- Email: [email protected]
Effects of long-term nutritional rehabilitation on body composition and clinical status in malnourished children and adolescents with cystic fibrosis
Effects of long-term nutritional rehabilitation on body composition and clinical status in malnourished children and adolescents with cystic fibrosis Fourteen patients aged 4.9 to 21.5 years" with cystic .fibrosis and moderate to severe lung disease, malnutrition, or growth failure were given nocturnal supplemental feeding by gastrostomy tube. Mean Jollow-up was for 1.1 years (range 0.8 to 2.78 years). Patients were studied to observe the effect of nutritional support on body composition, growth, pulmonary function, and quality of life. A contemporary group o f patients with CF was retrospectively pair matched to the study group. The supplemental feeding resulted in positive changes in body composition and in growth velocity. Weight, as a percentage o f standard in the control group, declined by 3% over 1 year, whereas it increased by 2 % in the treatment group (P <0.05). Pulmonary function, assessed as a percent o f predicted FVC and FEV~, did not change significantly in the treatment group over 1.1 years, whereas FVC declined by 12% (P <0.01) and FEV~ declined by 13% (P <0.01) in the control group. There was a marked increase in patient ability to participate in activities of daily living, even in those patients in whom pulmonary function deteriorated during the study. (J PEDIATR 1985;107:225-230)
Lance D. Levy, M.B., Ch.B., Peter R. Durie, M.D., Paul B. Pencharz, M.B., Ph.D., and Mary L. Corey, M.Sc. Toronto, O n t a r i o , C a n a d a
THE DEVELOPMENT of malnutrition and an accelerated decline in pulmonary function have been noted to occur together in patients with cystic fibrosis. I We and other authors have speculated that these two findings are causally related? It has been suggested that, as lung disease worsens, energy expenditure rises because of an increase in the work of breathing. 3 Eventually a point is reached where
From the Research Institute, The HospitalJor Sick Children, and the Departments o f Paediatrics and Nutritional Sciences, The University of Toronto. Supported by a grant from the Canadian Cystic Fibrosis Foundation. Dr. Levy was supported by a fellowship from the Human Nutrition Research Council o f Ontario for 1 year, and subsequently by a fellowship from the Canadian Cystic Fibrosis Foundation. Submitted for publication Sept. 12, 1984; accepted Jan. 25, 1985. Reprint requests: Dr. Paul B. Pencharz, Head, Division of Clinical Nutrition, The Hospital for Sick Children, 555 University Ave., Toronto, Ont., Canada M5G IX8.
the patient is no longer able to ingest sufficient calories to maintain energy balance. Progressive weight loss occurs, and with it, lean tissue wasting.4 If this process continues unchecked, a reduced survival rate is certain. To test our hypothesis that restoration of nutritional status would ameliorate the accelerated rate of decline in pulmonary CF FEVI FVC TBK TBN
Cystic fibrosis Forcedexpiratory volume in I second Forcedvital capacity Total body potassium Total body nitrogen
function and perhaps lengthen survival, we undertook, by means of nocturnal gastrostomy feeding, the nutritional rehabilitation of malnourished patients with CF who could not restore energy balance by the oral route. We describe those changes in body composition, growth rate, body water, and pulmonary function in 14 patients followed prospectively for up to 3 years after gastrostomy, Changes
TheJournalofPED1ATRICS
225
226
Levy et al.
in pulmonary function and nutritional status are compared with those observed in a group of patients with CF retrospectively matched from our clinic population. METHODS We describe the course in 14 patients from the time of gastrostomy until December 1, 1984. Baseline data were accumulated during a period of not more than 3 months prior to gastrostomy and while patients did not have an acute intercurrent chest infection. All patients but one required pancreatic enzyme supplementation. Two patients were too young to perform pulmonary function tests adequately. Patient selection. We screened patients in our CF clinic (n = 550) for possible malnutrition. Patients whose weight as a percentage of ideal 5 for age, sex, and height fell below 90%, or who had growth retardation (stunting), or who failed to regain weight lost during chest infection were referred to the Division of Clinical Nutrition for a formal nutritional assessment. After assessment, it was often possible to suggest changes in diet or to begin the use of high-calorie feeding supplements. Patients not responding to this approach over a 3-month trial period were further evaluated for entry into the gastrostomy program. We determined, by means of patient and parent interviews, which families were psychologically suited to this type of intervention. Although inclusion criteria were generally clear-cut (i.e., weight loss and growth failure), exclusion criteria were harder to establish. We excluded patients with clinically evident heart failure, insulin-dependent diabetes mellitus, and carbon dioxide retention. The subject of exclusion criteria became a major focus within the feeding program as the study proceeded, and will be the subject of a further report. Our gastrostomy feeding program was approved by this hospital's Human Subject Review Committee, and the risks and benefits of gastrostomy and gastrostomy feeding were explained in detail to both the parents and the patients. Patients entering the program were admitted to the hospital for further evaluation prior to gastrostomy. Nutritional assessment and body composition measurements. Patient height, weight, skinfold thickness (triceps, biceps, subscapular, and suprailiac), and total body potassium and nitrogen were measured as previously reported: Pregastrostomy growth velocities were calculated from data for each patient over the 12-month period immediately prior to treatment. Growth velocities (before and after gastrostomy) were normalized by expressing them as a percentage of the 50th percentile value for age and sex, from tables developed by Tanner et al) Total body water and extracellular water compartments were measured with ~80-labeled water 7 and sodium bromide, 8 respectively.
