Salt appetite was investigated in 14 patients with congenital adrenal hyperplasia of the salt-wasting form (SW group), 12 patients with the simple virilized form who are not salt losing, and 18 healthy siblings. Salt appetite was evaluated by questionnaire, preference tests, and dietary analyses. The findings showed that SW who were not therapeutically normalized showed increased salt appetite but no change in sweet preference. Their salt appetite correlated with symptoms of salt wasting, namely, plasma renin activity, plasma K+, and urine Na+ and (inversely) with blood pressure. Sensitivity to the taste of NaCl was not altered. Factor analyses of a larger group confirmed the distinction between salt appetite and sweet preference, but intake of dietary Na+ and sweet carbohydrates and intake of salty and sweet snacks did not reflect distinct salt or sweet preferences. We confirm that putative perinatal dehydration, due to maternal nausea and vomiting during pregnancy, childhood vomiting, and diarrhea with occasional saline infusion, was related to increased salt appetite in adolescence. The findings suggest that salt appetite in humans is determined by interdependent, innate, physiological, and acquired attributes. Salt appetite in SW patients is an adaptive response mediated by the renin-angiotensin system, an innate predisposition to acquire salt preference (in anticipation of both sodium loss and its consequence), and imprinting by perinatal hyponatremic occurrences. Our findings contribute to understanding human salt intake, provide insight into the motivation for salt in patients with congenital adrenal hyperplasia 21-OH deficiency, and may point the way to improvements in therapeutic compliance in these patients.
- long-term effect
- perinatal dehydration
- salt intake
- sodium depletion
- therapeutic compliance
humans have a salt appetite beyond that due to known sodium requirements. Little is known about its determinants, although some believe it is related to availability (25). It apparently is not learned because, although preference for salt in a specific food can be learned, there is no generalization to other foods (12, 40). One contributing factor to adult sodium appetite is perinatal sodium loss (3, 20, 21, 23, 39).
Whether salt appetite is increased in sodium-need states is better understood. Early reports of physiologically related salt appetite have been sporadic, lacked measurements or taste controls (13, 14, 41, 42, 44, 46), and have been inconclusive (8, 44). More recent reports, however, have shown increased salt preference after sodium loss, such as after dialysis or exercise (19, 22, 41, 45).
In contrast to what is known in humans, in animals, both spontaneous and need-induced salt intake are determined by hormonal, physiological, and genetic factors, and salt intake may be enduringly increased after episodes of sodium depletion and dehydration (4, 7, 23, 27, 32).
To investigate the relationship of sodium need to human sodium appetite, we studied salt appetite in patients with the classical form of congenital adrenal hyperplasia (CAH). In its most severe form, termed “salt wasting” (SW), untreated patients are unable to retain sodium.
CAH is an autosomal recessive disease caused by mutations of the CYP21 gene resulting in a deficiency of the 21-hydroxylase enzyme. The mutations restrict 21-hydroxylase synthesis, decreasing production of cortisol and aldosterone, increasing androgens, and inducing feedback elevation of ACTH. The severity of the CYP21 mutations determines the clinical category of CAH, with the most severe type being SW, presenting with salt loss dehydration from early infancy. The less severe form is the simple virilizing (SV) form, with prenatal virilization in females and precocious adrenarche in males. The least severe is the nonclassical or late-onset form, presenting in childhood or adolescence with symptoms of hyperandrogenism. Treatment of patients with the SW form includes hydrocortisone to replace the glucocorticoid deficiency and suppress ACTH secretion to maintain normal androgen production and fludrocortisone to replace mineralocorticoid deficiency (18, 37, 38).
Plasma levels of 17-hydroxy progesterone (17-OHP), ACTH, and plasma renin activity (PRA) are parameters of adequate therapy. Because aldosterone regulates sodium homeostasis, SW patients are in danger of excessive sodium excretion; if inadequately treated or noncompliant, patients are vulnerable to hypovolemia, hyponatremic dehydration, hyperkalemia, and shock (18, 37, 38). It is not known whether sodium appetite is increased in these patients.
