301
Psychiatry Research, 38:301-311
Elsevier
Elevated Serum Lactate Associated Induced by Hyperventilation Richard J. Maddock,
Cameron
S. Carter,
With
and Dorothy
Panic
Attacks
W. Gietzen
Received December 26, 1990; revised version received June 21, 1991;acceptedAugust
2.3, 1991.
Abstract. Several lines of evidence suggest that lactate metabolism may be altered in panic disorder. We recently reported exaggerated increases in serum lactate in panic patients following hyperventilation during glucose infusion. In the current study, lactate metabolism was stimulated by hyperventilation following glucose ingestion in 12 panic patients and 12 controls. The seven patients who panicked during hyperventilation exhibited larger increases in serum lactate levels than nonpanicking patients or controls. The lactate response was significantly correlated with peak ratings of anxiety and panic symptoms, but not correlated with insulin or cortisol levels, heart rate, pCO,, adiposity, exercise habits, or diet. Hyperventilation-induced panic appears to be associated with metabolic changes leading to elevated serum lactate. Key Words. Panic,
hyperventilation,
lactate,
glycolysis,
cortisol,
insulin.
There is a growing consensus that vulnerability to frequent recurrent panic attacks can be a consequence of both psychological and biological factors (Ballenger, 1989). Although recent studies have yielded abundant evidence for biological factors predisposing to panic disorder (Roy-Byrne and Cowley, 1988), the specific nature of these factors has not yet been identified. Several lines of evidence suggest that an abnormality in lactic acid metabolism may be present in persons vulnerable to panic attacks. Cohen and White (1950) found that patients with neurocirculatory asthenia (and panic attacks) had higher serum lactate levels under standard exercise conditions than control subjects matched for usual level of physical activity. Lactate infusion studies have found higher lactate levels (Liebowitz et al., 1985) and more rapid increases in lactate (Rainey et al., 1985) in panic patients compared with normal control subjects. Uhde and Boulenger (1989) recently reported that panic patients exhibited greater increases in serum lactate following caffeine ingestion than control subjects, and that patients who panicked after caffeine had greater increases in lactate than patients who did not panic. If lactate metabolism is altered in panic disorder, is this a stable characteristic of individuals with the disorder, or does it occur only in temporal association with acute anxiety or panic? In an earlier study from our laboratory, we demonstrated a greater Richard J. Maddock. M.D., is Director of the Anxiety Research Center and Assistant Professor of Psychiatry; Cameron S. Carter, M.B.B.S., is Associate Director of the Anxiety Research Center and Assistant Professor of Psychiatry; and Dorothy W. Gietzen. Ph.D., is Director of Research and of the Neurochemistry Laboratory and Associate Research Neurophysiologist in the Department of Psychiatry. University of California, Davis, School of Medicine. (Reprint requests to Dr. R.J. Maddock, Dept. of Psychiatry. University of California, Davis Medical Center, 4430 V St., Sacramento, CA 95817. USA.)
01651781/91/$03.50
@ 1991 Elsevier Scientific Publishers Ireland Ltd.
302 increase in serum lactate in panic patients compared with control subjects IO-20 min after the metabolic challenge of glucose infusion accompanied by hyperventilation (Maddock and Mateo-Bermudez, 1990). We now report on a second study using the glucose-hyperventilation technique to stimulate lactate metabolism in which the effects of acute panic are examined. Hyperventilation leads to a small increase in serum lactate in man and animals (Hood and Tannen, 1983). Recent studies have shown that hyperventilation during an intravenous infusion of glucose produces a greater increase in serum lactate than hyperventilation alone (Brautbar et al., 1983; Maddock and Mateo-Bermudez, 1990). Most circulating lactate in resting man is produced by muscle, brain, and erythrocytes, and is taken up by the liver, heart, and kidneys. Hyperventilation produces an intracellular alkalosis that causes stimulation of phosphofructokinase and other pH-sensitive enzymes in the glycolytic pathway and thereby leads to increased lactate production (Hood and Tannen, 1983). Glucose loading further increases serum lactate as insulin secretion increases entry of substrates (glucose and phosphate) into cells and decreases hepatic clearance by inhibition of gluconeogenesis (Brautbar et al., 1983). These observations provide the rationale for the technique used in the current study. Our earlier study found that patients with panic disorder had greater increases in serum lactate following hyperventilation combined withglucose infusion than control subjects. The current study attempts to extend these findings and evaluate various factors that may account for this phenomenon. In the first study, four of the eight patients experienced a panic attack during the hyperventilation condition. The increase in serum lactate may have been related to the occurrence of a panic attack, but the small number of subjects limited our power to test this hypothesis. The primary goal of the current study is to replicate our findings ina larger sample and to clarify the relationship between increased lactate and the occurrence of an acute panic attack. A secondary goal of the study is to assess several behavioral and physiological factors that may account for this altered lactate metabolism. Behavioral factors that may influence the regulation of enzymes involved in lactate metabolism include the proportion of carbohydrates in the diet (Gollnick et al., 1986) and the amount of vigorous physical activity usually engaged in by the subject (Ward et al., 1986). Both of these variables are assessed in the current study. Many physiological factors, including glucagon, catecholamines, cortisol, insulin, and body adiposity, can influence lactate metabolism (Massara and Camanni, 1970; Jacksonet al., 1986; Weissmanet al., 1986; Hagstrom et al., 1990). The last three factors are evaluated in this study. Methods Twelve outpatients from the University of California, Davis Anxiety Disorders Clinic and 12 normal control subjects participated in the study. There were nine women and three men in each group. The patients met DSM-I/I-R criteria (American Psychiatric Association, 1987) for panic disorder with agoraphobia. Six of the patients had mild, five had moderate, and one had severe agoraphobia. No patient had a concurrent diagnosis of major depression or substance abuse. The average age of the patients was 35.8 years (range 21-51). The average age of the control subjects was 36.0 years (range 23-52). All subjects were in good physical health and had no history of hepatic or metabolic disease. The control subjects were free of psychiatric illness. The patients had been free of psychoactive medications for a minimum of 2 weeks before
303 participation in this study. All patients reported at least weekly panic attacks (by DSM-III-R criteria) for the preceding 4 weeks. Subjects came to the laboratory at 8 a.m. after an overnight fast. A heparinized catheter was placed in a forearm vein and, after 30 min of rest, baseline measurements were made. Subjects then ingested glucose (1 g/kg body weight) in water. After a further 20 min of rest, prehyperventilation measurements were made. Subjects were then asked to “breathe deeply and rapidly” sufficient to maintain an end-tidal pC0, of 20 mmHg for 8 min. A visual analog monitor gave subjects feedback as to their end-tidal pC0, during hyperventilation, and they were instructed to maintain this level at or below 20 mmHg. After 8 min of hyperventilation (HV), subjects were instructed to “breathe normally.” Subjects were monitored during a 15-min recovery period following the end of the HV condition. Venous blood was sampled for lactate at baseline and 15 min after HV. This 15-min time point corresponded to the time of peak serum lactate observed in our first study (Maddock and Mateo-Bermudez, 1990). Lactate was measured by enzymatic methods (Tietz, 1976). In 10 patients and 10 control subjects, venous blood was sampled for insulin 20 min after the ingestion of glucose (before HV) and for cortisol at baseline and 15 min after HV. Insulin and cortisol were measured by radioimmunoassay. Venous blood was also sampled for plasma catecholamines, but technical problems made these data uninterpretable. End-tidal pCOZ was monitored and recorded continuously throughout the experiment from a loose-fitting face mask by a Godard infrared capnograph linked to a Grass polygraph. Heart rate was monitored continuously from electrocardiographic electrodes placed in the lead II position and recorded on a Grass polygraph. Sixty-second samples of heart rate and end-tidal pCOZ were hand scored every 2 min during baseline, HV, and for 15 min after HV. Anxiety symptoms were assessed by an 11-point (O-10) visual analog scale for overall anxiety rating (AR) and by the 17-item Acute Panic Inventory (API; Dillon et al., 1987) at baseline, immediately before HV, immediately after HV, and 15 min after HV. AR was also rated every 2 min during HV. Before beginning the experiment, all subjects described the type and degree of their regular physical activity over the past year using a modification of the Baecke Activity Scale (Baecke et al., 1982; Washburn and Montoye, 1986). This yielded an estimate of their average weekly energy expenditure from vigorous physical activity. In eight patients and eight control subjects, customary dietary patterns were assessed by having a trained dietitian instruct the subject in keeping a complete diary of daily food intake. Subjects then recorded their food intake for 4 consecutive days (3 weekdays and 1 weekend day). The diary information was evaluated using dietary analysis software (ESHA Research, 1988) to estimate average daily intake of carbohydrate calories and percentage of total caloric intake as carbohydrates. This technique is considered one of the most valid methods of assessing dietary intake (Block, 1982). Adiposity can be estimated from body weight relative to frame size. Of the various indices of relative body weight, body mass index (BMI q weight/ height*) is most closely correlated with direct measurements of adiposity (Harrison, 1985). Weight and height were measured for each subject, and BMI was calculated as kg/m*. Our criteria for a panic attack were an increase in intensity of at least four panic symptoms on the API, accompanied by either an increase of at least 2 points on the AR to a value of 5 or greater or any AR score of 7 or greater, plus the attending psychiatrist’s assessment that a panic attack occurred. Details regarding the assessment of anxiety symptoms and the demand characteristics of the hyperventilation procedure have been reported separately (Maddock and Carter, 1991). Group comparisons of physiological data were made by t test. The directional hypotheses that serum lactate would increase more in patients than control subjects, and more in patients who panicked during the procedure than nonpanicking patients or control subjects, were tested by one-tailed comparisons. All other comparisons were two-tailed. Group comparisons of rating scale data were made by the Mann-Whitney test. Pearson’s correlation coefficient and Kendall’s T were applied to measure correlations between physiological and rating scale data, respectively. Occurrence of panic attacks in the patient vs. the control group was compared by Fisher’s exact test.
