminotran,;.. °
-
ferase ; ‘~4g, serum magnesium; K, serum potassium; P. serum inorganic phosphorus; Hb, hemoglobin: NB, nitrogen apparent balance. + NS, not significant.
Downloaded from pen.sagepub.com at UNIV CALGARY LIBRARY on May 26, 2015
406
liter. These biochemical abnormalities were not associated with apparent clinical disturbances and promptly disappeared after iv administration of 150 mmol of phosphorus as fructose 1-6-diphosphate. Before treatment most patients showed a normochromic, normosideremic anemia of minor degree. After EN, mean hemoglobin concentration decreased somewhat. There was no significant change of red cell volume at post-treatment control. An increase of serum alanine aminotransferase and aspartate aminotransferase was observed in all of our subjects after EN. Both enzymes returned to normal levels of 1 to 2 weeks after stopping EN.
TPN, it induces a similar weight gain with only minor and transient complications such as hypo]phosphatemia and slight increase of serum aminotrans~Compared to
Department of Medicine
C.
Cattani,
I Division
AND
of Medicine, Ospedale Maggiore, Parma, Italy
ABSTRACT. Biochemical and anthropometric data were collected in 9 subjects with anorexia nervosa before and after enteral nutrition by the nasogastric route. All subjects but one accepted the treatment which was stopped as soon as an adequate spontaneous food intake was resumed. Enteral nutrition caused a significant increase of mid-arm muscle circumference and of tricipital skinfold. Body weight gain was remarkable, averaging a rate of 8.22 kg/month. This value was not different from the levels attained by anorectic subjects put on total
parenteral nutrition. Furthermore, a significant rise of some biochemical indexes, namely prealbumin and total iron binding capacity, was achieved. The most impressive drawbacks of enteral nutrition were occasional hypophosphatemia and mod-
The great variety of therapeutic efforts reported in the medical literature on anorexia nervosa (AN) reflects the present lack of knowledge of the pathogenesis of the disease, which seems to be increased in frequency’ and severity’ during the last decades. The poor response to treatment is emphasized by some recent research4.5 showing persistent personal difficulties and anomalous feeding behavior in spite of a resumption of normal body weight and regular menses. The same authors state that about 25% of these patients continue in amenorrhoea in spite of regained original weight, and one-third of the patients do not regain the original weight at all. A critical point, which seems to have been poorly investigated, regards the role of artificial nutrition in the treatment of AN. This approach aims not only at facing actual malnutrition of the anorectic patients, but also breaks off the vicious circle resulting from the manutrition itself, which could enforce more and more the anorectic behavior.’ In 32 cases7-10 total parenteral nutrition (TPN) was employed. The patients responded to the treatment with a very high rate of weight increase (about 10 kg/month), but they developed many complications (46%), half of which were clinically relevant and one, fatal. There are very few reports about the use of enteral nutrition (EN) in AN. This kind of nutritional support was administered to 3 patients by the intrajejunal route, with an energetic supply ranging from 2000 to 2400 Kcal/day.l A mean weight increase rate of 5.9 kg/month was obtained, the only complication reported being the occurrence of abdominal pain and diarrhea in 2 subjects. These symptoms disappeared after dilution of the nutrient and the use of an antispastic medication. Three other subjects with AN, part of a group of malnourished individuals with different disease states, received EN by the nasogastric route. 12 The caloric support was not specified and a mean weight increase of 5.1 kg/month was obtained. without any side
effect. This and the preceding reportll seem to point out that the slower weight increase with EN, in comparison to TPN, is balanced by the absence of significant com-
Reprint requests: Dr. Giovanni Bufano, Via Don Sturzo, Parma. Italy.
no.
6, 43100
erate rise of serum
aspartate and alanine aminotransferases. Both of these abnormalities were transient and apparently not associated with clinical disturbances. ( Journal of Parenteral and Enteral Nutrition 14:404-407, 1990)
plications. METHODS
Over the course of 5 years, nine young women were referred to us because of an extreme reduction of spontaneous food intake and prominent body weight loss. All patients met, the DSM III criteria for AN, except two, who did not reach the loss of 25% or more of the original body weight. None of the patients reported bulimia, vomiting, purging or diuretic abuse. All but one reported amenorrhea. In one case amenorrhea preceded loss of weight. The length of time elapsed between the onset of anorexia and amenorrhea and, respectively, our examination is reported in Table I. At admission, every patient underwent anthropometric measurements such as actual body weight, left tricipital skinfold, mid-arm muscle circumference, and blood chemical tests including total serum protein and serum albumin (both SMAC II, Technicon), serum t’tansferrin expressed by total iron
binding capacity of serum (TIBC), serum prealbumin (immunoprecipitation and nephelometric determination: Laser nephelometer, Hyland), serum total cholesterol (SMAC II, Technicon), serum potassium (SMAC II, Technicon), magnesium (Spectrophotometer Video 12), and inorganic phosphorus (SMAC II, Technicon), serum alanine aminotransferase and aspartate aminotransferase (SMAC II, Technicon), and hemoglobin concentration. A careful 24-hr urine collection allowed the determination of daily urea excretion which was employed for the estimate of apparent nitrogen balance (NB) according to the equation: NB (g/24 hr) protein intake (g/24 hr)/6.25 - (urinary urea nitrogen (g/24 hr) + 2.5). During the first 3 days after admission the patients were allowed free spontaneous feeding, with simultaneous recording by a dietist of caloric and proteic intake. Daily caloric intake of each subject in this phase averaged a value less than basal metabolic rate, estimated
404 Downloaded from pen.sagepub.com at UNIV CALGARY LIBRARY on May 26, 2015
=
405
proteins 19.7%, carbohydrates 44%, lipids 36%) was started with a gradual induction (25% of caloric and osmotic load on the 1st day, 50% on the 2nd day, 75% on the 3rd, 100% afterwards). Caloric requirements were
The administration of nutrient did not induce side effects such as vomiting, nausea, diarrhea, or abdominal pain. Acceptance of treatment was encouraged by previous agreement about the upper limit of weight increase to be obtained with EN: this limit was fixed at intermediate level between actual weight and the weight preceding anorexia. Some subjects complained of intestinal constipation: this disturbance was corrected with occa-
calculated using the Harris-Benedict equation modified
sional 10% glycerol
factor of about 2.0. This was equal to a mean daily caloric load of 2311 ± 607 Kcal. Every day nutrient was administered by continuous drop infusion for 14 hr. Weekly, during EN and at the end of the treatment,
Arethropometric and Biochemical Measurements
according to the Harris-Benedict equation. Thus, our patients were trained to EN after nasogastric tubing with a silicone Ch 0.9, mercury-tipped tube (Vygon 23967). Enteral feeding with
by
a
polymeric formula (Nutrisond;
enemas.
