Appetite 87 (2015) 76–80
Contents lists available at ScienceDirect
Appetite j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p p e t
Research report
The development of the predisposition to dehydration questionnaire David Benton *, Hayley Young, Kimberley Jenkins Department of Psychology, University of Wales Swansea, Swansea, Wales SA2 8PP, UK
A R T I C L E
I N F O
Article history: Received 1 September 2014 Received in revised form 14 November 2014 Accepted 24 November 2014 Available online 2 December 2014 Keywords: Dehydration Drinking Mood Osmolality Sweating Water consumption
A B S T R A C T
The role played by hydration in general health and well-being is an emerging public health issue, yet there are few tools available to monitor its status in large populations. The aim was therefore to develop a questionnaire that assesses individual differences in the tendency to lose body fluid in a warm environment and hence become dehydrated. Fifty-three subjects sat in a room at 30 °C for four hours and changes in mood and measures of hydration were monitored. There were marked individual differences in the loss of body mass that differed from 0.24% to 2.39%. Females who reported habitually drinking a lot had more water in their diet and at baseline the osmolality of urine was lower. After being subject to heat, those who reported habitually drinking more produced more urine, had a lower urine osmolality at the end of the study, and overall more body mass was lost. Females who reported that they responded badly to heat were more confused, unsure and depressed after four hours at 30 °C. In males those reporting that they habitually drank to a greater extent had more water in the diet, and also those who dealt badly with heat habitually drank more. It was concluded that particularly in females, questionnaire measures were able to predict changes in hydration that result from a warm environment. © 2014 Elsevier Ltd. All rights reserved.
Introduction The role played by hydration in general health and well-being is an emerging public health issue as water makes up 50–60% of lean bodyweight in men and 45–50% in women. Intracellular fluid makes up about 35%, extracellular fluid about 12% and plasma about 4–5% of lean bodyweight (Kumar & Clark, 2012). Water plays a role in all aspects of bodily functioning including the distribution of oxygen and nutrients, the removal of waste products, serving as a lubricant and regulating temperature. The loss of bodily water adversely affects both physical (Cheuvront, Carter, & Sawka, 2003) and psychological performance (Benton & Young, 2014), it disrupts thermoregulation and appetite and results in headaches, irritability and sleepiness. EFSA (2010) concluded that “a cause and effect relationship has been established between the dietary intake of water and maintenance of normal physical and cognitive functions”. However, to date, much of the study of the psychological correlates of hydration status has relied on taking physiological measurements in small samples, usually in those who have exercised, an approach that cannot be easily applied more generally. If hydration is to be studied in large populations there is a need for tools that can be used widely. The present study therefore considered one aspect of the topic, the measurement of individual differences in the response to a warm environment, as adequate
* Corresponding author. E-mail address: [email protected] (D. Benton). http://dx.doi.org/10.1016/j.appet.2014.11.029 0195-6663/© 2014 Elsevier Ltd. All rights reserved.
hydration is particularly important for thermoregulation (EFSA, 2010). The provision of water helps to maintain the body’s core temperature within the desirable range and both mood and cognition are influenced by changes in core body temperature (Holland et al., 1985; Wright, Hull, & Czeisler, 2002). An increased body temperature will result from both decreased sweating and a lower blood flow in the skin, both of which are associated with dehydration. Euhydration implies being in water balance; that is, the intake of water matches the amount lost. However, as there are well described homeostatic mechanisms that maintain hydration within a prescribed range, many suggest that it is improbable that minor changes in the intake of fluid will disrupt functioning (Benton, 2011). Dehydration is typically defined as the loss of a particular percentage of body mass, with a loss greater than one percent being defined as dehydration. Initially water is lost from the blood and then from the cells. A loss of body mass, greater than one percent, is associated with the enlargement of the brain ventricles, a reflection of the shrinking of brain cells (Benton & Young, 2014). A review concluded that “when dehydration reduces body mass by over 2% there are consistent reports that mood is influenced; fatigue is greater and alertness lower. The effects on cognition have been less consistent. . ..” (Benton & Young, 2014). In this context there is an interest in evaluating the water balance of populations and hence a need to develop suitable approaches. Recently, the Water Balance Questionnaire was developed by Malisova et al. (2012). This questionnaire uses a wide range of questions to estimate water intake and its loss. Water intake reflects the consumption of solid and fluid foods and the drinking of water, whereas its loss results from its excretion in urine, faeces and sweat.
D. Benton et al./Appetite 87 (2015) 76–80
In addition what is eaten or drunk, activity levels and the weather influence water balance. One factor that such an approach does not consider is individual differences in the response to a warm environment. The present study therefore related a series of questions concerning the reaction to a warm environment to the loss of bodily fluids over a four hour period spent at 30 °C.
Method Baseline data were collected from 110 students of Swansea University, of whom 49 were males and 61 females. Of these 23 males and 30 females were then exposed, without drinking, to a temperature of 30 °C for four hours. Of the females 78% were taking an oral contraceptive. They were on average 21.5 years. The average BMI of the males was 25.3 (4.4) and females 23.4 (3.6). The procedure was approved by the Swansea University Psychology Ethics Committee.
Procedure Subjects came to the laboratory at 0900 having consumed their usual breakfast. They emptied their bladder and gave a sample of urine. They were weighed, rated their mood, responded to the questions concerning the reaction to a warm environment and filled in the dietary diary. Half the subjects then remained in a room at 30 °C for four hours. After 230 minutes they were again weighed before emptying their bladder and giving a second sample of urine. Then after 240 minutes they were weighed a final time and mood was rated.
