Methodological and statistical considerations for exercise-related hormone evaluations.

Improvements in laboratory techniques have allowed research related to exercise endocrinology to flourish. The emerging literature, however, is often inconsistent and contradictory. The discrepancies in research findings are possibly the result of poor control of confounding variables and/or inappropriate methodologies or analyses. Environmental and pretesting behavioural conditions must be standardised to minimise the influence of variables not directly related to the investigation. Environmental temperature and relative humidity, alcohol, caffeine and nicotine intake, prandial state, sleep deprivation and previous exercise can each alter hormonal responses to exercise. Both prescription and over-the-counter medications can also modify normal hormonal secretions thereby confusing exercise-induced findings. Specimen collection and analysis procedures must be controlled carefully. Changes in plasma volume related to postural changes or tourniquet-induced stasis can confound attempts to isolate exercise-related endocrine responses. The established circadian and rhythmical variations characteristic of many hormones need to be controlled. The specimen selection (plasma, serum, urine, etc), collection, storage and analysis procedures should be carefully planned and evaluated. The magnitude of haemolysis, analytical and biological variation must also be monitored. Isolating the hormonal perturbations resulting from a particular exercise variable can be very difficult. Exercise intensity, duration, mode, frequency and volume may each have specific effects on the endocrine changes seen with exercise and training. Furthermore, hormonal responses to exercise are dependent upon initial training status and fitness level. The statistical procedures and data presentation options selected to convey experimental findings can bias experimental results. The descriptive and inferential statistics to be used for data analysis should be preplanned and consistent with the underlying assumptions of the analytical procedure. Careful consideration should be given to the biological relevance of statistically significant findings. In some cases, data transformations (e.g. absolute vs relative changes, logarithmic) should be considered for analysis or presentation. Given the individual nature of hormonal responses to exercise, emphasis should be placed presenting individual data. Other considerations, including age, sex, racial origin and disease conditions need to be controlled for when trying to examine exercise-induced hormone changes.