OUP user menu

Language dominance in neurologically normal and epilepsy subjects
A functional MRI study

Jane A. Springer, Jeffrey R. Binder, Thomas A. Hammeke, Sara J. Swanson, Julie A. Frost, Patrick S. F. Bellgowan, Cameron C. Brewer, Holly M. Perry, George L. Morris, Wade M. Mueller
DOI: http://dx.doi.org/10.1093/brain/122.11.2033 2033-2046 First published online: 1 November 1999

Summary

Language dominance and factors that influence language lateralization were investigated in right-handed, neurologically normal subjects (n = 100) and right-handed epilepsy patients (n = 50) using functional MRI. Increases in blood oxygenation-dependent signal during a semantic language activation task relative to a non-linguistic, auditory discrimination task provided an index of language system lateralization. As expected, the majority of both groups showed left hemisphere dominance, although a continuum of activation asymmetry was evident, with nearly all subjects showing some degree of right hemisphere activation. Using a categorical dominance classification, 94% of the normal subjects were considered left hemisphere dominant and 6% had bilateral, roughly symmetric language representation. None of the normal subjects had rightward dominance. There was greater variability of language dominance in the epilepsy group, with 78% showing left hemisphere dominance, 16% showing a symmetric pattern and 6% showing right hemisphere dominance. Atypical language dominance in the epilepsy group was associated with an earlier age of brain injury and with weaker right hand dominance. Language lateralization in the normal group was weakly related to age, but was not significantly related to sex, education, task performance or familial left-handedness.

  • cerebral dominance
  • language
  • functional MRI
  • lateralization
  • EHI = Edinburgh Handedness Inventory
  • fMRI = functional MRI
  • IAP = intracarotid amobarbital procedure
  • LI = laterality index

Introduction

Hemispheric specialization for language functions has been a central topic in brain research for well over a century (Dax, 1865). Despite the vast literature on this topic, several basic issues regarding language lateralization remain unresolved. Although it is generally accepted that most people have left hemisphere language dominance, the actual incidence of atypical dominance in the general population is not known. The risk of losing language function after right brain surgery for adult-onset disease is consequently estimated to be small, but has not been determined precisely. For the majority of the population with left hemisphere dominance, it is not known to what extent the right hemisphere also contributes to language processing. Although the dominant hemisphere is often assumed to carry out essentially all language functions, an alternative possibility is that there are varying degrees of dominance that determine vulnerability to language deficits following injury to either the right or left hemisphere. Finally, there is evidence that several genetic, developmental, environmental and pathological factors may influence language lateralization, but the power of these influences remains poorly defined.

Several techniques have been used to examine these issues. The intracarotid amobarbital procedure (IAP) (Wada and Rasmussen, 1960; Loring et al., 1992) enables testing of each cerebral hemisphere individually for language competence after temporary anaesthetization of the opposite hemisphere. It has been called `the least ambiguous method' for determining language dominance (Snyder et al., 1990) and is routinely used in the presurgical evaluation of epilepsy patients (Risse et al., 1997). Findings from IAP studies provide estimates of language dominance patterns in the general population (Hecaen and Albert, 1978; Kolb and Whishaw, 1990) and suggest that handedness, age of seizure onset and sex may all influence language lateralization. In one of the largest studies to date, Rasmussen and Milner examined the effects of early brain injury and handedness on language dominance (Rasmussen and Milner, 1977). Among 396 epilepsy patients who underwent the IAP, ~96% of right-handers and 70% of left-handers without early brain injury showed left hemisphere dominance for simple speech functions. The incidence of atypical speech representation was higher in patients with early brain injury. Results of other IAP studies (Branch et al., 1964; Mateer and Dodrill, 1983; Strauss and Wada, 1983; Rausch and Walsh, 1984; Powell et al., 1987; Rey et al., 1988; Woods et al., 1988; Zatorre, 1989; Loring et al., 1990; Helmstaedter et al., 1997; Risse et al., 1997) show that the incidence of left hemisphere language dominance ranges from approximately 63% to 96% for right-handers and from 48% to 75% for left-handed and ambidextrous patients. Investigations of sex effects on language lateralization using IAP are less common, although one study found that women with a left hemisphere seizure focus were more likely than men to show right hemisphere language functions (Helmstaedter et al., 1997). Despite the significant contributions of these studies to our current understanding of language laterality, use of the IAP is precluded in normal subjects because the procedure is invasive and has health risks (Rausch et al., 1993). Because epilepsy is associated with varying degrees of brain dysfunction and tissue damage, reorganization of language functions may occur relatively often in such patients (Woods et al., 1988). Extrapolations about the incidence of atypical language dominance in normals from the findings of IAP studies may thus be inaccurate, presumably resulting in overestimations (Risse et al., 1997).