77
Body mass Body mass was measured using an electronic scale (Kern KMSTM, Kern and Sohn GmbH, Germany) that, to avoid problems associated with movement, took 50 assessments over a 5 second period and produced an average. It was sensitive enough to weight to 5 grams and could pick up over short periods changes in body mass due to breathing and perspiration. Subjects were weighed on arrival and after 230 minutes. Finally having emptied the bladder they were weighed for a final time after 240 minutes. Changes in mass from baseline to 230 minutes reflected water loss largely due to perspiration and breathing, and changes from 230 to 240 minutes reflected urine production. Preliminary studies examined the possible need to weigh individuals naked, as it was possible that some weight loss might be masked if perspiration remained in clothing. In the event this proved unnecessary. The percentage loss of body mass, based on 32 participants weighed naked, was 0.26% (0.05) after 230 minutes and 0.60% (0.33) following urination at 240 minutes. These values were virtually identical to the same individuals measured having worn light clothing: the comparable percentage changes were 0.26% (0.07) and 0.61% (0.35). The data presented were therefore obtained from subjects who wore the same light clothing throughout the procedure. Osmolality At the beginning and end of the procedure a urine sample was collected and the osmolality measured using an Osmomat 3000 freezing point osmometer (Gonotec GmbH, Berlin, Germany). Humidity and room temperature
Questionnaire development Thirty questions were created that dealt with factors that could potentially influence any tendency to become dehydrated when exposed to a warm environment. Topics included whether in general you tended to feel warm or cold, the tendency to sweat, the frequency of urination, the frequency that drinks were habitually consumed, the extent to which room heating was used and how they felt in a hot environment. Examples of the questions are found in Table 1. Each question was associated with a 100 millimetre line with the statement ‘Not at all like me’ at one end, and the statement ‘Very like me’ at the other. The line was marked to indicate the extent to which the descriptions applied to the subject. Using a ruler, a score from 0 to 100 was obtained for each question.
Room temperature and humidity were measured using a meter supplied by Trotec GmbH, Heinsberg, Germany. The temperature of the room varied from 30 to 31 °C with an average humidity of 53% that, depending on the testing day, varied from 43% to 62%. Food intake Subjects were asked to recall the food and drink they had consumed later than 1700 the previous evening and in the morning before coming to the laboratory. The time period was chosen to gain an indication of the contribution of recent dietary consumption to the hydration status rather than to establish general dietary habits. Where possible the size of portions were described, for example a
Table 1 Factor analysis of questions related to fluid intake and the response to heat.
I sweat very easily on a warm day I find my clothes are damp on a hot day I sweat more easily than other people I often have sweaty hands When I cold I find it difficult to get warm I like the heating turned up I have difficulty getting warm in bed I frequently feel cold I am often thirsty I urinate less frequently than my friends I drink (non-alcoholic) more than other people I get headaches when the weather is hot I feel exhausted in hot weather I do not deal well with a hot temperature
Sweats easily
Feels the cold
Drinks little
Deals badly with heat
0.84 0.78 0.88 0.70 −0.06 −0.04 −0.06 −0.25 0.05 0.20 0.15 −0.11 0.01 0.15
−0.03 −0.17 −0.04 −0.13 0.85 0.77 0.79 0.83 −0.09 0.03 −0.18 0.32 0.04 -0.33
−0.06 0.09 −0.20 −0.02 0.03 −0.19 −0.16 −0.03 0.80 −0.82 0.80 0.11 0.01 0.03
0.08 0.18 −0.20 −0.05 −0.09 0.00 −0.01 0.07 0.23 0.09 0.08 0.63 0.86 0.73
The figures in bold indicate the questions used to create overall scores for the four dimensions that were obtained.
78
D. Benton et al./Appetite 87 (2015) 76–80
cup or mug or the number of slices of toast. The data were analyzed using CompEat Nutritional Analysis Software (Nutrition Systems, Banbury, UK) that uses the McCance and Widdowson (2002) food tables. Where information was not available nutritional information was obtained from food labels or other manufacturer’s data. The reliability of coding was established by comparing the results from pairs of coders that found typically an inter-observer reliability in excess of 0.9. Portion sizes, where they were not specified by the participant, were estimated using established norms (Food Portion Sizes, 1988; Davies & Dickerson, 1991). No attempt was made to offer standard meals, or to prescribe the dietary intake, as the study was conceived as an examination of individuals going about their normal life. Mood The Profile of Mood States (POMS; Lorr & McNair, 1984) is a selfreport questionnaire that measures six dimensions of mood: (1) Composed – Anxious; (2) Energetic – Tired; (3) Elated – Depressed; (4) Clear-headed – Confused; (5) Agreeable – Hostile; (6) Confident – Unsure. To allow rapid measurement, these dimensions were assessed using 100 mm visual analogue scales, rather than the full test, with one of each pair of adjectives placed at the end of a line. Each mood was measured as the point in millimetres that a line was marked. Mood was assessed on arrival and at the end of the testing session. Statistical analysis The responses of 110 subjects to 30 questions were subject to factor analysis using principal component analysis with varimax rotation. As initial analysis of the data from males and females resulted in a very similar factor structure they were then treated as one sample. A principal component analysis with a varimax (orthogonal) rotation was then carried out with 15 questions chosen because they exclusively weighted on one of the major types of responses that was obtained. The Kaiser–Meyer Olkin measure of sampling adequacy suggested that the sample was factorable (KMO = 0.725). In addition the Bartlett’s test of sphericity (Chi