Reduced cortical activity due to chronic low blood pressure: An EEG study

Abstract

Alterations in cortical activation processes due to chronic low blood pressure were investigated. In 40 hypotensive subjects and 40 normotensive controls, the contingent negative variation (CNV), induced by a constant foreperiod reaction time task, was assessed at nine scalp sites (F3, Fz, F4, C3, Cz, C4, P3, Pz, P4). Additionally, spontaneous EEG was recorded at resting conditions. In hypotensives, a reduced amplitude of both the early and the late component of the CNV were found at Cz. At Fz the early CNV was reduced. Hypotensives exhibited longer reaction times, and the reaction time was negatively correlated with the CNV amplitude. Resting alpha power correlated negatively with blood pressure. The findings can be related to cognitive deficits due to hypotension found in earlier studies. The effects of hypotension on cortical activity are discussed to be mediated by afferents from the cardiovascular system to the prefrontal cortex as well as by reduced cerebral blood flow.

Introduction

The concept of essential hypotension refers to a chronic condition of lowered blood pressure, independent of the presence of any other pathological factors. According to the WHO (1978), hypotension is diagnosed if systolic blood pressure falls below 110 mmHg in males and below 100 mmHg in females, regardless of the diastolic reading. Chronically low blood pressure is widespread. For the general population, a prevalence between 5 and 6% was reported with younger women being especially affected (Boschke, 1982, Boschke, 1983, Luft, 1981, Duschek and Schandry, 2005). Typical subjective symptoms include fatigue, reduced drive, faintness, dizziness, enhanced urge for sleep, headache, palpitations and cold limbs as well as cognitive impairments (e.g. Weiß and Donat, 1982, Cadalbert, 1997).

In contrast to elevated blood pressure, hypotension is not regarded as a dangerous medical condition. However, its considerable impact on subjective well-being has been shown in several large, population-based studies. For instance, impairments of general health (Wessely et al., 1990, Pilgrim et al., 1992) as well as quality of life (Rosengren et al., 1993) were reported. Barrett-Connor and Palinkas (1994) found higher scores of depressiveness due to low blood pressure. Moreover, chronic hypotension has been identified as a major risk factor in pregnancy (Ng and Walters, 1992, Warland and McCutcheon, 2002). In studies focusing on the elderly population, associations between hypotension and the prevalence and incidence of dementia were reported (e.g. Morris et al., 2002, Ruitenberg et al., 2001).

In a number of more recent studies, a reduced cognitive performance in essential hypotension was reported (for overview see, e.g. Duschek et al., 2003). The deficits are particularly prominent in the area of attention, which was established based on a variety of paper pencil tests (Richter-Heinrich et al., 1971, Stegagno et al., 1996, Costa et al., 1998, Duschek et al., 2003), as well as on computer-based assessment instruments (Weisz et al., 2002, Duschek et al., 2005). Duschek et al. (2005), for instance, applied a comprehensive, computerized test battery (Zimmermann and Fimm, 2002), and found poorer performance in tasks aiming at the specific attentional components of tonic and phasic alertness, selective attention, divided attention, sustained attention (vigilance) and working memory.

The EEG provides useful information about cerebral functioning, which can be related to attentional processes. The present study focused on the contingent negative variation (CNV) as well as on the spectral frequency content of the spontaneous EEG in hypotension. The CNV is an extensively studied, slow event-related potential, which occurs during the period between a warning signal (S1) and a second stimulus (S2) demanding a motor, verbal or cognitive response (Walter et al., 1964, Tecce and Cattanach, 1993, Gómez et al., 2004). The negativity is considered as indicating periods of increased cortical excitability during the preparation for demands on the information processing system (Rockstroh et al., 1989, Birbaumer et al., 1990, Elbert et al., 1991).

Some evidence for the relation between the CNV and attentional performance stems from pharmacological studies. The effects of many stimulant and sedative drugs on attention involve an increase, respectively decrease, of the CNV amplitude (e.g. Tecce et al., 1978, Coons et al., 1981, Fattaposta et al., 1984). Moreover, attentional deficits due to brain lesions or psychosurgical procedures are associated with a reduced CNV (Zappoli et al., 1976, Rizzo et al., 1978). Correlations between the extent of attentional impairment and attrition of the CNV amplitude were also reported in affective disorders (Claverie et al., 1984) and brain lesions resulting from alcohol abuse (Chao et al., 2003). In experiments with healthy subjects, the presentation of certain distracting stimuli during the interval between S1 and S2 lead to disruption of the CNV, which has been interpreted as being caused by an interference with attention (e.g. Tecce and Scheff, 1969, Brix et al., 1979, Rizzo et al., 1984).