Investigations of previously normal patients with unilateral lesions secondary to stroke show that the incidence of `crossed aphasia' (i.e. aphasia in right-handed patients after a right hemisphere injury) is <2% (Zangwill, 1967; Gloning et al., 1969; Hecaen and Albert, 1978). Compared with the IAP studies, these data thus suggest an even lower incidence of right hemisphere language dominance in right-handed individuals. Lesion studies also indicate greater incidence of atypical language organization in left-handed than in right-handed patients. Left-handed aphasic patients tend to have incomplete and unusual aphasic syndromes (Brown and Hecaen, 1976) and to recover more rapidly (Hecaen and Sauguet, 1971), although these findings may not be robust when lesion location and size are controlled for (Gloning et al., 1969; Naeser and Borod, 1986; Borod et al., 1990). Several lesion studies also suggest that men have more strongly lateralized language functions than women (McGlone, 1977, 1980), though this finding may also be explained by lesion size and location effects (Kertesz and Sheppard, 1981).

Adult-onset lesion studies may more closely reflect normal language dominance patterns than do studies of epilepsy patients, yet this method provides only an indirect and approximate measure of these patterns. Bilateral language representation cannot be determined in individual cases from a unilateral lesion. The likelihood of detecting aphasia after a lesion depends not only on the side of injury, but also on the extent and location of the lesion within the hemisphere and the sensitivity of the language tests used. Similar lesions often cause different effects in different patients, and it may be difficult or impossible to determine whether these differences reflect true variability in functional organization or simply small differences in lesion characteristics or assessment methodology. Most important, the method may underestimate the incidence of right hemisphere participation in language, since the absence of aphasia after right hemisphere injury could be due to sparing of relatively small language areas, insensitivity of the examination methods to subtle disturbances or rapid compensation for the injury by the dominant hemisphere (Albert et al., 1981).

Unlike IAP and lesion methods, which are based on inference of language localization from observation of language deficits, activation imaging techniques such as PET and functional MRI (fMRI) provide direct observation of brain activity during language processes. Such studies demonstrate left-lateralized activation in normal subjects during a variety of language tasks (Petersen et al., 1988; Frith et al., 1991; Wise et al., 1991; Démonet et al., 1992; Howard et al., 1992; Sergent et al., 1992; Zatorre et al., 1992; Price et al., 1996b; Vandenberghe et al., 1996; Binder et al., 1997). However, inferring language dominance from such data requires consideration of two critical issues. First, the activation task used may engage a number of brain systems not specifically related to language, including sensory, motor and attentional systems. Measurement of language lateralization with such techniques thus requires that activation of these systems be controlled for by a contrast or baseline task. Secondly, even when such controls are employed, it is conceivable that some areas engaged specifically by the language task may not be critically necessary for normal language performance. Thus, the extent to which activation imaging data correlate with conventional measures of language dominance based on lesion methods must be determined empirically for each language activation task.

These issues are underscored by the fact that in virtually all of the activation studies cited above, some right hemisphere activation was observed during the language tasks. How should this non-dominant activation be interpreted? Most of the reports involved averaged group data, so one possibility is that the right hemisphere activation reflects inclusion of a few right-dominant subjects in the group. However, studies that included individual data show that subjects typically have variable amounts of right hemisphere activation even when their overall pattern is left-lateralized (Hinke et al., 1993; Binder et al., 1995, 1996; Desmond et al., 1995; Xiong et al., 1998). Such findings suggest either that the controls for activation of non-linguistic systems were inadequate or that the right hemisphere typically participates in language processing even in subjects with left hemisphere language dominance. This last account also implies that language `dominance' observed in lesion studies reflects a relative rather than a complete lateralization of function (Loring et al., 1990) and that such relative lateralization may be continuously variable across individuals in a population.

