|
Estimations: Plasma
and urinary electrolytes,
phosphate, uric acid and creatinine were measured by routine methods (available
on request). Blood glucose was estimated by glucose oxidase. Magnesium in both
plasma and urine was measured by atomic absorption spectrophotometry (Fleming
& Stewart, 1966).
Results: Body
weight loss: During the 382 days of
the fast, the patient's weight decreased from 456 to 180 lb. Five years after
undertaking the fast, Mr A.B.'s weight remains around 196 lb.
Blood glucose levels; Blood
glucose concentrations decreased systematically (Fig. 1) and remained around 30
mg/100 ml from the fourth month onwards. Values below 20 mg/ 100 ml were
occasionally seen towards the end of the fast. [This explains his sometimes
eating small amounts towards the end of the fast. Despite the hypoglycaemia
the patient
remained symptom-free, felt well and walked about normally. Tests of glucose
tolerance showed unimpaired capacity for glucose uptake (Table 1). Apart from Day
7, the glucose assimilation coefficient remained between 10 and 1 6 (low
normal) throughout the fast… the 10 min increment in blood glucose remained
between 62 and 75 mg/ 100 ml until refeeding, when it became greater than 100
mg/100 ml. Responses to the intravenous tolbutamide test also remained normal
until Day 355, brief episode of fainting immediately after the intravenous
injection. Responses to intravenous glucagon were reduced from the early days
of the fast onwards and by Day 269 and subsequently, there was no responsive
increment in blood glucose at all. A normal increase was seen by Day 7 of the
refeeding period, at which time the only sustenance had been entirely
carbohydrate in nature (1000 kcal daily as modified liquid glucose BPC,
'Hycal'-Beechams).
Plasma Concentrations: Over
the first 4 months there was a systematic decrease
in plasma potassium concentrations but the small daily supplement of 13 mEq
potassium, which was then given, increased the potassium level. After a
further 10 weeks, no more potassium
was given, and plasma potassium remained normal. . For 6 months towards the
end of the fast
there was a limited period of hypercalcaemia
(10-8-11 6 mEq/l) but this remitted spontaneously during the final month.
[Reflects bone loss through remodeling and the loss of excess bone as weight
was reduced.] As expected, plasma urea
decreased during the first 2 weeks of fasting and thereafter remained steady
between 15 and 20 mg/100 ml. Plasma uric
acid, which was high at 12-6 mg/100 ml prior to the fast (possibly
reflecting pre-admission attempts at fasting), remained generally at this level
throughout with one or two brief increases to 17 mg/100 ml which were not
sustained [Possible reflecting muscle
loss. I think a protein shake should
have been given.] Symptomatic gout did
not occur. Plasma magnesium concentrations
(Fig. 2) decreased over the first few
weeks and thereafter stabilized between 1-2 and 1-4 mEq/l. Erythrocyte magnesium
concentration remained
unchanged and normal.
Urinary electrolyte excretion (Fig. 3): The excretion of sodium, potassium, calcium and inorganic phosphate decreased to low
levels throughout the first 100 days, but thereafter the excretion of all four
urinary constituents, as well as of magnesium, began to increase. During the
subsequent 200 days sodium excretion, previously between 2 and 20 mEq daily,
reached over 80 mEq/24 hr, potassium excretion increased to 30-40 mEq daily and
calcium excretion increased from 10-30 mg/24 hr to 250- 280 mg/24 hr. Magnesium
excretion (which was not measured during the first 100 days) reached 10 mEq/ 24
hr between Days 200-300. Phosphate excretion, which had decreased to under 200
mg/24 hr, also increased to around 800 mg/24 hr, even exceeding 1000 mg/24 hr
on occasion. Peak excretions of all these constituents were seen around Day
300, after which there was a marginal decrease, but excretion remained high. There
was no indication in the
characteristics of the weight-loss response, which was steady and sustained, or
in the persistently low blood glucose or urea levels, which would have
suggested that he had been surreptitiously eating.