squared 665.48, p < 0.001) demonstrated that the resulting factors were uncorrelated. Factor scores were created by adding together the responses for questions that weighted heavily on each factor: using Pearson product moment correlations these were then related to a range of measures of dehydration. Results There was clear evidence of individual differences in the response to being at 30 oC for four hours. In males on average there was a loss of 0.86% (0.41) of body mass and in females 0.63% (0.36), although these averages hide considerable individual differences in the response to identical conditions. In males the losses varied from 0.42% to 2.39% and in females from 0.24% to 1.52%. Osmolarity at baseline was 628 (236) in males and 656 (318) mOsm/kg in females. At the end of the morning these values had risen on average to 675 (214) in males and 670 (295) mOsm/kg in females. At baseline 30% males and 40% of females had values greater than 800 mOsm/kg, values that had risen to 35% of males and 45% of females after four hours at 30 °C. An initial factor analysis of thirty statements concerning the response to heat produced four factors: Table 1 reports the results of a subsequent analysis that considered the fifteen most heavily weighted questions. A four-factor solution resulted with a simple structure that accounted for 66.8% of the variance (26.4%; 15.9%; 14.6%; 9.9%). The first factor was labelled ‘Sweats easily’ and was heavily weighted on the first four questions (Table 1): “I sweat very
easily on a warm day”; “I find my clothes are damp on a hot day”; “I often have sweaty hands”; “I sweat more easily than other people”. When the responses were added together a high score indicated that sweating occurred easily. The second factor was called ‘Feels the cold’ (questions 5–8). Again four questions were added: “I frequently feel cold”; “I have difficulty getting warm in bed”; “When I cold I find it difficult to get warm”; “I like the heating turned up”. A high score meant that an individual is often cold. The third dimension ‘Drinks a lot’ (questions 9–11) involved adding three questions: “I am often thirsty”; “I urinate less frequently than my friends (subtracted from 100)”; “I drink (nonalcoholic) more than other people”. A high score indicated frequent drinking and urination. The final factor was labelled ‘Deals badly with heat’ (questions 12–15) and was measured using four questions: “When I am hot I find it difficult to cool down”; “I get headaches when the weather is hot”; “I do not deal well with hot temperature”; “I feel exhausted in hot weather”. A high score was associated with the perception that hot weather was experienced poorly. A visual examination of the pattern of weightings illustrates that these four factors were independent of each other (Table 1). Table 2 reports correlations, in females, between these four dimensions and the bodily response to sitting at 30 oC for four hours. Of the four dimensions two were predictive. Although those who reported that they were more likely to sweat did not in fact loose more weight due to perspiration (Tables 2 and 3), they produced a smaller volume of urine (−0.41, p < 0.02) and lost less overall body mass (sweat + urine; 0.35, p < 0.05). The most predictive dimension was an indication of the frequency you usually drank and urinated. Those who drank a lot had more water in their diet (0.38, p < 0.003) and at baseline the osmolality of urine was lower (−0.36, p < 0.005). After being subject to heat those who habitually drank more produced more urine (0.48, p < 0.007), had a lower urine osmolality at the end of the study (−0.54, p < 0.002) and lost greater overall body mass (sweat + urine, −0.44, p < 0.02). Table 3 reports similar correlations in males. Again at baseline, those who reported that they habitually drank more frequently in fact had more water in the diet (0.29, p < 0.04) and similarly those who dealt badly with heat habitually drank to a greater extent (0.30, p < 0.03). There were, however, no further significant associations in males between the four questionnaire dimensions and the response to a warm environment. Finally the four dimensions were related to changes in mood over the morning (Table 4). Females, who reported that they responded badly to heat, were at the end of the testing session more confused rather than clearheaded (−0.50, p < 0.005), hostile rather than agreeable (−0.36, p < 0.05), unsure rather than confident (−0.45, p < 0.01) and depressed rather than elated (−0.40, p < 0.03). Both males and females who reported themselves as habitually
Table 2 Correlations in females between questionnaire measures and the response to heat. Sweats easily Baseline Urine osmolarity Water in diet Response to 4 hours of heat Loss weight due sweat Volume of urine Total loss body mass Urine osmolarity at finish
Feels the cold
Drinks a lot
Deals badly with heat
0.23 0.00
−0.15 −0.06
−0.36;p < 0.005 0.38;p < 0.003
−0.09 −0.19
−0.13 −0.41; p < 0.02 0.35; p < 0.05 0.34
0.26 0.12
0.19 0.48;p < 0.007
0.26 −0.13
−0.10 0.05
−0.54;p < 0.02
0.11
−0.61;p < 0.002
0.06
Baseline sample N = 61; response to four hours heat N = 30.
D. Benton et al./Appetite 87 (2015) 76–80
Table 3 Correlations in males between questionnaire measures and the response to heat. Sweats easily
Feels the cold
Drinks a lot
Deals badly with heat
Baseline Urine osmolarity Water in diet
−0.11 0.16
0.00 −0.16
−0.21 0.29; p < 0.04
−0.04 0.30; p < 0.03
Response to 4 hours of heat Loss weight due to sweat Volume of urine Total loss body mass Urine osmolarity after 4 hours
0.08 −0.15 0.12 0.13
0.14 −0.18 0.19 −0.06
−0.04 0.22 −0.09 −0.01
0.18 0.06 0.14 0.07
79
Table 5 Correlations between dietary intake and the percentage loss of body mass in response to exposure to 30 oC for four hours.
Energy (kcal) Protein (g) Carbohydrate (g) Fat (g) Water (g)
Males
Correlation with loss of body mass
Females
Correlation with loss of body mass
307 (246) 12.0 (6.0) 41.4 (31.4) 10.7 (11.9) 509 (225)
0.20 −0.25 0.23 0.24 −0.14
234 (69) 9.0 (4.7) 31.4 (15.0) 7.6 (3.1) 415 (214)
0.25 0.23 0.05 0.23 −0.13
The data are means (standard deviations). Males n = 23; Females n = 30.