The Journal of Pediatrics August 1985
Intracellular water was calculated as the difference between total body water and extracellular water. Pregastrostomy screening included tests of renal, hepatic, and hematopoietic function. Samples were obtained during the week prior to gastrostomy and analyzed by standard techniques. Pulmonary function assessment. Pulmonary function was tested prior to gastrostomy. Results were expressed as a percentage of the predicted value.9 Testing was performed according to the recommendations of the GAP Conference Report. t~ FEVj was obtained before bronchodilation. Baseline or repeat pulmonary function data were excluded if a patient had an acute intercurrent chest infection at the time of measurement. Gastrostomy tube insertion and feeding. A gastrostomy tube was placed percutaneously under endoscopic control without incision.tt Sedation and local anesthetic were used in all but the youngest patients, who required a brief general anesthetic. The morning after gastrostomy tube placement, feedings were begun with either an elemental (Vivonex; Norwich Eaton, Paris, Ontario, or Flexical HN; Mead Johnson, Candiac, Quebec) or a semielemental formula (Vital; Ross Laboratories, Montreal, Quebec). This was given initially by continuous infusion 24 hours a day while the patient took sips of fluid by mouth. The concentration of the formula was advanced gradually over a period of 3 or 4 days. Generally by the fourth day after gastrostomy, patients began taking an ad libitum regular diet during the day. Meals were accompanied by supplemental enzyme capsules. Supplemental feeding by gastrostomy tube was then given nocturnally over 10 to 12 hours. Patients tolerated a volume of feeding solution that gave approximately 30% of estimated daily energy requirements) 2Patients remained in the hospital for 10 to 14 days postoperatively. After discharge, patients were encouraged to eat regular meals. Postgastrostomy estimates of oral caloric intake were not routinely done if patients reported a good appetite for lunch and dinner and gained weight. Enzyme supplements were not required with semielemental and elemental gastrostomy feeds. A pilot study (unpublished) showed that energy losses on these chemically defined diets were not more than 8%. Patients were seen in clinic at 2-week intervals during the first month after gastrostomy and then every 3 months. Adjustments were made in the energy density or volume of nocturnal feeding in the event that weight gain was inadequate. Feedings of all patients were changed to Vital during the second year of this study, because this preparation appeared to cause less gastric upset. At each clinic visit, anthropometric measurements were made. Blood was drawn to monitor renal, hepatic, and hematopoietic function. TBK was measured at each visit after the first month,
Volume 107 Number 2
Malnutrition and pulmonary function in cystic fibrosis
227
Table I. Changes in body composition, growth velocities, and body water compartments after gastrostomy in 14 patients with C F
Mean -+ SD Body composition (n = 14) ABodyfat~'(%) 2x Fat-free mass (kg) A TBK (gin) A TBN (kg) Growth (n = 14) A Height (em) A Weight (kg) A Weight per height (% std):~ % A Height velocityw % A Weight velocityw Body water (n = 5) A Total body water (L) A Extracellular water (L) A Intracellular water (L)
3.0 3.4 18.0 0.10
• -+ • +-
2.0 2.0 14 0.17
Range --1.0to7.0 0.0 to 7.0 -1.0 to 40.0 -0.17 to 0.30
6.0 _+ 5.0 6.0 -+ 3.5 5,0 _+ 5.0 321 700
0.0 to 14.0 0.6 to 13.0 -3.0 to 12.0
4.0 _+ 2.3 0.4 -+ 1.1 3.0 -+ 3.0
2.0 to 8.0 -1.0 to 2.0 2.0 to 8.0
Elapsed Time* (yr)
1.1 _+0.7
1.2 • 0,7
]
P 0.005 0.005 0.005 0.05 0.005 0.005 0.005 0.005 0.005 0.001 NS 0.05
TBK, total body potassium; TBN, total body nitrogen; NS, not significant. *Time elapsed from date of gastrostomy to most recent observation. 1"Percentage of body weight composed of fat as determined by triceps, biceps, subscapsular, and supradiliac skinfold anthropometry. :~Weight as percentage of standard for age, sex, and height (see reference 5). w change in mean height and weight velocities.
and T B N every second visit (6-monthly). In five patients, body water compartments were serially measured. Anthropometry, TBK, T B N , body water determinations, and blood testing data were all obtained on the same day to allow comparison analysis. Pulmonary function tests were repeated at 3-month intervals. The final data for growth, body composition, and pulmonary function were obtained at the last clinic visit prior to December 1, 1984. The total duration of gastrostomy feeds is shown in Table I. Selection of matched comparison subjects. To assess further the effects of the gastrostomy feeding on lung function and nutritional status, an effort was made to match the treated subjects with controls derived from the computerized data base of our C F clinic population. A listing of all patients whose sex and birth year were the same as those of one of the gastrostomy patients provided the pool of potential comparison subjects. Pulmonary function and weight as a percentage of ideal values at the age nearest the pregastrostomy age of the respective patient were retrieved. Average values for pulmonary function and weight as a percentage of ideal were computed for each sex/birth year group. The number of sex/birth year matches who had not survived to the relevant age was also noted. Comparison patients were selected to have weight for height within 10% and FEV1 within 20% of the respective study subject values. For those subjects without a matched control by these criteria, the matching procedure was repeated for the birth year before and after the study subject birth year. Computer
graphs of subjects and controls were used to verify the similarity of growth and pulmonary course in the 6 months before the gastrostomy age. Statistics. Because controls were selected retrospectively, at the end of the study, their assessment dates did not match those of the subjects. Therefore, pregastrostomy values in the two groups were compared by using each patient's average measurements in the year before the relevant gastrostomy age. Postgastrostomy values for each patient were calculated as the average of measurements from the gastrostomy age to the end of the study. Paired t tests were used to evaluate mean changes during the follow-up period in the study group and in the comparison group. Changes in percent of ideal weight for height and in percent of predicted FVC and FEVI were evaluated. RESULTS The 14 gastrostomy subjects (five males) had a mean age of 12.9 years • 4.6 SD (range 4.9 to 21.5 years). Weight as a percent of standard was 82% • 10% (range 65% to 104%). The individual with a weight as a percentage of ideal of 104% was stunted. Mean percentage of body weight as fat was 7.0% in males and 15% in females. Average values for normal children of the same age are 11% for males and 19% for females? 3 Pregastrostomy height velocity was 74% • 21% of standard, and weight velocity was 28% _+ 24% of standard. Initial FVC was 45% • 14% predicted, and FEV~ was 41% • 16% predicted.
228
Levy et al.