Support for the idea of relative lateralization and its relationship to conventional concepts of dominance would require not only a demonstration of variable involvement of the right hemisphere in language processing, but also demonstration that the degree of right hemisphere involvement in language predicts the degree of language impairment following right hemisphere dysfunction. We provided such evidence in a previous study comparing fMRI data with the IAP (Binder et al., 1996). The fMRI tasks, based on those of Démonet and colleagues (Démonet et al., 1992), consisted of a contrast between a semantic word categorization task and a perceptual discrimination task that controls for primary sensory, motor response, working memory and attentional requirements. Variable amounts of left and right hemisphere activation were observed using this task contrast in 22 consecutive epilepsy patients. A continuous measure of relative lateralization computed from each subject's fMRI activation map was strongly correlated with relative interhemispheric asymmetry of language deficits observed during the IAP. Thus, patients with relatively more activation in the right hemisphere during fMRI had relatively more language deficit during right hemisphere anaesthetization in the IAP. These results strongly support a model of language dominance based on variable degrees of relative lateralization. They also demonstrate that activation measured by this fMRI procedure is of direct functional significance and thus related to conventional (i.e. lesion based) measures of language dominance.

In this report we present fMRI data, obtained using the previously described language activation protocol, from a consecutive series of 100 right-handed, healthy subjects and from a consecutive series of 50 right-handed epilepsy patients. Our first goal was to describe the degree of variability in language lateralization that occurs in neurologically normal, right-handed subjects and, more specifically, to estimate the incidence of `atypical' language dominance in this group. Our second goal was to compare language lateralization in the normal and epilepsy groups. Many previous estimates of the incidence of atypical language dominance were based on studies of epilepsy patients. We assessed the validity of this approach by testing the hypothesis that atypical dominance is more frequent in epilepsy patients than in normal subjects. A third goal was to identify subject variables associated with the direction and extent of language lateralization. The only variable examined thus far using functional imaging methods in the normal population was subject gender, and these studies produced disparate findings (Buckner et al., 1995; Shaywitz et al., 1995; Price et al., 1996a; Pugh et al., 1996; Frost et al., 1999). The effects of early brain injury on language organization have not been studied with functional imaging. Based on data from previous IAP studies, it was predicted that either the presence of a left hemisphere seizure focus or an earlier age of seizure onset would be associated with a higher incidence of atypical lateralization in the epilepsy group (Rasmussen and Milner, 1977; Mateer and Dodrill, 1983; Rey et al., 1988; Woods et al., 1988; Strauss et al., 1990; Satz et al., 1994; Helmstaedter et al., 1997; Risse et al., 1997). Additional variables examined in one or both subject groups included age, sex, degree of right-handedness, presence of left-handed family members, education, level of task performance and IQ. For each of the three goals, parallel analyses were conducted using both a continuous measure of activation asymmetry as an index of language dominance and a more conventional categorical measure of dominance derived from the activation asymmetry data.

Method

Participants

The normal group consisted of 100 healthy subjects (48 women, 52 men) with no history of neurological or psychiatric disease. The epilepsy group consisted of 50 consecutive patients (29 women, 21 men) who met the following selection criteria. All patients were undergoing comprehensive evaluation for surgical treatment of epilepsy that was medically intractable, restricted their lifestyle and had a probable focal onset. In addition to fMRI, the presurgical evaluation for epilepsy subjects included long-term in-patient video-EEG monitoring, neuropsychological testing, routine brain MRI and a standardized IAP. All had a full scale IQ ≥70 as measured by the Wechsler Adult Intelligence Scale—Revised (Wechsler, 1981). All normal and epilepsy subjects were right-handed as defined by a handedness quotient ≥50 on the Edinburgh Handedness Inventory (EHI) (Oldfield, 1971). Subjects with inadequate performance on the activation tasks were excluded. Inadequate performance was defined as below the 5th percentile of a normative group for the tone decision task and below chance-level for the semantic decision task. English was the primary language of all subjects. Subjects were recruited on a voluntary basis. After full explanation of the risks and purposes of this study, all subjects gave written informed consent according to institutional guidelines. The study was approved by the Human Research Review Committee of the Medical College of Wisconsin.

Demographic information is summarized in Table 1. The two groups did not differ in handedness, sex or familial sinistrality—defined as having at least one left-handed parent, grandparent or sibling. The normal group was significantly younger (t = –10.18, P < 0.001) and had more years of formal education (t = 4.34, P < 0.001) than the epilepsy group. Information on IQ and neurological history in the 50 epilepsy patients is provided in Table 2. Seizure localization data are provided for the 41 patients who underwent surgery after fMRI.

