ADH due to neurological lesions, are less likely to be the cause of hyponatremia unless there are positive clinical features. Among 50 elderly hospitalized patients with mild to moderate hyponatremia, an exhaustive evaluation did not reveal these causes of inappropriate ADH syndrome. The investigators found that the hyponatremia was associated with pneumonia and medication, although 60% remained idiopathic.27
Many patients may have primary orthostatic hypotension, e.g. due to autonomic neuropathy in Parkinson’s disease and multiple system atrophy, or associated with low renin‐low aldosterone mineralocorticoid deficiency. Older patients may have excessive treatment of their hypertension. The elderly should be monitored for orthostatic blood pressure changes and have more moderate adjustment of systolic hypertension than younger individuals.28
Other medications not specifically used in the elderly are associated with hyponatremia. Anti‐neoplastic agents include vincristine and cyclophosphamide. Vincristine may cause hypothalamic neuropathy, and cyclophosphamide treatment may potentiate the ADH effect at the renal tubule and requires patients to drink large volumes of water to prevent cystitis.24 Uncommon causes of hyponatremia associated with common drugs include non‐steroidal anti‐inflammatory agents (NSAIDs), which may lower the levels of prostaglandins. Prostaglandins have an anti‐ADH effect on the renal tubules, so lowered levels result in potentiated ADH action.2 Trimethoprim‐sulfamethoxazole may act as a mild diuretic and cause hyponatremia if given in high doses or when given to a patient with renal impairment.24
Hyponatremia is associated with a poor overall prognosis.3,4 There may be two temporal components of the hyponatremia syndrome in the elderly: a chronic hyponatremia syndrome associated with underlying diseases or dementia and an acute deterioration toward hyponatremia as manifested by delirium. However, these may be difficult to distinguish. Hyponatremia is associated with gait disturbances, falls, and bone fractures.3,4 In case‐control studies (controlled for age and admitting diagnoses), the presence of hyponatremia was associated with an increased incidence of falls in patients admitted to an emergency medical department29 and hospitalized individuals.30 In another case‐control series, the presence of hyponatremia was one component of fall prediction in addition to age and pre‐existing neurological disorders.31 Furthermore, people with hyponatremia had a 2.36‐fold higher chance of also having dementia after correcting for age, gender, medications, and comorbidities.32
Subsequent case‐control studies (adjusting for age)33,34 and meta‐analyses35‐37 found higher incidences of bone fractures associated with hyponatremia. There may be a direct relationship of hyponatremia with activation of osteoclastic bone resorption. In a rat model of exogenous inappropriate ADH, Verbalis38 demonstrated that chronic hyponatremia is associated with a substantial reduction in bone mass, as determined by bone density and histomorphology. However, the case studies do not resolve the issue of assigning causality of the hyponatremia with adverse events. Many of the patients with falls and hyponatremia had higher incidences of underlying reasons for falls when compared with case controls.34,37 Individuals with hyponatremia also were more frequently taking SSRIs (21% vs. 15%, p = 0.006), benzodiazepines (39% vs. 31%, p = 0.007), or other CNS class drugs (59% vs. 49%, p = 0.0004) or had cognitive impairment (17% vs. 7%, p < 0.0001) or orthostatic hypotension (6% vs. 2%, p = 0.003).33 The simultaneous use of medications associated with neurocognitive impairment and hyponatremia may explain the association of concurrence of hyponatremia with the high rate of falls.