Determining salt appetite in humans is complicated by the absence of a ubiquitous definition. Ingestion of NaCl in aqueous solution (used in animal research) is not a good predictor of the human preference for salt in food (12, 25, 28). Moreover, salt intake in humans comprises involuntary and discretionary intake (6, 15, 24, 25, 28, 36). Involuntary intake includes sodium inherent in foods and added in industrial processing or in cooking. It may be partially regulated insofar as the food choices an individual makes may be conditioned by postingestive consequences of sodium intake. In turn, these might influence discretionary salt intake (2, 10, 12, 24, 25, 28, 36). Discretionary salt intake also comprises a number of sources, e.g., salting at table, choice of salty food items in meals, and consumption of high-salt snacks. Intake from these various sources of salt is not necessarily correlated (20).
Because of these considerations, a single measure of salt intake or preference is unlikely to reliably reflect the avidity for salt, and a number of indexes of salt preference are often combined to provide a measure of “salt appetite” (3, 6, 20, 21, 25, 36). A test of salting of soup provides a measure of discretionary salt use. The technique that we employed, of mixing solutions (20, 22, 28), ensures that the result is driven by taste preference rather than by habit (10, 36). Discretionary salt use was also evaluated by asking the participants how much salt they add to each of some 20 food categories or items (3, 20, 28). To evaluate the participants' conceptions of their salt preferences relative to others, we asked them how much they like salt and how much of it they add to their food. To obtain a measure of preference for high-salt concentrations, we compared intake of salty and sweet snacks (3, 20). We used dietary recall to obtain an indication of involuntary salt intake. Finally, we calculated the mean of these different indexes, equally weighted, to operationally define salt appetite (20, 22).
Here, we investigated salt appetite in CAH. To validate our operational definition of salt appetite, we also examined a larger group of patients and healthy siblings and reexamined the relationship of perinatal sodium loss as a determinant of the enduring avidity for salt (3, 20, 39).
MATERIALS AND METHODS
Out patients of the pediatric endocrinology unit of Ha'Emek Medical Center and members of their families volunteered for the study. The study was approved by the Helsinki Committee of the hospital, by the Israeli Ministry of Health, and by the University of Haifa Committee for Human Experiments and informed consent was obtained from patients or their guardians.
Twenty-six patients, ethnic Arabs, of the pediatric endocrinology unit of Ha'Emek Medical Center with the classical form of CAH (14 SW and 12 SV) and 18 sibling controls participated. The SW and SV forms were diagnosed by clinical, hormonal, and genetic parameters. SW patients had a history of hyponatremic-hyperkalemic dehydration in infancy and elevated PRA, indicating sodium loss. Their phenotype was confirmed by mutations of the severe type (null and A group: see Table 3). SV presented with ambiguous virilization at birth or androgen excess at childhood. The patients of these two groups were further divided into two subgroups: therapeutically “normalized” (respectively nSV, nSW) and “nonnormalized” (uSV and uSW). Nonnormalized were defined by blood levels of ACTH above 46 ng/ml and 17-OHP above 10 ng/ml (Tables 1 and 5) (18, 37, 38). Note that plasma sodium within normal levels is defended in all cases; lowered plasma sodium entails hyponatremic crisis.
Measurement of Salt Appetite
Questionnaires and interviews on taste and dietary habits.
There were four parts to this measure, including liking score, seasoning score, dietary questionnaire, and frequency of eating pure salt.
For the liking score, the investigator asked each participant how much they like sweet and salty foods (4 levels, with examples of familiar items).
For the seasoning (salting and sweetening) score, the investigator interviewed each participant using a questionnaire comprising 22 common food items (3, 20, 28). Participants were asked how much they season each item with salt, sugar (or honey, etc.), or oils/fats (olive oil, butter, etc.; not reported here). Items were scored on an ordinal scale of four levels and averaged.