The Journal of Pediatrics August 1985
Table II. Complications following gastrostomy (in order of importance) Complication
Treatment
Early (0 to 1 month) Early morning fullness and anorexia, nausea
Usually none; generally resolves with time; metoclopramide useful in some refractory cases Move T-bar to prevent skin chafing as weight gain occurs Clotrimazole cream with or without hydrocortisone Vigorous QID physiotherapy from the day of insertion for prevention; early ambulation Seen in two of 21 patients; resolved with orally administered cloxacillin
Local tenderness Monilial infection Chest infection (0 to 7 days after insertion) Local cellulitis Late Local proliferation of granulation tissue Accidental tube removal
Cautery with silver nitrate Tube replaced and position checked with gastrografin x-ray study
Table IIl. Changes in pulmonary function and in weight/height in gastrostomy patients and in comparison group FVC (% predicted)
FEV~ (% predicted)
Weight~height (% standard) Tilq'le *
Patients (n = 10) Difference P Comparison group
Initial
Fina'l
Initial
Final
Initial
Final
(yr)
53 + 13
49 _+ 16
47 _+ 15
42 _+ 17
85 • 12
87 + 14
1.1 -+ 0.43
61 _+ 20
-4.0 NS 49•
NS 81+10
1.01 _+ 0.24
-5.0 56+21
+2.0 NS 43•
8429
(n = lO)
Difference P
-12 <0.0I
-13 <0.01
-3.0 <0.02 P <0.05t
Values are mean _+SD. lntragroup comparison by Student paired t test. *Time elapsed from gastrostomy or gastrostomy age in matched comparison subjects. ~'Intergroup statistical comparison (unpaired t test) for changes in weight/height (% standard).
Effects of gastrostomy feeding. Significant increases in percent body fat, fat-free mass, TBK, and TBN occurred in the gastrostomy patients (Table I). The increase in TBK was matched by an increase in total body water, which resulted from growth of the intracellular water compartment. Most notably, extracellular water did not increase with gastrostomy feeding. A linear relationship between TBK and intracellular water was found (r = 0.68, P <0.025). Postgastrostomy height velocity increased, showing catch-up growth from a mean of 74% of standard before gastrostomy to 162% of standard in the first 0.81 year of supplemental feeding. By a mean time of 1.2 • 0.7 years this rate of growth had increased further to 238% of standard. Weight velocity showed a similar pattern of rapid initial increase, and did not reach a plateau until a subject approached ideal weight for height. Body fat increased a mean of 3.0% _+ 2.0% (P <0.005). A significant improvement in mean hemoglobin occurred during the initial 3 months of nutritional rehabil-
itation, from 128 _+ 14 g m / L to 138 +_ 11 g m / L (P <0.005), by which time, hemoglobin levels had stabilized. Serum sodium and albumin concentrations were within the normal range, and did not change significantly. Gastrostomy tube complications were generally mild and occurred either immediately after insertion or much later (Table II). Comparison of subjects and controls. For each gastrostomy-fed patient, there were at least four and as many as 20 patients with the same sex and birth year being followed in our clinic. In only one patient were pregastrostomy values similar to mean values for his sex/birth year control group; all others had markedly lower than average CF clinic values for weight, height, and FEV~. No gastrostomy patient had more than one control in the same birth year matched for weight for height and FEV~. Controls for two female gastrostomy patients came from an adjacent birth year. No match was found for the three youngest patients, all girls, whose weight for height and FEV~ (when mea-
Volume 107 Number 2
Malnutrition and pulmonary function in cystic fibrosis
I INFECTION
1-- .........
"
"~ I
[
l DETERIORATING
I
LUNGFU}NCTION I
229
.t .
[
WEIGHT LOSS
Figure. Model defining interrelationship between deteriorating lung function and nutritional status in patients with cystic fibrosis.
sured) were poorer than all other birth year or adjacent birth year control m~itches~ Comparison of pregastrostomy and follow-up values for FVC, FEVb and weight as a percent of standard for the 10 patients with CF and for their controls showed that pulmonary function in t h e gastrost0my subjects did not change significantly over the 0.8 to 2.8 years, whereas it declined in the controls (P <0.01). Mean weight for height in the 10 gastrostomy patients increased by 2%, but declined by 3% in the controls (P <0.05). Deaths during the study, Six patients died during the 3-year period to December 1, 1984. Three patients, who were our youngest (ages 4,9, 5.8, and 8.9 years) and the most seriously affected in terms of lung disease, did not have matched controls. The other three did have matches within the comparison group: a man aged 18.7 years at the time of gastrostomy survived 1.2 years longer than his match; a woman aged 17.4 years at the time of gastrostomy survived 1.8 years longer than her match; and a man aged 21.5 years at the time of gastrostomy lived 0.6 year after gastrostomy, whereas his match lived 0.8 year after gastrostomy age. These three older patients died of respiratory failur e secondai'y to acute pneumonia caused by Pseudomonas cepacia, which is resistant to all known antibiotics. One of the matches died before pulmonary function tests could be repeated, reducing the number of pairs from 11 to 10 (Table III). DISCUSSION We have Proposed a model ~4 (Figure) that defines the interrelationship between deteriorating lung function and nutritional status in patients with CF. Other workers 2' ~5.~6 have attempted to break this apparent cycle of progressive weight loss and deteriorating lung function by providing short-term nutritional support. None of these studies showed lasting improvements in pulmonary function or body composition. We have demonstrated that long-term nutritional Supplementation will produce catch-up and sustained growth in a group of patients with CF. The
prolonge d improvement noted in weight as a percent of ideal, and the stabilization in pulmonary function status are significant when compared with those of a matched comparison group. It is possible that several patients had an increase in growth related solely to the start of puberty. However, our experience with these and other patients with CF has shown that growth velocity slows when nutritiona! support is discontinued and malnutrition recurs. The increase in weight for height did not, i n most cases, result in a normalization of weight as a percentage of standard for height, age, and sex. It is probable that several as yet imprecisely understood patient-reported problems led to a reduction in voluntary food intake, so that weight gain was suboptimal. These problems include a subjective sensation of fullness and hence a decrease in appetite in the morning. Six patients with marked lung disease remained anorexic after gastrostomy, or continued to complain of early satiety so that daytime meals were small in size. Body composition changes occurred with a moderate recovery of adipose reserve and lean tissue mass. Initially, we were concerned that weight gains in these patients could represent merely increases in extracellular water, with the risk of compromising gaseous exchange and cardiac function. Our data show that weight gain early in the course of refeeding resulted principally from increases in fat and lean tissue rather than in extracellular water. A moderate anemia is quite common in patients with CF~7; the effects of chronic illness and infection on the hematopoietic system may in part be the causes. However, two additional factors, iron deficiency~7.~ and protein malnutrition, ~9need to be considered. Our patients had a significant rise in hemoglobin during the refeeding period despite the fact that only two received therapeutic doses of iron. We suggest, therefore, that nutritional support reversed the effects of malnutrition on the hematopoietic system. It is possible that the anemia previously accepted as inevitable in unwell and often malnourished patients with CF may be improved by refeeding.