304 Results All subjects completed the protocol and maintained their mean&O, below 20 mmHg during the HV condition (range = 11.2-18.7 mmHg). Seven of 12 patients and 1 of 12 control subjects had a panic attack during HV by our criteria. This represents a significantly greater occurrence of panic attacks in the patient group (p < 0.05). Table 1 presents serum lactate values. There were no significant differences in baseline lactate levels between patients and control subjects (t 0.5, df= 22, NS), or between patients who panicked and nonpanicking patients (1 = 1.1, df = 10, NS) or control subjects (t = 0.8, df = 17, NS). Patients who panicked had significantly greater increases in serum lactate than nonpanicking patients (t = 1.86, df= 10, p < 0.05, one-tailed) or control subjects (t = 1.84, df 17, p < 0.05, one-tailed). The increase in lactate in patients as a whole did not differ from that in control subjects (t = 0.98, df = 22, NS). Changes in serum lactate correlated significantly with peak scores in AR (T = 0.424, n = 24,~ < 0.005) and API (r = 0.34, n=24,p
Psychiatry Research, 38:301-311
Elsevier
Elevated Serum Lactate Associated Induced by Hyperventilation Richard J. Maddock,
Cameron
S. Carter,
With
and Dorothy
Panic
Attacks
W. Gietzen
Received December 26, 1990; revised version received June 21, 1991;acceptedAugust
2.3, 1991.
Abstract. Several lines of evidence suggest that lactate metabolism may be altered in panic disorder. We recently reported exaggerated increases in serum lactate in panic patients following hyperventilation during glucose infusion. In the current study, lactate metabolism was stimulated by hyperventilation following glucose ingestion in 12 panic patients and 12 controls. The seven patients who panicked during hyperventilation exhibited larger increases in serum lactate levels than nonpanicking patients or controls. The lactate response was significantly correlated with peak ratings of anxiety and panic symptoms, but not correlated with insulin or cortisol levels, heart rate, pCO,, adiposity, exercise habits, or diet. Hyperventilation-induced panic appears to be associated with metabolic changes leading to elevated serum lactate. Key Words. Panic,
hyperventilation,
lactate,
glycolysis,
cortisol,
insulin.
There is a growing consensus that vulnerability to frequent recurrent panic attacks can be a consequence of both psychological and biological factors (Ballenger, 1989). Although recent studies have yielded abundant evidence for biological factors predisposing to panic disorder (Roy-Byrne and Cowley, 1988), the specific nature of these factors has not yet been identified. Several lines of evidence suggest that an abnormality in lactic acid metabolism may be present in persons vulnerable to panic attacks. Cohen and White (1950) found that patients with neurocirculatory asthenia (and panic attacks) had higher serum lactate levels under standard exercise conditions than control subjects matched for usual level of physical activity. Lactate infusion studies have found higher lactate levels (Liebowitz et al., 1985) and more rapid increases in lactate (Rainey et al., 1985) in panic patients compared with normal control subjects. Uhde and Boulenger (1989) recently reported that panic patients exhibited greater increases in serum lactate following caffeine ingestion than control subjects, and that patients who panicked after caffeine had greater increases in lactate than patients who did not panic. If lactate metabolism is altered in panic disorder, is this a stable characteristic of individuals with the disorder, or does it occur only in temporal association with acute anxiety or panic? In an earlier study from our laboratory, we demonstrated a greater Richard J. Maddock. M.D., is Director of the Anxiety Research Center and Assistant Professor of Psychiatry; Cameron S. Carter, M.B.B.S., is Associate Director of the Anxiety Research Center and Assistant Professor of Psychiatry; and Dorothy W. Gietzen. Ph.D., is Director of Research and of the Neurochemistry Laboratory and Associate Research Neurophysiologist in the Department of Psychiatry. University of California, Davis, School of Medicine. (Reprint requests to Dr. R.J. Maddock, Dept. of Psychiatry. University of California, Davis Medical Center, 4430 V St., Sacramento, CA 95817. USA.)