Baseline sample N = 50; response to four hours heat N = 23.
drinking more rated themselves as more confused rather than clearheaded at the end of the morning (−0.42, p < 0.05 and −0.41, p < 0.03). Differences in males and females on the four dimensions were examined although with only one factor was there a significant difference; females were significantly more likely to report that they felt the cold (184 (86) v 97 (58), p < 0.001). There was in contrast no differences with ‘Sweats easily’ (male 221 (73) v female 198 (91), ‘Drinks a lot’ (male 180 (54) v female 160 (69) and ‘Deals poorly with heat’ (male 188 (84) v females 203 (70)). Finally the extent to which recent differences in dietary intake might be influential was examined (Table 5). In no instance was a macro-nutrient, or the water consumed as either drinks or as a component of food items, correlated significantly with the extent to which body mass was lost. The present findings were observed in those who had in the previous 16 hours (including a period when asleep) consumed in males 509 and females 425 millilitres of water. Discussion There were large differences in the extent to which fluid was lost under identical conditions: differences of a sufficient size that the nature of these responses to heat are likely to play a major part in the likelihood that somebody will become dehydrated. As such, an
Table 4 Correlations between questionnaire measures of the response to heat and mood after exposure to 30 °C for four hours. Sweats easily Clearheaded M F Agreeable M F Composed M F Confident M F Energetic M F Elated M F Thirsty M F
Feels the cold
Drinks a lot
Deals badly with heat
−0.42;p < 0.05 −0.41;p < 0.03
−0.08 −0.50;p < 0.005
0.12 −0.08
0.32 0.01
0.13 −0.28
0.03 −0.01
−0.31 0.08
0.32 −0.35;p < 0.05
0.07 0.02
0.21 −0.16
0.15 −0.08
0.26 −0.22
0.21 0.14
−0.23 −0.22
0.00 −0.25
0.41;p < 0.05 −0.45;p < 0.01
−0.06 −0.03
0.21 −0.20
−0.14 −0.15
0.15 −0.18
−0.16 −0.11
0.17 −0.29
−0.10 0.02
0.06 −0.40;p < 0.03
0.14 −0.08
−0.16 −0.31
0.19 −0.03
0.21 −0.12
Changes in mood from baseline to the end of the study were calculated such that a negative score indicated poorer mood. These were correlated with questionnaire measures of how individuals responded to heat. Males (M) n = 23; Females(F) n = 30. The figures in bold reached statistical significance.
ability to predict the extent of such responses would allow individuals susceptible to dehydration to be distinguished. Table 2 illustrates that in females the ‘Drinks a lot’ dimension was associated with various measures associated with the development of dehydration. The dimension predicted baseline osmolality and the amount of water in the diet. It was also associated with a greater production of urine, a greater loss of body mass and lower urine osmolality at the end of the session. In contrast, in males, although the ‘Drinks a lot’ factor was related to the amount of water in the diet, it did not predict the bodily response to heat (Table 3). However, Benton and Young (2014) noted that there had been surprisingly few studies that compared the responses of males and females; surprising as it is known that female hormones influence kidney functioning. As you progress through the menstrual cycle changes in the levels of antidiuretic hormone parallels those of oestrogen and it also rises if oestrogen is administered to postmenopausal women. High levels of oestrogen result in an increased intake of fluid, greater fluid retention and increases in total body water. Prashad, Fletcher, and Cooper (1987) found that urine production was less, and sodium release greater, in the post-menstrual phase, changes that were not related to the rate of glomerular filtration. Female sex hormones cause renal vasodilation and decrease the proportion of fluid that, after reaching the kidneys, passes into the renal tubules (Pechere-Bertschi & Burnier, 2007). In this context it is perhaps not surprising that the responses of males and females differed, although further work is required to establish the precise origin of these differences. As nearly 80% of the present sample of females was taking an oral contraceptive it was not possible to relate the findings to their hormonal status although it is an obvious suggestion that this question should be further examined. A related matter, that in the present context also needs further exploration, is the association between the stage of the menstrual cycle and the ability to control body temperature. Marsh and Jenkins (2002) noted that there is substantial evidence that the increased levels of progesterone that occur during the luteal phase of the menstrual cycle cause an increase in both core and skin temperatures, with a consequent change in the ambient temperature at which sweating begins. For example Inoue et al. (2005) reported that the body temperatures of women in the luteal phase were significantly higher than both men and women in the follicular phase. They also studied changes in the rate of sweating when legs were immersed in water at 42 °C. During heat exposure, sweat rates on the forehead, chest, back, and forearm of women were significantly lower than those of men although this phenomenon did not vary with the stage of the cycle. They concluded that, “compared with men, heat loss from women depends more on cutaneous vasodilation than on sweating, irrespective of the phase of the menstrual cycle”. Another factor that needs to be considered is that changes in body temperature might influence psychological functioning. When body temperature was raised to between 38.80 and 39.05 °C by immersion in water, the speed of performing various tests increased and
80
D. Benton et al./Appetite 87 (2015) 76–80
alertness declined (Holland et al., 1985). Temperature increases from a nadir in the early morning to a maximum level in the early evening; a variation of about 0.5 °C. In a study that uncoupled temperature from other physiological parameters Wright et al. (2002) found that memory, alertness and attention were better when body temperature was at its highest. As such the extent that variations in hormonal status interact with any environmentally induced changes in the body temperature of women may influence the development of dehydration and any related consequences. Given that it was the ‘Drinks a lot’ dimension that was most predictive it is important that Benton and Young (2014) noted that the influence of habitual levels of fluid intake had been little considered. Does the body adapt to a habitual level of intake so that a deviation from the usual pattern of intake has adverse consequences? Pross et al. (2014) compared those with a low (
Contents lists available at ScienceDirect
Appetite j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p p e t
Research report
The development of the predisposition to dehydration questionnaire David Benton *, Hayley Young, Kimberley Jenkins Department of Psychology, University of Wales Swansea, Swansea, Wales SA2 8PP, UK
A R T I C L E
I N F O
Article history: Received 1 September 2014 Received in revised form 14 November 2014 Accepted 24 November 2014 Available online 2 December 2014 Keywords: Dehydration Drinking Mood Osmolality Sweating Water consumption
A B S T R A C T
The role played by hydration in general health and well-being is an emerging public health issue, yet there are few tools available to monitor its status in large populations. The aim was therefore to develop a questionnaire that assesses individual differences in the tendency to lose body fluid in a warm environment and hence become dehydrated. Fifty-three subjects sat in a room at 30 °C for four hours and changes in mood and measures of hydration were monitored. There were marked individual differences in the loss of body mass that differed from 0.24% to 2.39%. Females who reported habitually drinking a lot had more water in their diet and at baseline the osmolality of urine was lower. After being subject to heat, those who reported habitually drinking more produced more urine, had a lower urine osmolality at the end of the study, and overall more body mass was lost. Females who reported that they responded badly to heat were more confused, unsure and depressed after four hours at 30 °C. In males those reporting that they habitually drank to a greater extent had more water in the diet, and also those who dealt badly with heat habitually drank more. It was concluded that particularly in females, questionnaire measures were able to predict changes in hydration that result from a warm environment. © 2014 Elsevier Ltd. All rights reserved.