Discussion: Body
weight loss and length of fast; The
amount of weight lost and the rate of loss were not strikingly different from
that of an earlier patient (Stewart, Fleming & Robertson, 1966) who reduced
his weight from 432 to 235 lb during 350 days of intermittent starvation. The
amount of weight lost by Mr A.B. was 276
lb, with an average rate of loss of 0-72 lb/day, comparing favourably with the
rates of weight loss in other long-term fasts (> 200 days) which have ranged
from 0-41 lb/day (Thomson, Runcie & Miller, 1966) to 0 67 lb/day (Runcie
& Thomson, 1970). The nearest comparable fasts are one of 256 days
(Collison, 1967, 1971), those of 249 and 236 days (Thomson et al., 1966) and
two of 210 days (Garnett et al., 1969; Runcie & Thomson, 1970).
There have been reports of five fatalities coinciding with the
treatment of obesity by total starvation (Cubberley, Polster & Schulman,
1965; Spencer, 1968; Garnett et al., 1969; Runcie & Thomson, 1970). One was
attributed to lactic acidosis during the refeeding period following a 3 week
fast (Cubberley et al., 1965). Two were considered to be due to ventricular
failure, occurring during the fast, at 3 and 8 weeks respectively in patients who
had shown evidence of heart failure before beginning the fast (Spencer, 1968).
One patient (Runcie & Thomson, 1970) died on the thirteenth day of his fast
from small bowel obstruction. Only one of the five 'fasting' deaths has been
associated with a fast of more than 200 days' duration. It occurred during the
refeeding period after a fast of 210 days in an apparently well young woman
(Garnett et al., 1969). Following this particular report doubt has been cast on
the safety of the treatment of obesity by total fasting (Garnett et al., 1969;
Rooth & Carlstrom, 1970). However, the allopurinol[1]
which had been given may have had unfavourable effects on nucleotide metabolism
(Stewart & Fleming, 1969).
Blood glucose changes: Decreased
(Jackson et al., 1968; Runcie & Thomson, 1970), increased (Schless &
Duncan, 1966) and normal (Anderson, Herman & Newcomer, 1969) glucose
tolerance have all been described as occurring during fasting. The
categorization of obese patients into 'diabetic' and 'non-diabetic' groups, even
although none was overtly diabetic prior to fasting, is according to some a
prerequisite in the evaluation of glucose tolerance during fasting, with the
'non-diabetic' showing decreased tolerance and the 'diabetic' showing unaltered
or improved tolerance (Schless & Duncan, 1966; Jackson et al., 1968). When
assessed by the rate of decrease in
glucose concentrations following intravenous glucose administration as
recommended, our 'non-diabetic' patient showed no systematic change in glucose
tolerance throughout the fast or afterwards. It would appear that there is
considerable individual variation, both in the level of blood glucose
maintained during prolonged fasting and in the response to a glucose load.
However, it is obvious that even patients such as ours, who do become markedly
hypoglycaemic, can tolerate their fasts as well as those who maintain more
normal sugar levels.
Plasma electrolyte changes Progressive hypokalaemia was not a
feature in our patient following the cessation of potassium supplementation.
Two of twenty-five patients have developed hypokalaemia, despite potassium
supplements (Munro et al., 1970). It may be that the release of potassium from
cellular tissue during fasting of the obese can be sufficient in some cases to
maintain plasma levels despite continued urinary loss. The temporary period of
hypercalcaemia is difficult to explain. It was unassociated with any other
evidence of parathyroid hyperactivity and coincided with a slight increase in
plasma phosphate concentration, within the normal range. The coincidence between
the hypercalcaemia
and the occurrence of markedly increased excretion of both calcium and
phosphate is impressive. It may be that the hypercalcaemia was associated with
resorption of some bone salt which was no longer required in the skeleton to
support the former excess weight. Hypercalcaemia,
to our knowledge, has not been remarked in any of the other long-term
therapeutic fasts.