For the dietary questionnaire, participants were asked to recall their frequency of consumption and quantity during the last week of 50 items of food and drink (e.g., hummus, soft drinks) or categories (breakfast cereals) covering the Israeli diet. Energy, macronutrient, electrolyte, sweet dietary carbohydrates, and fluid content were calculated (26).
The frequency of eating pure salt was obtained from questions during a semi-structured interview with the mother and child on the child's liking of salt and sweet foods, food preferences and restrictions (diets), forms of ingesting salt, ways of coping with the disease, and compliance in taking medication.
There were three parts to this measure, including number of sweet and salty snacks eaten, preferred concentrations of salt and sugar, and assessment of hedonics and intensity of oral sprays of NaCl solution.
During the interview, participants were encouraged to eat freely from four familiar commercial salty or sweet snack items presented on separate saucers in unwrapped bite-sized morsels (Table 2). The number of salty and sweet items eaten served in the analyses.
Preferred concentrations of NaCl in soup and sugar in tea (6, 15, 19, 20, 28) were assessed as follows. Tomato soup was prepared by diluting one part unsalted tomato paste concentrate (22BX) with nine parts water. A pinch of dried basil was also added. Participants were presented with two 100-ml cups of hot (∼45 ± 5°C) tomato soup, one unsalted and one with 3.3% (wt/wt) NaCl. They were asked to taste both cups, and then, using a 5-ml teaspoon, they were asked to mix them in a third cup until the soup was “most tasty.” The sodium content of the mixture was analyzed. The procedure for determining the preferred concentration of sugar in tea was similar, with one cup containing unsweetened tea and the other cup containing tea with 30% (wt/wt) sucrose.
Both tests were repeated, with 5 min between tests. Scores were averaged for the repeated tests.
Assessment of hedonics and intensity of oral sprays of NaCl solution were made using visual analog scales (VAS). Six concentrations of NaCl solution (0.0025, 0.01, 0.04, 0.16, 0.64, and 2.56 M) were sprayed (0.29 ml volume) into the oral cavity in ascending and descending order. Participants marked a VAS for intensity (anchored by “do not taste anything” and “very, very strong taste”) in Hebrew and another for hedonics (anchored by “most dislike this taste” and “I like this taste very, very much”) for each concentration. The two scores of each concentration were averaged. Because the three lowest concentrations were not distinguished (see Fig. 1), the mean of the three highest concentrations served for the appetite analyses.
We calculated salt appetite using the scores of the seven tests. The overall preference for sweet served as control and was derived from the five analogous measures to the first three tests, as well as the fifth and sixth test, above.
The interrelationships of the various measures were analyzed for a larger sample and are presented in General Salt Appetite.
In addition, participants' mothers completed a questionnaire rating the degree of nausea and vomiting they experienced, by trimester, while pregnant with the child. The severity and frequency of vomiting that the child experienced and the severity and frequency of diarrhea in infancy (birth to 6 yr of age) and childhood (6–12 yr) were also queried, which accounted for general tendencies (e.g., car sickness) and particular periods (e.g., illness). Mothers were also queried about whether the child experienced dehydration or hyponatremia that required saline infusions. These data were scored on ordinal scales and collected to determine the contribution of putative dehydrational events to offspring salt appetite (3, 9, 20, 21, 39).
Height, weight, and blood pressure were measured, and body-mass index (BMI) was calculated.
Serum concentrations of 17-OHP and aldosterone were measured by RIA with the use of commercial assays (DPC), ACTH levels with Immulite (2000, DPC), and PRA with RIA (Diasorin). Blood (∼6 ml) and urine samples were analyzed for levels of creatinine, sodium, and potassium. Fractional excretion of sodium was calculated.
