N. Custro et al., DEFICIENT PULSATILE THYROTROPIN SECRETION IN THE LOW-THYROID-HORMONE STATE OF SEVERE NONTHYROIDAL ILLNESS, European journal of endocrinology, 130(2), 1994, pp. 132-136
Twenty-four-hour thyrotropin (TSH) profiles in eight severely ill pati
ents were compared with those of six healthy subjects. The profiles we
re assessed using the cosinor method to evaluate circadian variations
and using the Pulsar algorithm to analyze episodic secretion. In the n
ormal subjects, the typical periodicity of TSH secretion showed a mean
level in the rhythm (mesor) of 2.03 mU/l. The amplitude (half the ext
ent of rhythmic change in the cycle) was 0.58 mU/l; the acrophase (the
delay from midnight (0 degrees) of the highest level in the rhythm) w
as - 9.9 degrees. In contrast, severely ill patients showed only sligh
t and anticipated elevations of serum TSH levels (mesor 0.93 mU/l, amp
litude 0.22 mU/l, acrophase + 82.4 degrees). Moreover, whereas the epi
sodic TSH secretion in healthy individuals consisted of 5-8 pulses/24
h, mainly clustered around midnight, only one pulse of reduced amplitu
de was detected in two of the eight severely ill patients and no pulse
s in the other six. Since earlier studies have indicated that the loss
of TSH pulsatility is associated with the relative insensitivity of t
he thyrotrophs to low thyroid hormone levels and our analytical proced
ures have demonstrated that 24 h pulsatile pattern of TSH closely over
lapped with baseline TSH secretion, it seems reasonable to assume that
low-thyroid-hormone state, deficient pulsatile TSH secretion and alte
red nyctohemeral TSH periodicity do not coincide by chance, but that t
here is a causal relationship between such abnormalities in severely i
ll patients.