230
Levy et al.
Pulmonary function data revealed a significantly slower rate of decline in FVC and FEV. in the gastrostomy group. Consistent with this. all gastrostomy subjects reported an increased ability to participate in activities of daily life. If such participation resulted in more physical exercise, this may have had the effect of "training" respiratory musculature2 ~ Whether nutritional rehabilitation materially affects the severity of acute chest infections is unknown. The exact mechanism of how nutritional support benefits lung function remains to be resolved. We were aware of the possible effects of the gastrostomy program on psychosocial functioning of the patient and family. None of the C F clinic team of physicians, nurses, social workers, or psychiatrist were able to identify any deterioration in patient or family function: rather, the converse was observed We set out to investigate the effectiveness of the gastrostomy program as a means of treating malnutrition in patients with CF with advanced lung disease who were unresponsive to intensive dietary counseling and supplementation. The question remains whether the treatment of malnutrition affects life quality and survival. Ultimately, these questions can only be answered by randomized control study. Presumably, the control group would continue to receive "optimal" dietary counseling and supplementation. Iriasmuch as all our patients had been unresponsive to at least three months of such dietary therapy before being offered gastrostomy, we believe that there may be serious practical and ethical problems in designing and conducting a truly randomized study. Based on the results of this study, we propose that the long-term maintenance of adequate nutrition in selected malnourished patients with C F can safely and practically be done using nocturnal gastrostomy feeding. Furthermore, we have evidence that nutritional support may delay a decline in lung function significantly, and that this may translate into improved health, quality of life, and length of survival in some patients with CF. We thank Dr. Henry Levis0n (Director), Dr. Gordon Forstner, and the staff physicians of The Hospital for Sick Children Chest and CF Clinic for support and encouragement; Dr. Robert Filler and Dr. David Wesson for insertion of the gastrostomy tubes; Dr. Joan Harrison (Director, Medical Physics Laboratory, Toronto General Hospital) for the body nitrogen and potassium measurements; Dr. R. Hancock (SLOWPOKE Reactor, University of Toronto) for help with the bromide analyses of extracellular water; the CF ward and clinic nursing staff; the CF dietitiannatritionists for help with the dietary assessments; Ruth Clarke and Linda Wykes for technical assistance; and Heather Keillor for manuscript preparation.
The Journal of Pediatrics August 1985
REFERENCES
I. Gurwii.z D Covey M, Francis PWJ, et al: Perspectives in cystic fibrosis. Pediatr Clin North Am 26:603, 1979. 2. Gaskin K, Gurwitz D, Durie P, et al: Improved respiratory prognosis in patien{s with normal fat asorption. J PEDIATR i00;857, 1982. 3. Levison H, Cherniak R: Ventilatory cost of exercise in chronic obstructiye pulmonary disease. J Appl Physiol 25:2l, 1968. 4. Shizgal HM: The effect of malnutrition on body composition. Surg Gynecol Obstet 152:22, 1981. 5. Tanner JM, Whitehouse RH, Takaishi M: Standards from birth to maturity for height, Weight, height velocity and weight velocity: British children, 1965. ll. Arch Dis Child 41:613, 1966. 6. Archibald EH, Harrison JE, Pencharz PB: Effect of a weight-reducing high-protein diet on the body composition 0f obese adolescents. Am J Dis Child 137:658, 1983. 7. Schoeller DA, Van Santen E, Peterson DW, et al: Total body watei measurement in humans with zsO and 2H labeled'water. Am J Clin Nutr 33:2686, 1980. 8. Holzbecher J, Ryan DE: The rapid determination of total bromine and iodine in biological fluids by neutron activation. Clin Biochem 13:277 1980. 9. weng T-R, Levison H: Standards of pulmonary function in children. Am Rev Respir Dis 99:8791 1969. 10. Cystic Fibrosis "GAP" Conference Report: Standardization of lung ftmction testing in children. Cystic Fib?osis Foundation, 1978. 11. Gauderer MWL, Ponsky JL, lzant JR: Gastrostomy without laparotomy: A percutaneous endoscopic technique. J Pediatr Surg 15:872, 1980. 12. Recommended nutrient intakes for canadians: Dietary Standard for Canada. Ottawa, Ont., Canada, Bureau of Nutritional Sciences, Food Directorate, Health Protection Branch, Department of National Health and Welfare. 13. Widdowson EM: Changes in body proportions and composition during growth. In Davis JA, Dobbing J, editors: Scientific foundations of paediatrics. London, 1981, William Heinemann Medical Books Ltd., pp 330-342. 14. Pencharz PBI Energy intakes and low fat diets in children with cystic fibrosis. J Pediatr Gastroenterol Nutr 2:400, 1983. 15. Berry HK, Kellogg Fw, Hunt MM, et aI: Dietary supplement and nutrition in children with cystic fibrosis. Am J Dis Child 129:165, 1975. 16. Mansell AL, Andersen JC, Muttard CR, et al: Short-term pulmonary effects of total parenteral nutrition in children with cystic fibrosis. J PEDJATR104:700, 1984. 17. Ater JL, Herbst J J, Landau SA, et al: Relative anemia and iron deficiency in cystic fibrosis. Pediatrics 71:810, 1983. 18. Vichinsky EP, Pennathur-Das R, Nickerson B, et ah Inadequate erythroid response to hypoxia in cystic fibrosis. J PEDIATR 105:15, 1984. 19. Lindenbaum J: The hematopoietic system. In Hodges RE, editor: Human nutrition, vol 4. New York, 1979, Plenum Press, pp 1-52. 20. Cesny F J, Pullano TP, Cropp GJA: Cardiorespiratory adaptations to exercise in cystic fibrosis. Am Rev Respir Dis 126:217, 1982.