View this table:
Table 1

Group demographic information

Normals (n = 100)Epilepsy (n = 50)P value
Mean (SD)Mean (SD)
n.s. = not significant. *EHI = Edinburgh Handedness Inventory.
Age (years)23.2 (5.0)35.8 (10.2)<0.001
Education (years)15.1 (2.7)13.2 (2.0)<0.001
Handedness (EHI score*)84.4 (14.1)87.9 (14.4)n.s.
Familial sinistrality (%)3738n.s.
Sex (% male/% female)52/4842/58n.s.
View this table:
Table 2

Descriptive information of epilepsy group

Epilepsy variableMean (SD)Range
*Based on Wechsler Adult Intelligence Scale—Revised; see text for definition and explanation.
Full scale IQ*93.6 (9.2)76–120
Age of onset (years)
Probable brain injury 9.7 (9.9)0–32
Intractable seizures14.7 (10.3)0–43
Lateralization of surgery
Left23/41(56%)
Right18/41(44%)
Localization
Temporal38/41(93%)
Extra-temporal 3/41(7%)

Image acquisition

Imaging was performed on a 1.5 Tesla GE Signa scanner (GE Medical, Milwaukee, Wis., USA) using a three-axis local gradient coil and insertable transmit/receive RF coil (Medical Advances, Milwaukee, Wis., USA). fMRI used a blipped, gradient echo, echo-planar sequence (TE 40 ms, TR 4000 ms, field of view 24 cm, matrix 64 × 64 pixel matrix). Functional slices were acquired in the sagittal plane and covered the lateral two-thirds of each hemisphere (six or seven slices per hemisphere, slice thickness 7 mm, voxel size 3.75 × 3.75 × 7 mm). High-resolution, T1-weighted anatomical reference images were obtained as a set of 124 contiguous sagittal slices using a 3D spoiled-gradient-echo sequence (GE Medical).

Stimuli and activation tasks

Stimuli were 16-bit, digitally synthesized tones and sampled male speech sounds presented binaurally at precise intervals using a computer playback system. Characteristics of the sound delivery apparatus have been described elsewhere (Binder et al., 1995). Each echo-planar image series consisted of multiple periods of linguistic activation, during which subjects performed a semantic decision task, alternating with periods of a control task, during which subjects performed a tone decision task. The rationale behind these tasks and details of the procedure were described previously (Binder et al., 1995, 1996, 1997); a brief summary follows. In the tone decision (control) task, subjects heard trains of 3–7 tones. Each tone had a duration of 150 ms and a frequency of either 500 Hz or 750 Hz. Subjects were instructed to press a button for any train containing two high-pitch (750 Hz) tones. In the semantic decision task, subjects heard names of animals (e.g. `turtle') and were instructed to press a button for animals they considered to be both `found in the United States' and `used by humans.' The two tasks were matched for stimulus intensity, average stimulus duration, average trial duration and frequency of positive targets. Responses consisted of a thumb press to a magnet-compatible response device held in the subject's left hand. Subjects were given instructions and a brief practice session before entering the scanner.

Table 3 lists the important cognitive systems thought to be engaged by the tone and semantic decision tasks. The tone task has been shown to produce activation of superior temporal auditory processing systems as well as lateral frontal lobe attentional and working memory systems. The semantic decision-tone decision subtraction produces strongly lateralized activation in prefrontal and temporoparietal heteromodal cortex of the dominant hemisphere in normal subjects, which is thought to represent engagement of language processes that are relatively specific to the semantic decision task (Binder et al., 1995, 1997).

View this table:
Table 3

Cognitive systems involved in the semantic decision and tone decision tasks

Cognitive systemTone decision taskSemantic decision task
Lexical–semantic processing+
Phonetic processing+
Attention, working memory++
Auditory processing++
Motor response++

fMRI data analysis

Identification of event-related MRI signal changes used the correlation approach of Bandettini and colleagues (Bandettini et al., 1993). In brief, this method correlates the time series echo-planar data (excluding signal equilibration images at the beginning of the series) with a reference vector on a voxel-by-voxel basis after removal of linear trends. A threshold value for the correlation coefficient is chosen to identify those voxels significantly correlated with the activation procedure. Reference vectors were derived directly from each data set using a two-step method that selects voxels strongly correlated (r ≥ 0.70) with a simple square wave model of the activation procedure (Binder et al., 1995) and then computes a data set-specific reference vector as the mean vector of this selected group of voxels. The correlation test was then performed again for each voxel using this empirically derived reference vector, with a correlation threshold of r ≥ 0.40. The probability that any given voxel will exceed this threshold by chance is approximately P < 0.0001 or P < 0.05 for each set of 500 voxels. Across all subjects, the mean number of supratentorial brain voxels included in these analyses was 7899 (SD = 916). Details of the analysis method were described previously (Binder et al., 1996).