01651781/91/$03.50
@ 1991 Elsevier Scientific Publishers Ireland Ltd.
302 increase in serum lactate in panic patients compared with control subjects IO-20 min after the metabolic challenge of glucose infusion accompanied by hyperventilation (Maddock and Mateo-Bermudez, 1990). We now report on a second study using the glucose-hyperventilation technique to stimulate lactate metabolism in which the effects of acute panic are examined. Hyperventilation leads to a small increase in serum lactate in man and animals (Hood and Tannen, 1983). Recent studies have shown that hyperventilation during an intravenous infusion of glucose produces a greater increase in serum lactate than hyperventilation alone (Brautbar et al., 1983; Maddock and Mateo-Bermudez, 1990). Most circulating lactate in resting man is produced by muscle, brain, and erythrocytes, and is taken up by the liver, heart, and kidneys. Hyperventilation produces an intracellular alkalosis that causes stimulation of phosphofructokinase and other pH-sensitive enzymes in the glycolytic pathway and thereby leads to increased lactate production (Hood and Tannen, 1983). Glucose loading further increases serum lactate as insulin secretion increases entry of substrates (glucose and phosphate) into cells and decreases hepatic clearance by inhibition of gluconeogenesis (Brautbar et al., 1983). These observations provide the rationale for the technique used in the current study. Our earlier study found that patients with panic disorder had greater increases in serum lactate following hyperventilation combined withglucose infusion than control subjects. The current study attempts to extend these findings and evaluate various factors that may account for this phenomenon. In the first study, four of the eight patients experienced a panic attack during the hyperventilation condition. The increase in serum lactate may have been related to the occurrence of a panic attack, but the small number of subjects limited our power to test this hypothesis. The primary goal of the current study is to replicate our findings ina larger sample and to clarify the relationship between increased lactate and the occurrence of an acute panic attack. A secondary goal of the study is to assess several behavioral and physiological factors that may account for this altered lactate metabolism. Behavioral factors that may influence the regulation of enzymes involved in lactate metabolism include the proportion of carbohydrates in the diet (Gollnick et al., 1986) and the amount of vigorous physical activity usually engaged in by the subject (Ward et al., 1986). Both of these variables are assessed in the current study. Many physiological factors, including glucagon, catecholamines, cortisol, insulin, and body adiposity, can influence lactate metabolism (Massara and Camanni, 1970; Jacksonet al., 1986; Weissmanet al., 1986; Hagstrom et al., 1990). The last three factors are evaluated in this study. Methods Twelve outpatients from the University of California, Davis Anxiety Disorders Clinic and 12 normal control subjects participated in the study. There were nine women and three men in each group. The patients met DSM-I/I-R criteria (American Psychiatric Association, 1987) for panic disorder with agoraphobia. Six of the patients had mild, five had moderate, and one had severe agoraphobia. No patient had a concurrent diagnosis of major depression or substance abuse. The average age of the patients was 35.8 years (range 21-51). The average age of the control subjects was 36.0 years (range 23-52). All subjects were in good physical health and had no history of hepatic or metabolic disease. The control subjects were free of psychiatric illness. The patients had been free of psychoactive medications for a minimum of 2 weeks before
303 participation in this study. All patients reported at least weekly panic attacks (by DSM-III-R criteria) for the preceding 4 weeks. Subjects came to the laboratory at 8 a.m. after an overnight fast. A heparinized catheter was placed in a forearm vein and, after 30 min of rest, baseline measurements were made. Subjects then ingested glucose (1 g/kg body weight) in water. After a further 20 min of rest, prehyperventilation measurements were made. Subjects were then asked to “breathe deeply and rapidly” sufficient to maintain an end-tidal pC0, of 20 mmHg for 8 min. A visual analog monitor gave subjects feedback as to their end-tidal pC0, during hyperventilation, and they were instructed to maintain this level at or below 20 mmHg. After 8 min of hyperventilation (HV), subjects were instructed to “breathe normally.” Subjects were monitored during a 15-min recovery period following the end of the HV condition. Venous blood was sampled for lactate at