Lance D. Levy, M.B., Ch.B., Peter R. Durie, M.D., Paul B. Pencharz, M.B., Ph.D., and Mary L. Corey, M.Sc. Toronto, O n t a r i o , C a n a d a
THE DEVELOPMENT of malnutrition and an accelerated decline in pulmonary function have been noted to occur together in patients with cystic fibrosis. I We and other authors have speculated that these two findings are causally related? It has been suggested that, as lung disease worsens, energy expenditure rises because of an increase in the work of breathing. 3 Eventually a point is reached where
From the Research Institute, The HospitalJor Sick Children, and the Departments o f Paediatrics and Nutritional Sciences, The University of Toronto. Supported by a grant from the Canadian Cystic Fibrosis Foundation. Dr. Levy was supported by a fellowship from the Human Nutrition Research Council o f Ontario for 1 year, and subsequently by a fellowship from the Canadian Cystic Fibrosis Foundation. Submitted for publication Sept. 12, 1984; accepted Jan. 25, 1985. Reprint requests: Dr. Paul B. Pencharz, Head, Division of Clinical Nutrition, The Hospital for Sick Children, 555 University Ave., Toronto, Ont., Canada M5G IX8.
the patient is no longer able to ingest sufficient calories to maintain energy balance. Progressive weight loss occurs, and with it, lean tissue wasting.4 If this process continues unchecked, a reduced survival rate is certain. To test our hypothesis that restoration of nutritional status would ameliorate the accelerated rate of decline in pulmonary CF FEVI FVC TBK TBN
Cystic fibrosis Forcedexpiratory volume in I second Forcedvital capacity Total body potassium Total body nitrogen
function and perhaps lengthen survival, we undertook, by means of nocturnal gastrostomy feeding, the nutritional rehabilitation of malnourished patients with CF who could not restore energy balance by the oral route. We describe those changes in body composition, growth rate, body water, and pulmonary function in 14 patients followed prospectively for up to 3 years after gastrostomy, Changes
TheJournalofPED1ATRICS
225
226
Levy et al.
in pulmonary function and nutritional status are compared with those observed in a group of patients with CF retrospectively matched from our clinic population. METHODS We describe the course in 14 patients from the time of gastrostomy until December 1, 1984. Baseline data were accumulated during a period of not more than 3 months prior to gastrostomy and while patients did not have an acute intercurrent chest infection. All patients but one required pancreatic enzyme supplementation. Two patients were too young to perform pulmonary function tests adequately. Patient selection. We screened patients in our CF clinic (n = 550) for possible malnutrition. Patients whose weight as a percentage of ideal 5 for age, sex, and height fell below 90%, or who had growth retardation (stunting), or who failed to regain weight lost during chest infection were referred to the Division of Clinical Nutrition for a formal nutritional assessment. After assessment, it was often possible to suggest changes in diet or to begin the use of high-calorie feeding supplements. Patients not responding to this approach over a 3-month trial period were further evaluated for entry into the gastrostomy program. We determined, by means of patient and parent interviews, which families were psychologically suited to this type of intervention. Although inclusion criteria were generally clear-cut (i.e., weight loss and growth failure), exclusion criteria were harder to establish. We excluded patients with clinically evident heart failure, insulin-dependent diabetes mellitus, and carbon dioxide retention. The subject of exclusion criteria became a major focus within the feeding program as the study proceeded, and will be the subject of a further report. Our gastrostomy feeding program was approved by this hospital's Human Subject Review Committee, and the risks and benefits of gastrostomy and gastrostomy feeding were explained in detail to both the parents and the patients. Patients entering the program were admitted to the hospital for further evaluation prior to gastrostomy. Nutritional assessment and body composition measurements. Patient height, weight, skinfold thickness (triceps, biceps, subscapular, and suprailiac), and total body potassium and nitrogen were measured as previously reported: Pregastrostomy growth velocities were calculated from data for each patient over the 12-month period immediately prior to treatment. Growth velocities (before and after gastrostomy) were normalized by expressing them as a percentage of the 50th percentile value for age and sex, from tables developed by Tanner et al) Total body water and extracellular water compartments were measured with ~80-labeled water 7 and sodium bromide, 8 respectively.