Activation volumes were determined in each subject by counting the significantly activated voxels in the lateral two-thirds of each hemisphere. Due to the contralateral connections of the cerebral hemispheres with the cerebellar hemispheres, only supratentorial structures were included for analysis. A laterality index (LI) was calculated for each subject as the ratio [VLVR] / [VL + VR] × 100, where VL and VR are activation volumes for the left and right hemispheres. This approach yields LIs ranging between +100 (strong left hemisphere dominance) and –100 (strong right hemisphere dominance). LIs were subsequently classified as left hemisphere language dominant (defined as LI > 20), symmetric (–20 ≤ LI ≤ +20) or right hemisphere dominant (LI < –20). These cut-off points were derived statistically by determining the mean number of activated voxels across all subjects (mean = 286) and testing different hypothetical partitions of this total between hemispheres against a null hypothesis of equal activation in each hemisphere (i.e. 143 activated voxels on each side) using χ2 analysis (Binder et al., 1995). It was determined that for asymmetries greater than 173 : 113 the null hypothesis of symmetry was rejected at a probability P < 0.01. This partition of active voxels corresponds to an LI cut-off point of ±20.

A combination of non-parametric, correlational and multiple regression techniques was used to analyse the continuous variable LI. Because the distributions of LI in normals and epilepsy subjects were not normal, a non-parametric statistical procedure (Mann–Whitney U test) was used to evaluate group differences in LI between normal and epilepsy subjects, males and females, those with positive and those with negative familial sinistrality, and epilepsy subjects with left and those with right seizure focus. Because side of seizure focus was defined conservatively using the side of surgery, analysis of this variable was restricted to the 41 patients who underwent surgery. Simple correlational analysis was used to identify numerical variables associated with LI in each group. Variables examined in the normal group were age, education, EHI quotient and task performance. The relationships between these variables and a functional neuroimaging index of language lateralization have not as yet been systematically studied in a large sample of normals. Variables examined in the epilepsy group were age, education, EHI quotient, task performance, full scale IQ and ages of onset of `probable brain injury' and `intractable seizures' (see below). In the event that more than one variable was significantly correlated with LI, stepwise regression analysis (Pin = 0.05, Pout = 0.10) was used to investigate the incremental variance accounted for by predictor variables within each group.

After subjects were classified into left, symmetric or right (with symmetric and right being considered `atypical') dominance patterns, group differences in prevalence rates were studied with χ2 analyses (or Fisher's Exact Test in situations with cell counts <5). These included comparisons of epilepsy versus normal subjects, men versus women in the normal group, men versus women in the epilepsy group and right versus left seizure focus in the epilepsy group (using side of surgery as the grouping criterion). These latter comparisons were done in an attempt to replicate previous studies that have found males at greater risk than females for atypical language lateralization (Strauss et al., 1992a, b) and a higher prevalence of atypical language patterns in epilepsy patients with left than with right temporal seizure focus (Risse et al., 1997).

Similarly, because early brain injury has been associated with a greater incidence of atypical language dominance, we wished to investigate differences in prevalence rates for dominance patterns related to this variable. However, the criteria used by various investigators to determine early brain injury have varied considerably, likely reflecting the availability of evidence and the investigators' wishes to be conservative or liberal in their inclusion. Some investigators relied predominantly on radiological evidence, e.g. CT evidence of brain pathology that presumably occurred early in life (Mateer and Dodrill, 1983), while others used a combination of clinical and medical historical information, e.g. a neurological event early in life associated with an objective sign such as hemiparesis (Rasmussen and Milner, 1977). In the context of limited access to early medical records, onset of seizures is a commonly used marker for age of brain injury, although it is recognized that a seizure disorder may develop well after the precipitating neurological event (Salazar et al., 1985). Many investigators who used seizure onset to define the age of brain injury have not clearly specified whether the defining event was the first seizure of any type or rather the onset of recurrent, medically intractable seizures. Moreover, many investigators are unclear whether febrile seizures were included as a legitimate first sign of a seizure disorder or brain injury. Lastly, investigators have varied with regard to the age cut-off used to define `early,' with cut-offs ranging between 1 and 6 years of age. In the present study, two classification criteria were used to designate onset of brain injury, one representing a relatively liberal (or overly inclusive) criterion and the other a relatively conservative (or stringent inclusion) criterion. The aim of the first criterion was to identify the earliest age of brain injury by identifying the earliest potential sign of a neurological problem. This age, referred to as the age of probable brain injury, was identified by the age of the earliest known neurological event thought to precipitate epilepsy (e.g. birth trauma, head trauma, bout of meningitis or encephalitis) or, if none was identified, then the first seizure of any type (including febrile seizures). The aim of the second criterion was to provide a more stringent definition of early brain injury and emphasized unquestionable evidence of brain dysfunction. For this criterion, the age of onset of medically intractable seizures was used. Onset at age 5 years or less was defined as `early' onset for both probable brain injury and intractable seizures.