a
blood chemistry tests were repeated. Body weight was checked daily, whereas other anthropometric data were controlled at the end of the treatment. This was considered accomplished when the patients asked to be returned to spontaneous feeding and showed a food intake exceeding the basal metabolic rate, calculated according to the Harris-Benedict equation, by at least 30% for 3 consecutive days. At this point the patients were discharged. The follow-up included outpatient visits every 2 weeks and psychiatric consultation if needed. Statistics The difference between the
anthropometric and blood
chemistry tests before and after the treatment were analyzed by means of Student’s test for paired data. RESULTS
Treatment Course and
± 14 days (range 9-51 days). The shift to spontaneous eating was the result of a surge of intense appetite which occurred after a variable length of treatment. Two patients showed a shift toward a very high spontaneous caloric intake (more than 3000 Kcal/day) which lasted about a month after the dis-
distinguished by
treatment
and
rose
to
Five subjects showed serum total cholesterol concentration higher than 200 mg/dl; these subjects decreased their cholesterol concentration after the treatment. In one subject, whose serum total cholesterol level was 116 mg/dl before treatment, the level rose to 300 mg/dl at the end of the treatment. Serum electrolytes, scarcely altered in the pretreatment phase, were modified by EN
different extent. Serum potassium levels were in 3 subjects and were normalized after EN. Serum magnesium was in the normal range in pretreatment phase and showed scanty and not significant variations after EN. Serum phosphorus pretreatment levels were normal in all of our subjects. Three of them, after 2 weeks of EN, showed a marked fall of serum phosphorus level, which in 2 cases fell below 0.2 mmol/ a
slightly subnormal
Complications
was
markedly negative before positive values after EN. was
to
One patient, after 2 weeks of EN, refused to continue the treatment and asked to be discharged from the hospital. The other subjects accepted EN surprisingly well, showing a cooperative behavior after the first days of treatment. The length of EN was variable, averaging 21
charge. Their eating behavior
EN induced a remarkable change in the anthropometric, and some of the biochemical, variables of our subjects (Table II). Body weight increased at an average of 8.22 ± 3.43 kg/month. Also, tricipital skintold and mid-arm muscle circumference rose to a significant extent. Among biochemical parameters, only the faster components of serum proteins (prealbumin and TIBC) significantly increased, whereas serum albumin and total serum proteins showed no significant change. Apparent nitrogen balance
TABLE II Parameters measured before and afthr F’N
an
increase of the number of meals (5-6 meals/day) and by a relative rise of carbohydrate in comparison to protein
intake. TABLE I
* BNN’, body weight: ~’I11~, left mid-arm muscle circum!’ left tricipital skiiifold, .~, serum albumin; PA. serum prealtai._ total iron binding capacity of serum; C, total serum cholestl.,......BL1’. serum alanine aminotransferase: AST. serum aspartate aminotran,;.. °
-
ferase ; ‘~4g, serum magnesium; K, serum potassium; P. serum inorganic phosphorus; Hb, hemoglobin: NB, nitrogen apparent balance. + NS, not significant.
Downloaded from pen.sagepub.com at UNIV CALGARY LIBRARY on May 26, 2015
406
liter. These biochemical abnormalities were not associated with apparent clinical disturbances and promptly disappeared after iv administration of 150 mmol of phosphorus as fructose 1-6-diphosphate. Before treatment most patients showed a normochromic, normosideremic anemia of minor degree. After EN, mean hemoglobin concentration decreased somewhat. There was no significant change of red cell volume at post-treatment control. An increase of serum alanine aminotransferase and aspartate aminotransferase was observed in all of our subjects after EN. Both enzymes returned to normal levels of 1 to 2 weeks after stopping EN.
TPN, it induces a similar weight gain with only minor and transient complications such as hypo]phosphatemia and slight increase of serum aminotrans~Compared to