The Journal of Pediatrics August 1985
Intracellular water was calculated as the difference between total body water and extracellular water. Pregastrostomy screening included tests of renal, hepatic, and hematopoietic function. Samples were obtained during the week prior to gastrostomy and analyzed by standard techniques. Pulmonary function assessment. Pulmonary function was tested prior to gastrostomy. Results were expressed as a percentage of the predicted value.9 Testing was performed according to the recommendations of the GAP Conference Report. t~ FEVj was obtained before bronchodilation. Baseline or repeat pulmonary function data were excluded if a patient had an acute intercurrent chest infection at the time of measurement. Gastrostomy tube insertion and feeding. A gastrostomy tube was placed percutaneously under endoscopic control without incision.tt Sedation and local anesthetic were used in all but the youngest patients, who required a brief general anesthetic. The morning after gastrostomy tube placement, feedings were begun with either an elemental (Vivonex; Norwich Eaton, Paris, Ontario, or Flexical HN; Mead Johnson, Candiac, Quebec) or a semielemental formula (Vital; Ross Laboratories, Montreal, Quebec). This was given initially by continuous infusion 24 hours a day while the patient took sips of fluid by mouth. The concentration of the formula was advanced gradually over a period of 3 or 4 days. Generally by the fourth day after gastrostomy, patients began taking an ad libitum regular diet during the day. Meals were accompanied by supplemental enzyme capsules. Supplemental feeding by gastrostomy tube was then given nocturnally over 10 to 12 hours. Patients tolerated a volume of feeding solution that gave approximately 30% of estimated daily energy requirements) 2Patients remained in the hospital for 10 to 14 days postoperatively. After discharge, patients were encouraged to eat regular meals. Postgastrostomy estimates of oral caloric intake were not routinely done if patients reported a good appetite for lunch and dinner and gained weight. Enzyme supplements were not required with semielemental and elemental gastrostomy feeds. A pilot study (unpublished) showed that energy losses on these chemically defined diets were not more than 8%. Patients were seen in clinic at 2-week intervals during the first month after gastrostomy and then every 3 months. Adjustments were made in the energy density or volume of nocturnal feeding in the event that weight gain was inadequate. Feedings of all patients were changed to Vital during the second year of this study, because this preparation appeared to cause less gastric upset. At each clinic visit, anthropometric measurements were made. Blood was drawn to monitor renal, hepatic, and hematopoietic function. TBK was measured at each visit after the first month,
Volume 107 Number 2
Malnutrition and pulmonary function in cystic fibrosis
227
Table I. Changes in body composition, growth velocities, and body water compartments after gastrostomy in 14 patients with C F
Mean -+ SD Body composition (n = 14) ABodyfat~'(%) 2x Fat-free mass (kg) A TBK (gin) A TBN (kg) Growth (n = 14) A Height (em) A Weight (kg) A Weight per height (% std):~ % A Height velocityw % A Weight velocityw Body water (n = 5) A Total body water (L) A Extracellular water (L) A Intracellular water (L)
3.0 3.4 18.0 0.10
• -+ • +-
2.0 2.0 14 0.17
Range --1.0to7.0 0.0 to 7.0 -1.0 to 40.0 -0.17 to 0.30
6.0 _+ 5.0 6.0 -+ 3.5 5,0 _+ 5.0 321 700
0.0 to 14.0 0.6 to 13.0 -3.0 to 12.0
4.0 _+ 2.3 0.4 -+ 1.1 3.0 -+ 3.0
2.0 to 8.0 -1.0 to 2.0 2.0 to 8.0
Elapsed Time* (yr)
1.1 _+0.7
1.2 • 0,7
]
P 0.005 0.005 0.005 0.05 0.005 0.005 0.005 0.005 0.005 0.001 NS 0.05
TBK, total body potassium; TBN, total body nitrogen; NS, not significant. *Time elapsed from date of gastrostomy to most recent observation. 1"Percentage of body weight composed of fat as determined by triceps, biceps, subscapsular, and supradiliac skinfold anthropometry. :~Weight as percentage of standard for age, sex, and height (see reference 5). w change in mean height and weight velocities.
and T B N every second visit (6-monthly). In five patients, body water compartments were serially measured. Anthropometry, TBK, T B N , body water determinations, and blood testing data were all obtained on the same day to allow comparison analysis. Pulmonary function tests were repeated at 3-month intervals. The final data for growth, body composition, and pulmonary function were obtained at the last clinic visit prior to December 1, 1984. The total duration of gastrostomy feeds is shown in Table I. Selection of matched comparison subjects. To assess further the effects of the gastrostomy feeding on lung function and nutritional status, an effort was made to match the treated subjects with controls derived from the computerized data base of our C F clinic population. A listing of all patients whose sex and birth year were the same as those of one of the gastrostomy patients provided the pool of potential comparison subjects. Pulmonary function and weight as a percentage of ideal values at the age nearest the pregastrostomy age of the respective patient were retrieved. Average values for pulmonary function and weight as a percentage of ideal were computed for each sex/birth year group. The number of sex/birth year matches who had not survived to the relevant age was also noted. Comparison patients were selected to have weight for height within 10% and FEV1 within 20% of the respective study subject values. For those subjects without a matched control by these criteria, the matching procedure was repeated for the birth year before and after the study subject birth year. Computer
graphs of subjects and controls were used to verify the similarity of growth and pulmonary course in the 6 months before the gastrostomy age. Statistics. Because controls were selected retrospectively, at the end of the study, their assessment dates did not match those of the subjects. Therefore, pregastrostomy values in the two groups were compared by using each patient's average measurements in the year before the relevant gastrostomy age. Postgastrostomy values for each patient were calculated as the average of measurements from the gastrostomy age to the end of the study. Paired t tests were used to evaluate mean changes during the follow-up period in the study group and in the comparison group. Changes in percent of ideal weight for height and in percent of predicted FVC and FEVI were evaluated. RESULTS The 14 gastrostomy subjects (five males) had a mean age of 12.9 years • 4.6 SD (range 4.9 to 21.5 years). Weight as a percent of standard was 82% • 10% (range 65% to 104%). The individual with a weight as a percentage of ideal of 104% was stunted. Mean percentage of body weight as fat was 7.0% in males and 15% in females. Average values for normal children of the same age are 11% for males and 19% for females? 3 Pregastrostomy height velocity was 74% • 21% of standard, and weight velocity was 28% _+ 24% of standard. Initial FVC was 45% • 14% predicted, and FEV~ was 41% • 16% predicted.
228
Levy et al.