Results

Task performance

All subjects learned the tone decision task easily and performed well above chance levels (normal group mean = 97% correct, SD = 3.9; epilepsy group mean = 91.3% correct, SD = 10.4). Performances were also adequate on the semantic decision task in the normal group (mean = 90.7% correct, SD = 6.2) and the epilepsy group (mean = 86.3% correct, SD = 7.9). However, there was a significant difference between group means on both the tone decision task [t(148) = 4.85, P < 0.001] and the semantic decision task [t(148) = 3.68, P < 0.001], indicating that the epilepsy patients had more difficulty on both tasks compared with the normals.

Patterns of language dominance

Activation was observed in all subjects in a group of supratentorial regions previously reported to be activated during this task contrast, including prefrontal cortex of the superior, middle and inferior frontal gyri; mid-anterior cingulate gyrus; posterior cingulate/retrosplenial cortex; anterior superior temporal sulcus; middle temporal and posterior inferior temporal gyri; mid-fusiform and anterior parahippocampal gyri; anterior hippocampus; angular gyrus; anterior thalamus; and head of caudate (Binder et al., 1995, 1996, 1997; Frost et al., 1999). The median LI was 66.1 for the normal group and 60.5 for the epilepsy group, a difference that was not statistically significant (Zu = –1.42, P > 0.05). As illustrated in Fig. 1, the LIs of the normal subjects varied continuously over a range from very strong left hemisphere dominance (+97) to symmetric language distribution (–5). The LIs of the epilepsy group showed greater variability, ranging from exclusive left hemisphere representation (+100) to strong right hemisphere dominance (–81).

Fig. 1

Frequency distributions of language LI in normal and epilepsy subjects.

Using dominance classification criteria, the majority of the subjects in the normal group (94%) had left hemisphere dominance and the remainder had symmetric language representation (6%). Although most subjects showed clear left dominance, no normal subject had exclusive activation of the left hemisphere and the average normal subject had 19% of activated voxels in the right hemisphere. The majority of the epilepsy patients (78%) also had left hemisphere dominance. However, a significantly larger proportion (22%) of the epilepsy group had atypical language lateralization, including 16% with symmetric and 6% with right hemisphere dominance (χ2 = 8.50, P < 0.01; see Table 4).

View this table:
Table 4

Incidence (percentage) rates of language dominance in right-handed subjects from IAP studies and the present fMRI study

Language lateralization*
LeftBilateralRight
*To facilitate comparisons across studies, some values were recombined from original articles with emphasis given to presenting data that separates subgroups of early and late brain injury, and classifications using figures based on relative dominance when possible; values include subjects with early seizure onset or brain injury; see text for definition.
Present fMRI study
Normals (n = 100)94.0 6.0 0.0
Epilepsy
All78.016.0 6.0
Early
Brain injury ≤5 (n = 25)64.028.0 8.0
Intractable seizures ≤5 (n = 12)50.041.7 8.3
Late
Brain injury >5 (n = 25)92.0 4.0 4.0
Intractable seizures >5 (n = 38)86.8 7.9 5.3
IAP studies
Branch et al., 1964 (n = 48)90.0 0.010.0
Kurthen et al., 1994 (n = 142)90.0 6.0 4.0
Loring et al., 1990 (forced relative dominance; n = 91)91.2 4.4 4.4
No early injury (<6 years; n = 66)92.4 4.5 3.0
Mateer and Dodrill, 1983 (n = 66)92.4 6.1 1.5
Rasmussen and Milner, 1977
Early damage (n = 42)81.0 7.012.0
No early damage (n = 140)96.0 0.0 4.0
Rausch and Walsh, 1984 (n = 56)89.3 3.6 7.1
Rey et al., 1988 (n = 29)93.1 6.9 0.0
Risse et al., 1997 (n = 304)87.0 9.0 4.0
Strauss and Wada, 1983 (n = 66)86.4 6.0 7.6
Woods et al., 1988 (n = 118)92.4 4.2 3.4
Zatorre, 1989 (n = 38)63.032.0 5.0