Introduction The role played by hydration in general health and well-being is an emerging public health issue as water makes up 50–60% of lean bodyweight in men and 45–50% in women. Intracellular fluid makes up about 35%, extracellular fluid about 12% and plasma about 4–5% of lean bodyweight (Kumar & Clark, 2012). Water plays a role in all aspects of bodily functioning including the distribution of oxygen and nutrients, the removal of waste products, serving as a lubricant and regulating temperature. The loss of bodily water adversely affects both physical (Cheuvront, Carter, & Sawka, 2003) and psychological performance (Benton & Young, 2014), it disrupts thermoregulation and appetite and results in headaches, irritability and sleepiness. EFSA (2010) concluded that “a cause and effect relationship has been established between the dietary intake of water and maintenance of normal physical and cognitive functions”. However, to date, much of the study of the psychological correlates of hydration status has relied on taking physiological measurements in small samples, usually in those who have exercised, an approach that cannot be easily applied more generally. If hydration is to be studied in large populations there is a need for tools that can be used widely. The present study therefore considered one aspect of the topic, the measurement of individual differences in the response to a warm environment, as adequate
* Corresponding author. E-mail address: [email protected] (D. Benton). http://dx.doi.org/10.1016/j.appet.2014.11.029 0195-6663/© 2014 Elsevier Ltd. All rights reserved.
hydration is particularly important for thermoregulation (EFSA, 2010). The provision of water helps to maintain the body’s core temperature within the desirable range and both mood and cognition are influenced by changes in core body temperature (Holland et al., 1985; Wright, Hull, & Czeisler, 2002). An increased body temperature will result from both decreased sweating and a lower blood flow in the skin, both of which are associated with dehydration. Euhydration implies being in water balance; that is, the intake of water matches the amount lost. However, as there are well described homeostatic mechanisms that maintain hydration within a prescribed range, many suggest that it is improbable that minor changes in the intake of fluid will disrupt functioning (Benton, 2011). Dehydration is typically defined as the loss of a particular percentage of body mass, with a loss greater than one percent being defined as dehydration. Initially water is lost from the blood and then from the cells. A loss of body mass, greater than one percent, is associated with the enlargement of the brain ventricles, a reflection of the shrinking of brain cells (Benton & Young, 2014). A review concluded that “when dehydration reduces body mass by over 2% there are consistent reports that mood is influenced; fatigue is greater and alertness lower. The effects on cognition have been less consistent. . ..” (Benton & Young, 2014). In this context there is an interest in evaluating the water balance of populations and hence a need to develop suitable approaches. Recently, the Water Balance Questionnaire was developed by Malisova et al. (2012). This questionnaire uses a wide range of questions to estimate water intake and its loss. Water intake reflects the consumption of solid and fluid foods and the drinking of water, whereas its loss results from its excretion in urine, faeces and sweat.
D. Benton et al./Appetite 87 (2015) 76–80
In addition what is eaten or drunk, activity levels and the weather influence water balance. One factor that such an approach does not consider is individual differences in the response to a warm environment. The present study therefore related a series of questions concerning the reaction to a warm environment to the loss of bodily fluids over a four hour period spent at 30 °C.
Method Baseline data were collected from 110 students of Swansea University, of whom 49 were males and 61 females. Of these 23 males and 30 females were then exposed, without drinking, to a temperature of 30 °C for four hours. Of the females 78% were taking an oral contraceptive. They were on average 21.5 years. The average BMI of the males was 25.3 (4.4) and females 23.4 (3.6). The procedure was approved by the Swansea University Psychology Ethics Committee.
Procedure Subjects came to the laboratory at 0900 having consumed their usual breakfast. They emptied their bladder and gave a sample of urine. They were weighed, rated their mood, responded to the questions concerning the reaction to a warm environment and filled in the dietary diary. Half the subjects then remained in a room at 30 °C for four hours. After 230 minutes they were again weighed before emptying their bladder and giving a second sample of urine. Then after 240 minutes they were weighed a final time and mood was rated.