baseline and 15 min after HV. This 15-min time point corresponded to the time of peak serum lactate observed in our first study (Maddock and Mateo-Bermudez, 1990). Lactate was measured by enzymatic methods (Tietz, 1976). In 10 patients and 10 control subjects, venous blood was sampled for insulin 20 min after the ingestion of glucose (before HV) and for cortisol at baseline and 15 min after HV. Insulin and cortisol were measured by radioimmunoassay. Venous blood was also sampled for plasma catecholamines, but technical problems made these data uninterpretable. End-tidal pCOZ was monitored and recorded continuously throughout the experiment from a loose-fitting face mask by a Godard infrared capnograph linked to a Grass polygraph. Heart rate was monitored continuously from electrocardiographic electrodes placed in the lead II position and recorded on a Grass polygraph. Sixty-second samples of heart rate and end-tidal pCOZ were hand scored every 2 min during baseline, HV, and for 15 min after HV. Anxiety symptoms were assessed by an 11-point (O-10) visual analog scale for overall anxiety rating (AR) and by the 17-item Acute Panic Inventory (API; Dillon et al., 1987) at baseline, immediately before HV, immediately after HV, and 15 min after HV. AR was also rated every 2 min during HV. Before beginning the experiment, all subjects described the type and degree of their regular physical activity over the past year using a modification of the Baecke Activity Scale (Baecke et al., 1982; Washburn and Montoye, 1986). This yielded an estimate of their average weekly energy expenditure from vigorous physical activity. In eight patients and eight control subjects, customary dietary patterns were assessed by having a trained dietitian instruct the subject in keeping a complete diary of daily food intake. Subjects then recorded their food intake for 4 consecutive days (3 weekdays and 1 weekend day). The diary information was evaluated using dietary analysis software (ESHA Research, 1988) to estimate average daily intake of carbohydrate calories and percentage of total caloric intake as carbohydrates. This technique is considered one of the most valid methods of assessing dietary intake (Block, 1982). Adiposity can be estimated from body weight relative to frame size. Of the various indices of relative body weight, body mass index (BMI q weight/ height*) is most closely correlated with direct measurements of adiposity (Harrison, 1985). Weight and height were measured for each subject, and BMI was calculated as kg/m*. Our criteria for a panic attack were an increase in intensity of at least four panic symptoms on the API, accompanied by either an increase of at least 2 points on the AR to a value of 5 or greater or any AR score of 7 or greater, plus the attending psychiatrist’s assessment that a panic attack occurred. Details regarding the assessment of anxiety symptoms and the demand characteristics of the hyperventilation procedure have been reported separately (Maddock and Carter, 1991). Group comparisons of physiological data were made by t test. The directional hypotheses that serum lactate would increase more in patients than control subjects, and more in patients who panicked during the procedure than nonpanicking patients or control subjects, were tested by one-tailed comparisons. All other comparisons were two-tailed. Group comparisons of rating scale data were made by the Mann-Whitney test. Pearson’s correlation coefficient and Kendall’s T were applied to measure correlations between physiological and rating scale data, respectively. Occurrence of panic attacks in the patient vs. the control group was compared by Fisher’s exact test.
304 Results All subjects completed the protocol and maintained their mean&O, below 20 mmHg during the HV condition (range = 11.2-18.7 mmHg). Seven of 12 patients and 1 of 12 control subjects had a panic attack during HV by our criteria. This represents a significantly greater occurrence of panic attacks in the patient group (p < 0.05). Table 1 presents serum lactate values. There were no significant differences in baseline lactate levels between patients and control subjects (t 0.5, df= 22, NS), or between patients who panicked and nonpanicking patients (1 = 1.1, df = 10, NS) or control subjects (t = 0.8, df = 17, NS). Patients who panicked had significantly greater increases in serum lactate than nonpanicking patients (t = 1.86, df= 10, p < 0.05, one-tailed) or control subjects (t = 1.84, df 17, p < 0.05, one-tailed). The increase in lactate in patients as a whole did not differ from that in control subjects (t = 0.98, df = 22, NS). Changes in serum lactate correlated significantly with peak scores in AR (T = 0.424, n = 24,~ < 0.005) and API (r = 0.34, n=24,p
