Article Text

  1. G. R. Merriam,
  2. M. Kletke,
  3. S. Barsness,
  4. M. V. Vitiello
  1. Tacoma and Seattle, WA.


Ghrelin is a potent growth hormone (GH) secretagogue as well as an orexigen. Enterically secreted ghrelin circulates in concentrations sufficient to stimulate GH, but it is not known whether it contributes significantly to physiologic GH secretion. We have previously shown (Cummings et al., AFMR 2003) that GH secretion is normal after bypass surgery despite reduced ghrelin levels, and that ghrelin and GH levels did not seem linked in relatively brief sampling profiles; but the latter series were too brief to be conclusive. Since GH and ghrelin both respond to meals, even if there is an appearance of linkage, it is possible that some or all might represent cueing of both by meals, rather than direct internal synchronization. We therefore studied the correlations between GH and ghrelin levels in a more extensive sampling protocol in older men.

Methods Healthy men aged 60-85 were placed on an isocaloric diet and admitted to the UW GCRC. Meals were served at standardized times. After habituation, samples were drawn q 20 min for 24 h and assayed for GH and total immunoreactive ghrelin. Profiles were plotted together and analyzed in two ways—by standard parametric cross-correlation analysis, and by non-parametric scoring of synchronous and near-synchronous occurrence of peaks identified by computer algorithm in the two profiles—in both the full 24 h series, and in 12 h subseries including (7am-7pm) or excluding (7 pm-7am) mealtimes. These analyses were performed both with GH and ghrelin levels drawn from the same subject (“matched” series), and with GH from one subject and ghrelin from a different subject (“mismatched” series). These latter pairings would be expected to retain linkages due to similar schedules, but not those due to direct internal synchronization.

Results “Matched” daytime series showed significantly higher correlations (r=0.46) than nighttime series (r=0.16, p≤0.007). Correlations in “mismatched” series were lower both by day (r=0.16) and by night (r=0.07; ns for day vs. night). There was also a much higher temporal synchronization of pulses in daytime vs. nighttime profiles, but little overall difference between matched vs. mismatched series. This suggests that daytime events common to all subjects accounted for most of the observed synchrony.

Conclusions These results show that circulating profiles of GH and ghrelin exhibit substantial temporal correlation as analyzed by two independent methods, but that much of this correlation may be attributed to external cueing, such as by meals, rather than to internal mechanisms of synchronization.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.