77
Body mass Body mass was measured using an electronic scale (Kern KMSTM, Kern and Sohn GmbH, Germany) that, to avoid problems associated with movement, took 50 assessments over a 5 second period and produced an average. It was sensitive enough to weight to 5 grams and could pick up over short periods changes in body mass due to breathing and perspiration. Subjects were weighed on arrival and after 230 minutes. Finally having emptied the bladder they were weighed for a final time after 240 minutes. Changes in mass from baseline to 230 minutes reflected water loss largely due to perspiration and breathing, and changes from 230 to 240 minutes reflected urine production. Preliminary studies examined the possible need to weigh individuals naked, as it was possible that some weight loss might be masked if perspiration remained in clothing. In the event this proved unnecessary. The percentage loss of body mass, based on 32 participants weighed naked, was 0.26% (0.05) after 230 minutes and 0.60% (0.33) following urination at 240 minutes. These values were virtually identical to the same individuals measured having worn light clothing: the comparable percentage changes were 0.26% (0.07) and 0.61% (0.35). The data presented were therefore obtained from subjects who wore the same light clothing throughout the procedure. Osmolality At the beginning and end of the procedure a urine sample was collected and the osmolality measured using an Osmomat 3000 freezing point osmometer (Gonotec GmbH, Berlin, Germany). Humidity and room temperature
Questionnaire development Thirty questions were created that dealt with factors that could potentially influence any tendency to become dehydrated when exposed to a warm environment. Topics included whether in general you tended to feel warm or cold, the tendency to sweat, the frequency of urination, the frequency that drinks were habitually consumed, the extent to which room heating was used and how they felt in a hot environment. Examples of the questions are found in Table 1. Each question was associated with a 100 millimetre line with the statement ‘Not at all like me’ at one end, and the statement ‘Very like me’ at the other. The line was marked to indicate the extent to which the descriptions applied to the subject. Using a ruler, a score from 0 to 100 was obtained for each question.
Room temperature and humidity were measured using a meter supplied by Trotec GmbH, Heinsberg, Germany. The temperature of the room varied from 30 to 31 °C with an average humidity of 53% that, depending on the testing day, varied from 43% to 62%. Food intake Subjects were asked to recall the food and drink they had consumed later than 1700 the previous evening and in the morning before coming to the laboratory. The time period was chosen to gain an indication of the contribution of recent dietary consumption to the hydration status rather than to establish general dietary habits. Where possible the size of portions were described, for example a
Table 1 Factor analysis of questions related to fluid intake and the response to heat.
I sweat very easily on a warm day I find my clothes are damp on a hot day I sweat more easily than other people I often have sweaty hands When I cold I find it difficult to get warm I like the heating turned up I have difficulty getting warm in bed I frequently feel cold I am often thirsty I urinate less frequently than my friends I drink (non-alcoholic) more than other people I get headaches when the weather is hot I feel exhausted in hot weather I do not deal well with a hot temperature
Sweats easily
Feels the cold
Drinks little
Deals badly with heat
0.84 0.78 0.88 0.70 −0.06 −0.04 −0.06 −0.25 0.05 0.20 0.15 −0.11 0.01 0.15
−0.03 −0.17 −0.04 −0.13 0.85 0.77 0.79 0.83 −0.09 0.03 −0.18 0.32 0.04 -0.33
−0.06 0.09 −0.20 −0.02 0.03 −0.19 −0.16 −0.03 0.80 −0.82 0.80 0.11 0.01 0.03
0.08 0.18 −0.20 −0.05 −0.09 0.00 −0.01 0.07 0.23 0.09 0.08 0.63 0.86 0.73
The figures in bold indicate the questions used to create overall scores for the four dimensions that were obtained.
78
D. Benton et al./Appetite 87 (2015) 76–80
cup or mug or the number of slices of toast. The data were analyzed using CompEat Nutritional Analysis Software (Nutrition Systems, Banbury, UK) that uses the McCance and Widdowson (2002) food tables. Where information was not available nutritional information was obtained from food labels or other manufacturer’s data. The reliability of coding was established by comparing the results from pairs of coders that found typically an inter-observer reliability in excess of 0.9. Portion sizes, where they were not specified by the participant, were estimated using established norms (Food Portion Sizes, 1988; Davies & Dickerson, 1991). No attempt was made to offer standard meals, or to prescribe the dietary intake, as the study was conceived as an examination of individuals going about their normal life. Mood The Profile of Mood States (POMS; Lorr & McNair, 1984) is a selfreport questionnaire that measures six dimensions of mood: (1) Composed – Anxious; (2) Energetic – Tired; (3) Elated – Depressed; (4) Clear-headed – Confused; (5) Agreeable – Hostile; (6) Confident – Unsure. To allow rapid measurement, these dimensions were assessed using 100 mm visual analogue scales, rather than the full test, with one of each pair of adjectives placed at the end of a line. Each mood was measured as the point in millimetres that a line was marked. Mood was assessed on arrival and at the end of the testing session. Statistical analysis The responses of 110 subjects to 30 questions were subject to factor analysis using principal component analysis with varimax rotation. As initial analysis of the data from males and females resulted in a very similar factor structure they were then treated as one sample. A principal component analysis with a varimax (orthogonal) rotation was then carried out with 15 questions chosen because they exclusively weighted on one of the major types of responses that was obtained. The Kaiser–Meyer Olkin measure of sampling adequacy suggested that the sample was factorable (KMO = 0.725). In addition the Bartlett’s test of sphericity (Chi