Therefore, it is important to assess the role of hyponatremia within the context of a cause‐and‐effect relationship and determine whether the neurocognitive symptoms are reversible with correction of the hyponatremia independent of correction of the underlying illness or removal of potentially causative medications. Renneboog et al.29 performed a retrospective analysis of patients admitted to an acute care hospital with serum sodium levels between 115 mEq/L and 132 mEq/L. In those with hyponatremia, there was a 21% incidence of falls compared with 5% in the controls (p < 0.001). The researchers then prospectively tested eight subjects with hyponatremia for abnormalities in gait and cognition. The tests for gait and cognitive dysfunction were compared prospectively when they had hyponatremia (serum sodium 128±3 mEq/L) and after correction of hyponatremia (138±2 mEq/L). There were significant improvements in gait abnormalities after correction of hyponatremia (as determined through a standardized test of distance travelled by tandem gait on a pressure‐sensitive calibrated platform). There were improvements in cognitive tests (determined by decreased reaction time on attention tests (from 673 ± 182 msec vs. 615 ± 184 msec, p < 0.001). Despite Renneboog’s finding of an increased incidence of falls, it was not determined whether the falls were due directly to the hyponatremia or whether the hyponatremia was a marker of the severity of the underlying illness or the doses of the offending medication causing the hyponatremia.
Therefore, it was hoped that the issue of causation of mental status changes with hyponatremia would be resolved by the correction of the hyponatremia independent of modifications of either the underlying disease or medication adjustment. Prospective studies using vasopressin antagonists (see the discussion of vaptans below) have shown only modest clinical benefits with correction of hyponatremia.
Lixivaptan, an oral vasopressin receptor2 (V2) antagonist (not approved in the US for hyponatremia), increased the serum sodium by 3 mEq/L over that of controls.39 There were small differences at various time points but no overall benefit compared with controls in a standardized Trail Making Test or Medical Outcomes Survey‐6 Cognitive Function Scale (MOS‐6) score at 28 days.
Tolvaptan, a specific vasopressin receptor antagonist, was assessed in a combined randomized prospective international multicenter study of 448 subjects, the SALT‐1 (US cohort), and SALT‐2 (international cohort).23 There was a preplanned 30‐day analysis of a standardized quality of life scale, the Medical Outcomes Study 12‐item Short Form of the General Health Survey [SF‐12]. Although subjects were considered to have asymptomatic hyponatremia, about one‐third had cirrhosis, one‐quarter had congestive heart failure, and 40% were diagnosed with SIADH. Baseline quality of life scores for both physical and mental components were lower than the 25th percentile for age‐adjusted measures. In the combined study, there was a small improvement in the overall SF‐12, p = 0.02. However, the changes were significant in the SALT‐1 sub‐study (US, baseline score 42.3 ± 11.7, treatment effect 3.9, p = 0.04) but not in the SALT‐2 sub‐study (international, baseline score 44.7 ± 12.0, treatment effect 2.2, p = 0.15). There were no changes in the physical component summary of the SF‐12. The benefits in the SF‐12 were more relevant in those with marked hyponatremia (baseline serum sodium <130 mEq/L, p = 0.04).23 Changes in the SF‐12 were not considered in the primary outcome analysis. It is unclear whether there are any clinically relevant benefits of this modest increase in the quality of life score or that the desired improvement in hyponatremia could override other underlying disease symptoms. There was also no predesigned assessment of gait, balance, or cognitive functions.
The neurocognitive effect of tolvaptan was further assessed in 107 patients with asymptomatic hyponatremia with the sole diagnosis of SIADH.40 Improvement in serum sodium from 129 mEq/L to 136 mEq/L did not improve the primary cognitive endpoint (composite score on the Cognitive Research system) reaction times, although there was a change in a subset psychomotor speed domain.
A systematic review of the vaptans found that all vasopressin receptor antagonists were beneficial in improving serum sodium, with limited evidence of impact on quality of life or mortality.41
Workup and treatment of hyponatremia
There may be a recognized history of excessive water intake that overwhelms the ability of the kidneys to clear free water, such as primary polydipsia (purposeful intake of 16–20 litres per day), so‐called beer potomania (beer intake of similar volumes), or excessive water intake during prolonged exercise (e.g. in long‐distance runners).1,8 However, without that history, the usual cause of hyponatremia is increased water retention, although there may also be a component of solute loss. Tests should be obtained of the serum for basal metabolic