Mutation screening was preformed for all subjects (patients and siblings). DNA was extracted from the peripheral blood leukocytes in accordance with standard methods. PCR amplification of two fragments, extending over the entire gene, and primers that distinguish the gene CYP21B from the pseudogene CYP21A were used. The set of probes was developed to detect the nine common mutations: V281L in exon 7, P30L in exon 1, I172N in exon 4, the splice site mutation in intron 2, the cluster of three mutations (1236N, V237E, M239K) in exon 6 designed as CL6, Q318X stop codon in exon 8, and the 8-bp deletion in exon 3 (16). Table 3 presents the individual mutations in our population.
Analyses employed SPSS 10 and 11, with the different tests employed as reported. Three analyses of variance (ANOVAs) were necessary to compare the various groupings: 1) controls, SV, SW; 2) controls nSW, uSW, SV; and 3) controls, uSW, uSV. Least significant difference served for a priori-determined post hoc comparisons. The Mann-Whitney U-test served for analyses of ordinal data and for some cases where group sizes were small and individual scores of some measures disparate.
Alpha was determined at 0.05, and SE is the measure of variability throughout this report.
To derive salt appetite, the different measures were equated by dividing each participant's score by the highest score for that measure. In turn, these scores were averaged to derive the salt appetite score for each participant. Sweet preference was similarly obtained.
Participants were invited to the pediatric endocrinology unit of the hospital. They answered the questionnaires and did the behavioral tests, were interviewed, and were given a clinical examination by the pediatric endocrinologist. Blood and urine samples were collected. Sessions lasted ∼2 h and were always in the mornings between 0800 and 1000. Testing started or ended (randomly) with the clinical evaluation and collection of blood and urine samples. The behavioral testing started with the interview, followed by the questionnaires, during which time the snacks were offered, and then the other taste tests. After the sessions, there was a debriefing conversation during which the aims of the experiment were explained and participants and their parents asked questions and offered additional information about the disease and coping, etc. All the tests and questions about salt were embedded in other tests and questions on taste and dietary preferences to ensure that the participants were not aware that the focus of the research was salt. Debriefing confirmed this; the patients understood the research to be on taste and food preferences.
Because parents often introduced the children as healthy or CAH, the study was not fully blind. However, the study was blind with respect to the classification of the CAH form, as well as therapeutic adequacy, both of which were unknown at the time of testing.
Salt Appetite in CAH
Salt appetite and sweet preference.
The results of the one-way ANOVAs for evaluation of salt appetite are presented in Table 4. Table 4 shows that SW and uSW had a significantly greater sodium appetite than controls, even though not all the subtests showed a similar trend. uSW salt appetite was greater than that of SV or uSV. There were no significant differences between the groups in sweet preference.
Hedonics and intensity of oral NaCl spray.
Figure 1 presents the results of the VAS scores. For the hedonic VAS, the three highest concentrations were combined, and SW rated them as less aversive than SV (P < 0.05, Mann-Whitney U-test). There were no group differences in intensity ratings.
Ingestion of solid salt.
Thirteen participants reported licking salt: 1 of 18 controls and 12 [SV = 5, SW = 7 (uSW = 6)] of 26 CAH patients (χ2 = 8.42, P < 0.01). Seven participants reported eating salt “sometimes” and 6 reported eating it “frequently.” They reported a variety of ways they ingested salt, including licking from the palm of the hand, wetting a finger in the mouth and then dipping it into the salt dish,1 pouring salt from the salt holder onto the finger or using a finger to gather spilled salt from the table, drinking the salt brine from pickle jars, eating the salt crumbs from salt snack bags, and liking particularly salty food items.
Three uSW also reported additional unusual forms of eating salt such as licking lemon salt, salting foods such as bread, cheese, instant coffee powder (and eating it), figs, green almonds, oranges, and licking salt off a cut lemon. One child reported eating sugar with salt.