Plasma magnesium changes: Decreased
plasma magnesium concentrations have not previously been reported during
prolonged fasting. On the contrary normal plasma magnesium concentrations,
despite magnesium 'depletion' in muscle tissue, have been described (Drenick et
al., 1969) during short-term fasting (1-3 months). The only other relevant
report is a remark (Runcie & Thomson, 1970) that one patient who fasted 71
days had a normal plasma magnesium level of 2-2 mEq/l at the time when she
developed latent tetany. [Tetany or tetany seizure is a medical
sign consisting
of the involuntary contraction of muscles,
Can be cxaused by low levels of calcium, carbon dioxide, magnesium,
excess potassium.]
Urinary excretion patterns:
A breakdown in the renal control of electrolyte excretion has been
reported to occur in some patients after 40 or 100 days during prolonged
fasting (Runcie & Thomson, 1970). The increased urinary excretion of not
only sodium and potassium but also calcium, magnesium and phosphate, noted in
our patient, may be of similar origin. The concurrent development of these
losses represents a definite change in the pattern of renal excretion after
about 100 days of fasting. A similar increase in the urinary excretion pattern
of sodium, potassium, calcium, phosphate and magnesium was noted during
protracted intermittent starvation in a previous patient (Stewart et al.,
1966). It is conceivable that some specific deficiency, perhaps of potassium or
magnesium, may have brought about this 'nephropathy' but its real nature is at
present conjectural. The concentrations of plasma monovalent cations as well as
phosphate, appeared to be essentially stable. Once the long-term fasting levels
were established, plasma magnesium, glucose and urea levels also remained
stable. Apart from the plasma calcium and bicarbonate concentrations, there
were no other alterations in plasma constituents coincidental with the change
in urinary excretion. The overall stability of the plasma constituents studied
therefore contrasts vividly with the concurrently developing increase in the
urinary losses of both monovalent and divalent cations, together with inorganic
phosphate which occurred after about 100 days of fasting. Such symptom-free
developments are possibly
more compatible with the release of surplus structural or other tissue
components than with disadvantageous decompensatory loss of electrolytes from
the extracellular fluid occurring as a result of some kind of nephropathy. It
has been shown (Naeye, 1969) that hyperplasia and hypertrophy, not only of
adipose tissue, but also of heart, kidneys, pancreas, liver and spleen occurs
in obese subjects. It is suggested that the increased excretions described may
have originated in dissolution of this soft tissue and associated skeletal
excess. Accordingly, the electrolyte losses described are not regarded as
representing a necessarily disadvantageous development during fasting.
Conclusions:
Short-term fasts, although
demonstrating to the obese patient his ability to lose weight, have a poor
long-term outlook with respect to subsequent weight gain (MacCuish et al.,
1968). We have found, like Munro and colleagues (1970), that prolonged
supervised therapeutic starvation of the obese patient can be a safe therapy,
which is also effective if the ideal weight is reached. There is, however,
likely to be occasionally a risk in some individuals, attributable to failures
in different aspects of the adaptative response to fasting. Until the
characteristics of these variations in response are identified, and shown to be
capable of detection in their prodromal stages, extended starvation therapy
must be used cautiously. In our view, unless unusual hypokalaemia is seen,
potassium supplements are not mandatory. Xanthine oxidase inhibitors (or
uricosuric agents) are not always necessary and could even be potentially
harmful (British Medical Journal, 1971) perhaps particularly in the long-term
fasting situation. In most obese
patients, there exists a weight loss 'barrier' representing a balance between
the desire to lose weight on the one hand and the desire to eat on the other.
When the first desire continues to overcome the second, the prolongation of
therapeutic fasting should not be ruled out. The various biochemical features
which developed in our patient are not considered to be expressions of a
failure in adaptation. Starvation therapy can be completely successful, as in
the present instance.
|