In the neuropsychological literature, the CNV is usually linked to the concept of “phasic alertness” (e.g. van Zomeren et al., 1984, van Zomeren and Brouwer, 1994). Phasic alertness denotes the sudden increase of attentiveness that immediately follows a warning signal, which requires a quick response (Posner and Rafal, 1987). In more detailed concepts, the CNV has been associated with multiple sensory, cognitive and physiological processes relating to the analysis of the warning signal, the anticipation of the imperative stimulus and the preparation of the behavioral response (e.g. Birbaumer et al., 1990, Brunia, 2003).

For longer inter-stimulus intervals, two components of the CNV can be distinguished: an early wave with frontal scalp predominance, and a late wave, which is maximal at the vertex (Andreassi, 2000). The early CNV is thought to reflect the processing of information provided by the warning stimulus, as well as activity associated with response selection (Rohrbaugh and Gaillard, 1983, Gevins and Cutillo, 1986, Gaillard and v. Beijsterveldt, 1991, van Boxtel et al., 1993). Moreover, as cited in earlier literature, this component was related to tonic cortical arousal. Tecce (1972) postulated an inverted U-shaped relationship between the early CNV amplitude and the arousal level, which has been confirmed in some of the more recent experiments (Higuchi et al., 1997a, Higuchi et al., 1997b). The late CNV, also referred to as the “stimulus preceding negativity” (Brunia, 1988), has been linked to anticipatory attention towards the imperative stimulus as well as to cognitive and motor preparation of the response (Rohrbaugh and Gaillard, 1983, Rosahl and Knight, 1995, Kotani and Aihara, 1999). In tasks which require a motor response to S2, the late component is superimposed by the readiness potential (RP), which precedes self-paced movements (Kornhuber and Deecke, 1965). Nevertheless, according to contemporary knowledge, the late CNV and the RP can be viewed as different brain electric phenomena (Tecce and Cattanach, 1993, Brunia, 2003).

Two earlier studies provided the first hints for a lower CNV due to hypotension. In a pioneering study, Costa et al. (1998) identified a reduced amplitude in a constant foreperiod reaction time task (light and tone as S1 and S2). A relatively short inter-stimulus interval of 2 s was chosen. This was suboptimal for a discrimination between the early versus late component, as well as between the physiological and cognitive processes related to each of them. Weisz et al. (2002) used tones for S1 and S2 separated by a 4 s inter-stimulus interval, and merely found the early CNV to be significantly reduced in hypotensives. The authors could not provide an explanation for the specific reduction of the early wave. The missing effect on the late wave can possibly be ascribed to the relatively high blood pressure in their hypotensive sample. The mean systolic reading in the exclusively female group was 102 mmHg, and almost two-thirds of the subjects exceeded the criteria of a maximal blood pressure of 100 mmHg as suggested by the WHO. Both Costa et al. (1998) and Weisz et al. (2002) assessed the CNV only at the vertex. Especially, because of the frontal scalp distribution of the early CNV, a multiple electrode array also including frontal sites would be more appropriate.

The power spectrum of the spontaneous EEG provides a useful tool for the assessment of tonic cortical arousal (Andreassi, 2000, Davidson et al., 2000, Stern et al., 2001). It can be linked to “tonic alertness”, which implies a generalized physical and mental state of activation (van Zomeren et al., 1984; Posner and Rafal, 1987). The preponderance of alpha waves indicates a comparatively low activation, whereas higher levels of tonic alertness are accompanied by higher portions of beta activity. Concerning hypotension, no data on spontaneous EEG activity have been reported so far. Given the subjectively reported fatigue (e.g. Wessely et al., 1990) and the reduced performance in cognitive tasks aiming, e.g. at tonic alertness and vigilance (Duschek et al., 2005), a reduced level of cortical activation may be assumed.

The purpose of the present study was to establish brain electric correlatives of the cognitive deficits due to chronic low blood pressure. Based on the reported deficits in tonic and phasic alertness, reduced cortical activation processes can be assumed to be related to hypotension. This reduction should be expressed in alterations of the CNV as well as in spontaneous EEG activity. Considering this, as well as the results of the former EEG studies on hypotension which have been mentioned, the following two hypotheses may be deduced. (1) The early as well as the late component of the CNV are reduced in essential hypotension. (2) Hypotensive persons exhibit a higher portion of resting EEG alpha power and less beta power.