Asked when they discovered they liked salt, one (uSW) reported eating salt at 2 yr old (according to his parents, he was already drinking pickle brine at that age) and 12 reported starting after 6 yr old. Seven participants reported discovering they liked salt by themselves, and six reported having been shown the habit by siblings, relatives, or friends; one participant was prescribed salt pills with his medication by his pediatrician.
All but two parent couples (one of a control) disapproved and were angry with their children for eating salt. Some parents did not realize their children might require salt because of their illness.
Relationship of salt appetite and physiological measures.
Table 5 presents the physiological data. Compared with controls and with nSW, uSW have elevated ACTH, (for which they were selected), 17-OHP, plasma K+, and urine Na+-to-K+ ratio and lower urine K+. Plasma aldosterone did not differ from controls in the SW groups. Sodium levels are normal in all CAH. The only evidence for electrolyte imbalance is lower urine K+ and higher Na+/K+. The higher BMI of uSV may be related to their earlier maturation due to virilization, to increased glucocorticoid treatment in this group (5, 18, 37, 38), or to their higher average age [although it did not differ significantly from the other groups (Table 1)].
Table 6 presents correlations of the physiological measures with salt appetite. In SW, there is a consistent relationship between salt appetite and symptomology of salt wasting, with salt appetite correlating with PRA, plasma K+, urine Na+, and inversely with BP. The significant correlation between sweet preference and plasma Na+ in SW (irrespective of therapeutic status) is inexplicable.
Dietary intake of energy, macronutrients, and most electrolytes, water, and total daily fluid was higher in uSW than in the other groups. Drinking of water was also elevated in uSV (Table 7).
The increased dietary intake in SW is likely due to elevated hormone levels or iatrogenic glucocorticoids (5, 38), although the increase in BMI was not significant. Dietary intake and BMI are further dissociated in uSV who did not show increased appetite but did have a higher BMI. As mentioned, this may be due to the greater age of uSV.
It should be noted that the increased dietary Na+ intake of the SW is not significant when caloric intake is held constant.
There was an inverse correlation of BMI with salt appetite and sweet preference in the uSW and SW categories, respectively.
The pattern of drinking is consistent with hypovolemia in uSV, SW, and uSW. Over all 44 participants, drinking of water correlated with salt appetite (r = 0.35, P < 0.05), whereas drinking of sweet drinks correlated with sweet preference, (r = 0.34, P < 0.05) (with sweet drinks excluded from the preference measure).
General Salt Appetite
These analyses of salt appetite utilized the full data set of 82 participants [age 13.6 ± 0.6 yr old (range 6–30 yr old)], with one-half of them CAH (44 Arabs and 38 Jews, 52 females and 30 males).
Analyses of salt appetite.
Factor analyses of the tests yielded four components that confirmed a fair dissociation of salt and sweet preference. Components 1 and 2 comprise tests of salt appetite and sweet preference, respectively, together explaining 38% of the variance. Components 3 and 4 comprise dietary intakes and intake of snacks, respectively, that do not distinguish between salty and sweet (Table 8).
Salt appetite and “biography of dehydration.”
Salt appetite correlated with prenatal and postnatal episodes of putative sodium loss (3, 20, 39) (Table 9). The recall data could not reveal the reasons for the infusions (clearly not all were for dehydration or hyponatremia), but the correlation with salt appetite was obtained from both controls (r = 0.40, P < 0.01) and CAH (r = 0.35, P < 0.05), although more CAH received infusions than controls (23/41 vs. 14/41, χ2 = 4.0, P < 0.05) and had them more often (4.8 ± 0.8 vs. 2.1 ± 0.6, P < 0.01).
There were no significant correlations with sweet preference.
We find that SW and uSW individuals have a greater sodium appetite than their healthy siblings. Comparisons with other patient groups showed that the increased appetite was not related to CAH per se but specifically to the SW group of patients that were not clinically normalized. In terms of our measures of sodium appetite, they scored ∼140–170% higher than sibling controls and other patients. This did not occur with sweet preference; thus this is not due to a general effect on taste (13, 14).