The Journal of Pediatrics August 1985
Table II. Complications following gastrostomy (in order of importance) Complication
Treatment
Early (0 to 1 month) Early morning fullness and anorexia, nausea
Usually none; generally resolves with time; metoclopramide useful in some refractory cases Move T-bar to prevent skin chafing as weight gain occurs Clotrimazole cream with or without hydrocortisone Vigorous QID physiotherapy from the day of insertion for prevention; early ambulation Seen in two of 21 patients; resolved with orally administered cloxacillin
Local tenderness Monilial infection Chest infection (0 to 7 days after insertion) Local cellulitis Late Local proliferation of granulation tissue Accidental tube removal
Cautery with silver nitrate Tube replaced and position checked with gastrografin x-ray study
Table IIl. Changes in pulmonary function and in weight/height in gastrostomy patients and in comparison group FVC (% predicted)
FEV~ (% predicted)
Weight~height (% standard) Tilq'le *
Patients (n = 10) Difference P Comparison group
Initial
Fina'l
Initial
Final
Initial
Final
(yr)
53 + 13
49 _+ 16
47 _+ 15
42 _+ 17
85 • 12
87 + 14
1.1 -+ 0.43
61 _+ 20
-4.0 NS 49•
NS 81+10
1.01 _+ 0.24
-5.0 56+21
+2.0 NS 43•
8429
(n = lO)
Difference P
-12 <0.0I
-13 <0.01
-3.0 <0.02 P <0.05t
Values are mean _+SD. lntragroup comparison by Student paired t test. *Time elapsed from gastrostomy or gastrostomy age in matched comparison subjects. ~'Intergroup statistical comparison (unpaired t test) for changes in weight/height (% standard).
Effects of gastrostomy feeding. Significant increases in percent body fat, fat-free mass, TBK, and TBN occurred in the gastrostomy patients (Table I). The increase in TBK was matched by an increase in total body water, which resulted from growth of the intracellular water compartment. Most notably, extracellular water did not increase with gastrostomy feeding. A linear relationship between TBK and intracellular water was found (r = 0.68, P <0.025). Postgastrostomy height velocity increased, showing catch-up growth from a mean of 74% of standard before gastrostomy to 162% of standard in the first 0.81 year of supplemental feeding. By a mean time of 1.2 • 0.7 years this rate of growth had increased further to 238% of standard. Weight velocity showed a similar pattern of rapid initial increase, and did not reach a plateau until a subject approached ideal weight for height. Body fat increased a mean of 3.0% _+ 2.0% (P <0.005). A significant improvement in mean hemoglobin occurred during the initial 3 months of nutritional rehabil-
itation, from 128 _+ 14 g m / L to 138 +_ 11 g m / L (P <0.005), by which time, hemoglobin levels had stabilized. Serum sodium and albumin concentrations were within the normal range, and did not change significantly. Gastrostomy tube complications were generally mild and occurred either immediately after insertion or much later (Table II). Comparison of subjects and controls. For each gastrostomy-fed patient, there were at least four and as many as 20 patients with the same sex and birth year being followed in our clinic. In only one patient were pregastrostomy values similar to mean values for his sex/birth year control group; all others had markedly lower than average CF clinic values for weight, height, and FEV~. No gastrostomy patient had more than one control in the same birth year matched for weight for height and FEV~. Controls for two female gastrostomy patients came from an adjacent birth year. No match was found for the three youngest patients, all girls, whose weight for height and FEV~ (when mea-
Volume 107 Number 2
Malnutrition and pulmonary function in cystic fibrosis
I INFECTION
1-- .........
"
"~ I
[
l DETERIORATING
I
LUNGFU}NCTION I
229
.t .
[
WEIGHT LOSS
Figure. Model defining interrelationship between deteriorating lung function and nutritional status in patients with cystic fibrosis.
sured) were poorer than all other birth year or adjacent birth year control m~itches~ Comparison of pregastrostomy and follow-up values for FVC, FEVb and weight as a percent of standard for the 10 patients with CF and for their controls showed that pulmonary function in t h e gastrost0my subjects did not change significantly over the 0.8 to 2.8 years, whereas it declined in the controls (P <0.01). Mean weight for height in the 10 gastrostomy patients increased by 2%, but declined by 3% in the controls (P <0.05). Deaths during the study, Six patients died during the 3-year period to December 1, 1984. Three patients, who were our youngest (ages 4,9, 5.8, and 8.9 years) and the most seriously affected in terms of lung disease, did not have matched controls. The other three did have matches within the comparison group: a man aged 18.7 years at the time of gastrostomy survived 1.2 years longer than his match; a woman aged 17.4 years at the time of gastrostomy survived 1.8 years longer than her match; and a man aged 21.5 years at the time of gastrostomy lived 0.6 year after gastrostomy, whereas his match lived 0.8 year after gastrostomy age. These three older patients died of respiratory failur e secondai'y to acute pneumonia caused by Pseudomonas cepacia, which is resistant to all known antibiotics. One of the matches died before pulmonary function tests could be repeated, reducing the number of pairs from 11 to 10 (Table III). DISCUSSION We have Proposed a model ~4 (Figure) that defines the interrelationship between deteriorating lung function and nutritional status in patients with CF. Other workers 2' ~5.~6 have attempted to break this apparent cycle of progressive weight loss and deteriorating lung function by providing short-term nutritional support. None of these studies showed lasting improvements in pulmonary function or body composition. We have demonstrated that long-term nutritional Supplementation will produce catch-up and sustained growth in a group of patients with CF. The
prolonge d improvement noted in weight as a percent of ideal, and the stabilization in pulmonary function status are significant when compared with those of a matched comparison group. It is possible that several patients had an increase in growth related solely to the start of puberty. However, our experience with these and other patients with CF has shown that growth velocity slows when nutritiona! support is discontinued and malnutrition recurs. The increase in weight for height did not, i n most cases, result in a normalization of weight as a percentage of standard for height, age, and sex. It is probable that several as yet imprecisely understood patient-reported problems led to a reduction in voluntary food intake, so that weight gain was suboptimal. These problems include a subjective sensation of fullness and hence a decrease in appetite in the morning. Six patients with marked lung disease remained anorexic after gastrostomy, or continued to complain of early satiety so that daytime meals were small in size. Body composition changes occurred with a moderate recovery of adipose reserve and lean tissue mass. Initially, we were concerned that weight gains in these patients could represent merely increases in extracellular water, with the risk of compromising gaseous exchange and cardiac function. Our data show that weight gain early in the course of refeeding resulted principally from increases in fat and lean tissue rather than in extracellular water. A moderate anemia is quite common in patients with CF~7; the effects of chronic illness and infection on the hematopoietic system may in part be the causes. However, two additional factors, iron deficiency~7.~ and protein malnutrition, ~9need to be considered. Our patients had a significant rise in hemoglobin during the refeeding period despite the fact that only two received therapeutic doses of iron. We suggest, therefore, that nutritional support reversed the effects of malnutrition on the hematopoietic system. It is possible that the anemia previously accepted as inevitable in unwell and often malnourished patients with CF may be improved by refeeding.
