Breath tests (BT) using stable isotopically labelled substrates seem t
o fullfil all the demands and desires for a non-invasive investigation
. There are no radiation hazards, substrates are given in tracer amoun
ts perorally, breath and urine samples can be collected easily, and te
sts can be done repeatedly, thus easily allowing the monitoring of fun
ction with time. There are, however, some disadvantages. Any BT has th
e same assumption: after intake of the C-13 tracer the substrate is me
tabolized to (CO2)-C-13. An increase of (CO2)-C-13 above baseline leve
ls is said to reflect the function investigated-in C-13 sucrose studie
s, the amount of carbohydrate absorbed; in C-13 aminopyrine BT, the li
ver function; in C-13 glucose BT in a diabetic child, the impaired han
dling of glucose. However, as only the end product (CO2)-C-13 is measu
red, there is no information on all the pools and fluxes the labelled
substrate and its metabolites have to pass. At least in inborn errors
of metabolism, probably in any disease, one has to assume that these f
luxes and pools are substantially changed. Therefore all calculations
are weak and finally one has to resort to invasive methods, i.e. drawi
ng blood to measure pools and fluxes to allow a correct interpretation
of the BT data. Furthermore, changes in the basal exhalation of (CO2)
-C-13 during the test will have an impact on the BT calculation. Anoth
er problem is that for an exact calculation, the basal metabolic rate
(BMR) and the actual endogenous CO2 production in the patient is neede
d, which in most instances is unknown. It is not easy to maintain a st
able endogenous CO2 production, particularly in younger children who w
ill not rest or in neonates and toddlers who may fight against taking
breath samples. Taking together these limitations are the reason why B
T have not been able to reach the level of routine clinical methods, e
specially in the diagnostic work up of impaired liver function or inbo
rn errors of metabolism.