SW and uSW patients reported that they liked salt more, that they used it more often, and that they ate solid salt more frequently. In a test of preferred levels of salt in soup, these patients added 130–160% more NaCl than other patients or sibling controls, and they found high concentrations of oral NaCl solution less aversive. This last was not due to altered sensory evaluation of the intensity of salt taste, which was identical in all participant groups, unlike previous reports of Addison's patients (13, 14).
The dietary questionnaires of SW and uSW also revealed a greater intake of dietary sodium, but this could not be dissociated from intake of other macronutrients and electrolytes and is likely a function of a greater nutrient intake in SW, possibly because of glucocorticoid-induced hyperphagia (37, 38).
In less quantified form, a measure of the increased appetite is that a much higher proportion of CAH patients reported ingesting salt, including licking crystalline salt and drinking brine from salt-pickle jars. Three uSW patients also reported unusual forms of salt ingestion (not reported by healthy siblings or other CAH), including salting fruit, eating salt with instant coffee powder, and eating salt with sugar. It is noteworthy that these adolescents' motivations to eat salt were expressed despite their parents' disapproval and anger.
We interviewed the participants to understand how they first discovered their love of salt. CAH overwhelmingly recounted initiation of salt eating after school age (>6 yr old) rather than infancy; one-half reported that they learned to do so from others, whereas the other one-half discovered themselves that they liked salt. One SW patient remembered well her revelation, that of her first taste of pure salt in a kindergarten taste experiment. These observations suggest that eating salt to ameliorate or prevent hyponatremic crisis is an acquired strategy, rather than innate in the sense that is appears in infancy. However, it could also be that, although the appetite itself is not innate, the benefits of salt need to be discovered and learned, and there may be an innate predisposition for such learning (29, 31, 32, 48). Alternatively, the late discovery of salt could be because younger infants have less access to salt or because parents may be controlling medication better in younger infants.
Patient MM was 12.3 yr old and raised as a boy, although chromosomally 46XX (after this study, he received dispensation for gender-change surgery from his religious court). He was a SW patient, nonnormalized for the 18 mo of this study. He likes salt very much but not sweet. He licks salt frequently, from his hand and from the salt cellar, and puts a lot of salt on his food. He eats a lot of pickled (in salt-brine) olives and cucumbers and salty cheeses. He eats lemons, figs, and instant-coffee powder with salt. His parents are permissive of these habits. Patient MM discovered his liking for salt by himself. Patient UM, 9.8 yr old, his brother, is also uSW, with similar habits, which he learned from patient MM. In the tomato soup test, he liked best the undiluted high concentration (3.3% NaCl, more concentrated than sea water). Their cousin, patient NM, 6.4 yr old, also uSW, learned to eat salt from his cousins patients MM and UM, but his parents object to the habit. Three of their siblings were controls, all heterozygote carriers.
What might be the physiological mechanism driving this increased avidity for salt? Among SW patients, salt appetite correlated with symptoms of salt wasting, namely PRA, plasma K+, urine Na+, and (inversely) with BP. The central renin-angiotensin system and peripheral aldosterone work in concert to arouse sodium appetite in sodium-depleted animals (4, 7, 8, 17, 23, 32, 33, 35, 43), and the peripheral renin-angiotensin system has also been implicated (7, 8, 43). Thus heightened PRA could contribute to the appetite by increased ANG II, in turn increasing the palatability of sodium (1, 7, 11, 17, 23, 31, 33, 48).
It is noteworthy that aldosterone did not correlate with sodium appetite, nor did ACTH, which may also arouse sodium appetite in animals (7, 47). Finally, we have no explanation for the correlation of sweet preference with plasma Na+ in SW, but it may caution against the vagaries of small-sample correlations.