squared 665.48, p < 0.001) demonstrated that the resulting factors were uncorrelated. Factor scores were created by adding together the responses for questions that weighted heavily on each factor: using Pearson product moment correlations these were then related to a range of measures of dehydration. Results There was clear evidence of individual differences in the response to being at 30 oC for four hours. In males on average there was a loss of 0.86% (0.41) of body mass and in females 0.63% (0.36), although these averages hide considerable individual differences in the response to identical conditions. In males the losses varied from 0.42% to 2.39% and in females from 0.24% to 1.52%. Osmolarity at baseline was 628 (236) in males and 656 (318) mOsm/kg in females. At the end of the morning these values had risen on average to 675 (214) in males and 670 (295) mOsm/kg in females. At baseline 30% males and 40% of females had values greater than 800 mOsm/kg, values that had risen to 35% of males and 45% of females after four hours at 30 °C. An initial factor analysis of thirty statements concerning the response to heat produced four factors: Table 1 reports the results of a subsequent analysis that considered the fifteen most heavily weighted questions. A four-factor solution resulted with a simple structure that accounted for 66.8% of the variance (26.4%; 15.9%; 14.6%; 9.9%). The first factor was labelled ‘Sweats easily’ and was heavily weighted on the first four questions (Table 1): “I sweat very
easily on a warm day”; “I find my clothes are damp on a hot day”; “I often have sweaty hands”; “I sweat more easily than other people”. When the responses were added together a high score indicated that sweating occurred easily. The second factor was called ‘Feels the cold’ (questions 5–8). Again four questions were added: “I frequently feel cold”; “I have difficulty getting warm in bed”; “When I cold I find it difficult to get warm”; “I like the heating turned up”. A high score meant that an individual is often cold. The third dimension ‘Drinks a lot’ (questions 9–11) involved adding three questions: “I am often thirsty”; “I urinate less frequently than my friends (subtracted from 100)”; “I drink (nonalcoholic) more than other people”. A high score indicated frequent drinking and urination. The final factor was labelled ‘Deals badly with heat’ (questions 12–15) and was measured using four questions: “When I am hot I find it difficult to cool down”; “I get headaches when the weather is hot”; “I do not deal well with hot temperature”; “I feel exhausted in hot weather”. A high score was associated with the perception that hot weather was experienced poorly. A visual examination of the pattern of weightings illustrates that these four factors were independent of each other (Table 1). Table 2 reports correlations, in females, between these four dimensions and the bodily response to sitting at 30 oC for four hours. Of the four dimensions two were predictive. Although those who reported that they were more likely to sweat did not in fact loose more weight due to perspiration (Tables 2 and 3), they produced a smaller volume of urine (−0.41, p < 0.02) and lost less overall body mass (sweat + urine; 0.35, p < 0.05). The most predictive dimension was an indication of the frequency you usually drank and urinated. Those who drank a lot had more water in their diet (0.38, p < 0.003) and at baseline the osmolality of urine was lower (−0.36, p < 0.005). After being subject to heat those who habitually drank more produced more urine (0.48, p < 0.007), had a lower urine osmolality at the end of the study (−0.54, p < 0.002) and lost greater overall body mass (sweat + urine, −0.44, p < 0.02). Table 3 reports similar correlations in males. Again at baseline, those who reported that they habitually drank more frequently in fact had more water in the diet (0.29, p < 0.04) and similarly those who dealt badly with heat habitually drank to a greater extent (0.30, p < 0.03). There were, however, no further significant associations in males between the four questionnaire dimensions and the response to a warm environment. Finally the four dimensions were related to changes in mood over the morning (Table 4). Females, who reported that they responded badly to heat, were at the end of the testing session more confused rather than clearheaded (−0.50, p < 0.005), hostile rather than agreeable (−0.36, p < 0.05), unsure rather than confident (−0.45, p < 0.01) and depressed rather than elated (−0.40, p < 0.03). Both males and females who reported themselves as habitually
Table 2 Correlations in females between questionnaire measures and the response to heat. Sweats easily Baseline Urine osmolarity Water in diet Response to 4 hours of heat Loss weight due sweat Volume of urine Total loss body mass Urine osmolarity at finish
Feels the cold
Drinks a lot
Deals badly with heat
0.23 0.00
−0.15 −0.06
−0.36;p < 0.005 0.38;p < 0.003
−0.09 −0.19
−0.13 −0.41; p < 0.02 0.35; p < 0.05 0.34
0.26 0.12
0.19 0.48;p < 0.007
0.26 −0.13
−0.10 0.05
−0.54;p < 0.02
0.11
−0.61;p < 0.002
0.06
Baseline sample N = 61; response to four hours heat N = 30.
D. Benton et al./Appetite 87 (2015) 76–80
Table 3 Correlations in males between questionnaire measures and the response to heat. Sweats easily
Feels the cold
Drinks a lot
Deals badly with heat
Baseline Urine osmolarity Water in diet
−0.11 0.16
0.00 −0.16
−0.21 0.29; p < 0.04
−0.04 0.30; p < 0.03
Response to 4 hours of heat Loss weight due to sweat Volume of urine Total loss body mass Urine osmolarity after 4 hours
0.08 −0.15 0.12 0.13
0.14 −0.18 0.19 −0.06
−0.04 0.22 −0.09 −0.01
0.18 0.06 0.14 0.07
79
Table 5 Correlations between dietary intake and the percentage loss of body mass in response to exposure to 30 oC for four hours.
Energy (kcal) Protein (g) Carbohydrate (g) Fat (g) Water (g)
Males
Correlation with loss of body mass
Females
Correlation with loss of body mass
307 (246) 12.0 (6.0) 41.4 (31.4) 10.7 (11.9) 509 (225)
0.20 −0.25 0.23 0.24 −0.14
234 (69) 9.0 (4.7) 31.4 (15.0) 7.6 (3.1) 415 (214)
0.25 0.23 0.05 0.23 −0.13
The data are means (standard deviations). Males n = 23; Females n = 30.
