Caffeine, Fluid-Electrolyte Balance, Temperature Regulation, and Exercise-Heat Tolerance

Lawrence E. Armstrong; Douglas J. Casa; Carl M. Maresh; Matthew S. Ganio

Disclosures

Exerc Sport Sci Rev. 2007;35(3):135-140. 

In This Article

Fluid Balance

Diuretics are substances that promote the formation of urine within the kidneys. Water, cranberry juice, and alcohol are weak diuretics. Caffeine (in coffee, tea, and soft drinks) and theophylline in tea stimulate increased renal glomerular filtration and inhibit reabsorption of sodium (Na+) within nephrons, thereby stimulating an increased Na+ and water excretion. The diuretic effect of caffeine at rest has been recognized since 1928.[10] Although this landmark study has influenced practitioner beliefs for decades, it involved only three subjects, and the observations spanned only a few hours.

When physicians, physiologists, and dietitians recommend that caffeinated beverages not be consumed before or during exercise,[16,28] they rarely cite research findings to support their assumptions. Opposing this viewpoint, other professionals view caffeine as a mild diuretic that poses no harm to health or exercise performance.[7,9,26] In fact, three previously published review articles[2,20,21] concluded that caffeinated beverages and water affect body water balance similarly during exercise.

Table 1 considers the diuretic effect of caffeine versus a control fluid (i.e., water or placebo) as published in 15 different research studies; seven involved exercise. Specifically, this table considers the amount of caffeine and the volume of fluid consumed; all investigations are ranked on the basis of caffeine dose. All studies in which caffeine doses were less than 226 mg (see rows 16-23 in Table 1 ) reported no statistical difference. As shown in column 3, a significantly greater within-study diuresis occurred in 6 of 23 total experiments when caffeine (vs control) was consumed at doses ranging from 240[25] to 642 mg.[22] These doses are equivalent to approximately 1.5 to 4.3 cups of brewed coffee (150 mg of caffeine/150 mL) and approximately 5 to 13 cans of cola soft drink (50 mg of caffeine/360 mL). No difference was reported in 17 experiments ( Table 1 ), when up to 553 mg of caffeine was consumed. Furthermore, there seems to be no relationship between the volume of fluid consumed (column 4) and the appearance of a significant diuresis because of caffeine (column 3).

The duration of observations (i.e., 1 h to 11 d) is important when considering column 1 because 73% of these 15 studies involved acute measurements (1 h to 6 h). Brief observation periods may result in considerably different conclusions than longer periods (16 h to 11 d) because human water balance is challenged by diuresis several times each day. Acute diuresis per se must be distinguished from chronic depletion of body water. That is, diuresis subsequent to caffeine consumption (column 3) does not necessarily mean that 24-h or weekly total body water turnover results in dehydration or hypohydration; neuroendocrine mechanisms regulate fluid volume successfully as indicated by plasma osmolality across a wide range of 24-h fluid intakes (i.e., 1.7-7.9 L·d-1[23]) and urine volumes. Indeed, the only investigations in Table 1 to observe chronic effects of caffeine consumption[3,12] found no indication of dehydration in numerous indices of hydration status. Thus, although some experiments reported that caffeine caused diuresis exceeding that of water, a) no evidence of chronic dehydration exists in the scientific literature, and b) restricting dietary intake of caffeine is not scientifically and physiologically supported. This is compounded by the lack of a universally accepted "gold standard" that serves to assess hydration state.

Figure 1 provides another method to analyze the influence of caffeine on body water balance. If the ingested volume of a fluid exceeds the resulting urine volume, the difference represents the volume that is retained by the body during the observation period. This figure depicts water retention as a positive value and net fluid loss as a negative value. These values differ considerably among studies because of differences of experimental protocols and fluid volumes ingested. When caffeinated beverages (gray bars) are compared with control fluids (e.g., water or placebo (black bars)), the mean values for fluid balance (fluid intake-urine volume) are not statistically different (mean ± SD: caffeine, 994 ± 1839 mL; placebo, 1068 ± 1707 mL); if the 72-h study by Fiala et al.[12] is removed from Figure 1, the values remain statistically similar (P > 0.05), but the means and SD are smaller (caffeine, 511 ± 625 mL; placebo, 624 ± 618 mL). We interpret this to mean that caffeinated beverages, in general, contribute to euhydration similarly to water.

Figure 1.

Mean fluid retention (above dashed line) when acutely consuming a control fluid (placebo or water) vs a fluid containing caffeine or (below dashed line) when 24-h fluid intake (beverages + water content of food) contains caffeine vs no caffeine. Negative values depict studies that reported a net fluid loss. CAF indicates caffeine; CON = control fluid; E = exercise; R = rehydration after exercise.

For each experiment, the reader should note that the duration of observations is labeled in hours (h) or days (d) and that observations of exercise (E) and rehydration (R) are annotated at the end of each bar. Thus, only four studies involved observations greater than 4 h, three studies observed fluid turnover during exercise, and three studies observed the effects of rehydration after exercise. These underpowered categories do not allow conclusions to be drawn about specific time, exercise, or rehydration matters.

The reader also should note that the hydration state, and the amount of fluid consumed during the previous 12 h, of test subjects was generally not controlled in these studies. This suggests that the negative values in Figure 1 (i.e., representing a net fluid loss) result from overhydration before experiments began, because even placebo trials resulted in a net fluid loss.

processing....