The differences between normalized and nonnormalized patients may be instructive in understanding the nature of the increased appetite. All SW patients were recommended medication; however, of the 14, only 5 were therapeutically normalized. Because the severities of the mutation of nSW and uSW were quite similar (unlike nSV and uSV), it is possible that the difficulty in achieving stabilization is compounded by noncompliance, i.e., most SW preferred eating salt to taking medication. Similar rates of compliance, as assessed by physicians, have been noted (compliance was “good” in 28 and “bad” in 37 of 65 patients) (5). Noncompliance is quite persistent: in an attempt to obtain within-subject comparisons of the appetite, we retested eight nonnormalized patients 3–6 mo later; however, only two had normalized (during that time, another normalized patient became nonnormalized).
Thus, even though medication is better for patients in addressing their pathologies, most patients seem to prefer living on the edge with salt vs. taking their medication. This suggests that self-medication with palatable salt might be inbred in humans as it is in sodium-depleted animals. Humans might be biologically and innately predisposed to learn that salt remedies their affliction (4, 7, 33, 48), whereas learning to take medication has to be achieved without such an evolutionary prop (29). These speculations can have implications for compliance; they may provide doctors and patients with insight into the difficulties of compliance and, based on learning theory, might suggest novel strategies to improve it, such as flavoring medication with salt.
For the entire population of 82 participants, factor analysis indicated a fair dissociation between salt and sweet preference. This is useful evidence for salt appetite in humans as a discrete behavioral category, such as is familiar in animals, rather than a general preference for high levels of seasoning or strong flavors (25, 48).
However, the analyses also suggested that not all forms of salt ingestion are related to a salt appetite: intake of salt and sweet snacks and dietary intake of Na+ and sweet carbohydrates are largely independent of salt appetite. The latter are probably determined by food intake rather than sodium requirement. Nevertheless, in SW patients, increased sodium intake from these sources may contribute to preventing and alleviating hyponatremia.
We also show that maternal nausea and vomiting during pregnancy, childhood vomiting and diarrhea, and infusions in childhood correlated with increased salt appetite in offspring. There were no significant correlations with sweet preference. These results are consistent with previous findings in humans and rats showing that perinatal sodium loss increases salt appetite enduringly (3, 20, 21, 23, 27, 39, 48).
Together, our findings suggest that CAH adolescent patients who are SW but not therapeutically normalized because of poor compliance have a greater salt appetite. It is characterized by an increased avidity for salt in food; they find it more palatable and less aversive at high concentrations, and half of the patients have a habit of eating pure salt in ways that are rare in healthy controls. They ingest more dietary sodium than their siblings, which, although it may be due to hyperphagia rather than salt appetite, may serve them equally well in allaying their sodium loss.
The increased appetite is multifactorial in origin. Perinatal episodes of sodium deficit would enhance the appetite by imprinting (4, 30). Most SW report initiation of salt intake at school age rather than at infancy, and all report that they learned to do so. Thus eating salt to prevent, or ameliorate, hyponatremic crisis is also an acquired strategy. It may be acquired by advice from others, such as physicians and relatives with CAH, and by conditioning to the postingestive remedial effects of salt in hyponatremic crisis. In addition, PRA levels and salt appetite correlate; thus renin- or deficit-heightened sodium palatability could reinforce these learning processes (1, 11, 23, 31, 33, 48). Finally, the learning could also be facilitated by an innate predisposition to learn the benefits of salt to the depleted body (2, 29, 31, 33).
These findings and ideas might provide helpful insight into the difficulties of compliance in CAH and advance our comprehension of salt appetite in humans and our predilection for excess salt, with its attendant health risks for the vulnerable.
This work was supported by a Ministry of Health Grant to M. Leshem and A. Kochli and by Israel Science Foundation Grant 902/00 to M. Leshem.
↵1 In many Galilean homes, salt is available at table in a dish, rather than in a salt cellar.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society