Baseline sample N = 50; response to four hours heat N = 23.
drinking more rated themselves as more confused rather than clearheaded at the end of the morning (−0.42, p < 0.05 and −0.41, p < 0.03). Differences in males and females on the four dimensions were examined although with only one factor was there a significant difference; females were significantly more likely to report that they felt the cold (184 (86) v 97 (58), p < 0.001). There was in contrast no differences with ‘Sweats easily’ (male 221 (73) v female 198 (91), ‘Drinks a lot’ (male 180 (54) v female 160 (69) and ‘Deals poorly with heat’ (male 188 (84) v females 203 (70)). Finally the extent to which recent differences in dietary intake might be influential was examined (Table 5). In no instance was a macro-nutrient, or the water consumed as either drinks or as a component of food items, correlated significantly with the extent to which body mass was lost. The present findings were observed in those who had in the previous 16 hours (including a period when asleep) consumed in males 509 and females 425 millilitres of water. Discussion There were large differences in the extent to which fluid was lost under identical conditions: differences of a sufficient size that the nature of these responses to heat are likely to play a major part in the likelihood that somebody will become dehydrated. As such, an
Table 4 Correlations between questionnaire measures of the response to heat and mood after exposure to 30 °C for four hours. Sweats easily Clearheaded M F Agreeable M F Composed M F Confident M F Energetic M F Elated M F Thirsty M F
Feels the cold
Drinks a lot
Deals badly with heat
−0.42;p < 0.05 −0.41;p < 0.03
−0.08 −0.50;p < 0.005
0.12 −0.08
0.32 0.01
0.13 −0.28
0.03 −0.01
−0.31 0.08
0.32 −0.35;p < 0.05
0.07 0.02
0.21 −0.16
0.15 −0.08
0.26 −0.22
0.21 0.14
−0.23 −0.22
0.00 −0.25
0.41;p < 0.05 −0.45;p < 0.01
−0.06 −0.03
0.21 −0.20
−0.14 −0.15
0.15 −0.18
−0.16 −0.11
0.17 −0.29
−0.10 0.02
0.06 −0.40;p < 0.03
0.14 −0.08
−0.16 −0.31
0.19 −0.03
0.21 −0.12
Changes in mood from baseline to the end of the study were calculated such that a negative score indicated poorer mood. These were correlated with questionnaire measures of how individuals responded to heat. Males (M) n = 23; Females(F) n = 30. The figures in bold reached statistical significance.
ability to predict the extent of such responses would allow individuals susceptible to dehydration to be distinguished. Table 2 illustrates that in females the ‘Drinks a lot’ dimension was associated with various measures associated with the development of dehydration. The dimension predicted baseline osmolality and the amount of water in the diet. It was also associated with a greater production of urine, a greater loss of body mass and lower urine osmolality at the end of the session. In contrast, in males, although the ‘Drinks a lot’ factor was related to the amount of water in the diet, it did not predict the bodily response to heat (Table 3). However, Benton and Young (2014) noted that there had been surprisingly few studies that compared the responses of males and females; surprising as it is known that female hormones influence kidney functioning. As you progress through the menstrual cycle changes in the levels of antidiuretic hormone parallels those of oestrogen and it also rises if oestrogen is administered to postmenopausal women. High levels of oestrogen result in an increased intake of fluid, greater fluid retention and increases in total body water. Prashad, Fletcher, and Cooper (1987) found that urine production was less, and sodium release greater, in the post-menstrual phase, changes that were not related to the rate of glomerular filtration. Female sex hormones cause renal vasodilation and decrease the proportion of fluid that, after reaching the kidneys, passes into the renal tubules (Pechere-Bertschi & Burnier, 2007). In this context it is perhaps not surprising that the responses of males and females differed, although further work is required to establish the precise origin of these differences. As nearly 80% of the present sample of females was taking an oral contraceptive it was not possible to relate the findings to their hormonal status although it is an obvious suggestion that this question should be further examined. A related matter, that in the present context also needs further exploration, is the association between the stage of the menstrual cycle and the ability to control body temperature. Marsh and Jenkins (2002) noted that there is substantial evidence that the increased levels of progesterone that occur during the luteal phase of the menstrual cycle cause an increase in both core and skin temperatures, with a consequent change in the ambient temperature at which sweating begins. For example Inoue et al. (2005) reported that the body temperatures of women in the luteal phase were significantly higher than both men and women in the follicular phase. They also studied changes in the rate of sweating when legs were immersed in water at 42 °C. During heat exposure, sweat rates on the forehead, chest, back, and forearm of women were significantly lower than those of men although this phenomenon did not vary with the stage of the cycle. They concluded that, “compared with men, heat loss from women depends more on cutaneous vasodilation than on sweating, irrespective of the phase of the menstrual cycle”. Another factor that needs to be considered is that changes in body temperature might influence psychological functioning. When body temperature was raised to between 38.80 and 39.05 °C by immersion in water, the speed of performing various tests increased and
80
D. Benton et al./Appetite 87 (2015) 76–80
alertness declined (Holland et al., 1985). Temperature increases from a nadir in the early morning to a maximum level in the early evening; a variation of about 0.5 °C. In a study that uncoupled temperature from other physiological parameters Wright et al. (2002) found that memory, alertness and attention were better when body temperature was at its highest. As such the extent that variations in hormonal status interact with any environmentally induced changes in the body temperature of women may influence the development of dehydration and any related consequences. Given that it was the ‘Drinks a lot’ dimension that was most predictive it is important that Benton and Young (2014) noted that the influence of habitual levels of fluid intake had been little considered. Does the body adapt to a habitual level of intake so that a deviation from the usual pattern of intake has adverse consequences? Pross et al. (2014) compared those with a low (