230
Levy et al.
Pulmonary function data revealed a significantly slower rate of decline in FVC and FEV. in the gastrostomy group. Consistent with this. all gastrostomy subjects reported an increased ability to participate in activities of daily life. If such participation resulted in more physical exercise, this may have had the effect of "training" respiratory musculature2 ~ Whether nutritional rehabilitation materially affects the severity of acute chest infections is unknown. The exact mechanism of how nutritional support benefits lung function remains to be resolved. We were aware of the possible effects of the gastrostomy program on psychosocial functioning of the patient and family. None of the C F clinic team of physicians, nurses, social workers, or psychiatrist were able to identify any deterioration in patient or family function: rather, the converse was observed We set out to investigate the effectiveness of the gastrostomy program as a means of treating malnutrition in patients with CF with advanced lung disease who were unresponsive to intensive dietary counseling and supplementation. The question remains whether the treatment of malnutrition affects life quality and survival. Ultimately, these questions can only be answered by randomized control study. Presumably, the control group would continue to receive "optimal" dietary counseling and supplementation. Iriasmuch as all our patients had been unresponsive to at least three months of such dietary therapy before being offered gastrostomy, we believe that there may be serious practical and ethical problems in designing and conducting a truly randomized study. Based on the results of this study, we propose that the long-term maintenance of adequate nutrition in selected malnourished patients with C F can safely and practically be done using nocturnal gastrostomy feeding. Furthermore, we have evidence that nutritional support may delay a decline in lung function significantly, and that this may translate into improved health, quality of life, and length of survival in some patients with CF. We thank Dr. Henry Levis0n (Director), Dr. Gordon Forstner, and the staff physicians of The Hospital for Sick Children Chest and CF Clinic for support and encouragement; Dr. Robert Filler and Dr. David Wesson for insertion of the gastrostomy tubes; Dr. Joan Harrison (Director, Medical Physics Laboratory, Toronto General Hospital) for the body nitrogen and potassium measurements; Dr. R. Hancock (SLOWPOKE Reactor, University of Toronto) for help with the bromide analyses of extracellular water; the CF ward and clinic nursing staff; the CF dietitiannatritionists for help with the dietary assessments; Ruth Clarke and Linda Wykes for technical assistance; and Heather Keillor for manuscript preparation.
The Journal of Pediatrics August 1985
REFERENCES
I. Gurwii.z D Covey M, Francis PWJ, et al: Perspectives in cystic fibrosis. Pediatr Clin North Am 26:603, 1979. 2. Gaskin K, Gurwitz D, Durie P, et al: Improved respiratory prognosis in patien{s with normal fat asorption. J PEDIATR i00;857, 1982. 3. Levison H, Cherniak R: Ventilatory cost of exercise in chronic obstructiye pulmonary disease. J Appl Physiol 25:2l, 1968. 4. Shizgal HM: The effect of malnutrition on body composition. Surg Gynecol Obstet 152:22, 1981. 5. Tanner JM, Whitehouse RH, Takaishi M: Standards from birth to maturity for height, Weight, height velocity and weight velocity: British children, 1965. ll. Arch Dis Child 41:613, 1966. 6. Archibald EH, Harrison JE, Pencharz PB: Effect of a weight-reducing high-protein diet on the body composition 0f obese adolescents. Am J Dis Child 137:658, 1983. 7. Schoeller DA, Van Santen E, Peterson DW, et al: Total body watei measurement in humans with zsO and 2H labeled'water. Am J Clin Nutr 33:2686, 1980. 8. Holzbecher J, Ryan DE: The rapid determination of total bromine and iodine in biological fluids by neutron activation. Clin Biochem 13:277 1980. 9. weng T-R, Levison H: Standards of pulmonary function in children. Am Rev Respir Dis 99:8791 1969. 10. Cystic Fibrosis "GAP" Conference Report: Standardization of lung ftmction testing in children. Cystic Fib?osis Foundation, 1978. 11. Gauderer MWL, Ponsky JL, lzant JR: Gastrostomy without laparotomy: A percutaneous endoscopic technique. J Pediatr Surg 15:872, 1980. 12. Recommended nutrient intakes for canadians: Dietary Standard for Canada. Ottawa, Ont., Canada, Bureau of Nutritional Sciences, Food Directorate, Health Protection Branch, Department of National Health and Welfare. 13. Widdowson EM: Changes in body proportions and composition during growth. In Davis JA, Dobbing J, editors: Scientific foundations of paediatrics. London, 1981, William Heinemann Medical Books Ltd., pp 330-342. 14. Pencharz PBI Energy intakes and low fat diets in children with cystic fibrosis. J Pediatr Gastroenterol Nutr 2:400, 1983. 15. Berry HK, Kellogg Fw, Hunt MM, et aI: Dietary supplement and nutrition in children with cystic fibrosis. Am J Dis Child 129:165, 1975. 16. Mansell AL, Andersen JC, Muttard CR, et al: Short-term pulmonary effects of total parenteral nutrition in children with cystic fibrosis. J PEDJATR104:700, 1984. 17. Ater JL, Herbst J J, Landau SA, et al: Relative anemia and iron deficiency in cystic fibrosis. Pediatrics 71:810, 1983. 18. Vichinsky EP, Pennathur-Das R, Nickerson B, et ah Inadequate erythroid response to hypoxia in cystic fibrosis. J PEDIATR 105:15, 1984. 19. Lindenbaum J: The hematopoietic system. In Hodges RE, editor: Human nutrition, vol 4. New York, 1979, Plenum Press, pp 1-52. 20. Cesny F J, Pullano TP, Cropp GJA: Cardiorespiratory adaptations to exercise in cystic fibrosis. Am Rev Respir Dis 126:217, 1982.