Factors associated with language lateralization

Normal group

There was no significant difference between the LIs of men (median = 65.4) and women (median = 67.4) in the normal group (Zu = –0.56, P > 0.05). χ2 analysis of sex differences using categorized LIs also did not reach significance (χ2 = 0.55, P > 0.05). LIs were not different between those with (median = 63.0) and those without (median = 66.1) familial sinistrality (Zu = –1.02, P > 0.05).

Simple correlation analysis found that education, EHI laterality quotient and task performance data were not correlated with LI. Only age correlated with LI (r = –0.23, P < 0.05) such that older subjects showed less strong language lateralization. Because visual inspection of the scatterplot relating age to LI revealed two older age outliers who may have been contributing disproportionately to this correlation, a non-parametric rank order correlation was computed. This post hoc test also showed a significant but relatively small negative correlation between age and LI (rs = –0.20, P < 0.05). The age versus LI scatterplot is shown in Fig. 2.

Fig. 2

Scatterplot of language LI as a function of age in the normal group, including the regression line, LI = 85.05 – 0.95 (age).

Epilepsy group

LIs of men (median = 67.0) and women (median = 57.5) in the epilepsy group were not significantly different (Zu = –0.49, P > 0.05). χ2 analysis using categorized LIs also did not show differences between the sexes (χ2 = 0.07, P > 0.05). Similarly, LIs did not differ between those with and without familial sinistrality (Zu = –1.57, P > 0.05).

LI was correlated with age of probable brain injury (r = 0.39, P < 0.01), age of onset of intractable seizures (r = 0.50, P < 0.001) and the EHI quotient (r = 0.50, P < 0.001), but did not correlate with age, education, full scale IQ or task performance. In the stepwise regression analysis, age, EHI quotient, full scale IQ, age of intractable seizures and task performance were used as predictor variables for LI. The first variable entered into the equation was age of intractable seizures (r2 = 0.25, P < 0.001), followed by EHI quotient, accounting for an additional 15% of variance in LI (r2 = 0.40, P < 0.001). Partial correlations of the other variables were not significant. When age of probable brain injury was used in the predictor variable list in place of age of intractable seizures, the first variable entered into the equation was EHI (r2 = 0.25, P < 0.001), and was followed by age of probable brain injury, accounting for an additional 14% of variance in LI (r2 = 0.39, P < 0.001).

Epilepsy patients were also categorized according to age of brain injury in order to compare language lateralization patterns of patients with and without an early age of brain injury or onset of intractable seizures. Twenty-five patients had an age of probable brain injury ≤5 years, and 25 had probable brain injury after 5 years of age. There was a significant difference between the LIs of the early (median = 53.0) and late (median = 69.0) onset groups (Zu = –2.60, P < 0.01). Likewise, a χ2 test using categorized LIs demonstrated that a significantly greater proportion of the group with early brain injury had atypical (symmetric or right dominant) language dominance (χ2 = 5.71, P < 0.02; see Table 4). Similar group differences were found when patients with early (n = 12, median = 25.0) and late (n = 38, median = 68.0) onset of intractable seizures were compared using either the continuous variable LI (Zu = –3.06, P < 0.01) or categorized dominance patterns (χ2 = 7.21, P < 0.01). Figure 3 illustrates how differences in the frequency of atypical language dominance relate to three intervals of age of onset of intractable seizures: 0–5 years, 6–15 years and 16 years or greater.

Fig. 3

Frequency of atypical language dominance in normal subjects and in epilepsy patients with onset of intractable seizures (IS) after age 15 years, between age 6 and 15 or before age 6.

LIs in subgroups of epilepsy patients with left (median = 53.0) and right (median = 64.2) hemisphere seizure focus were not significantly different (Zu = –0.09, P > 0.05). Likewise, the χ2 test comparing left and right hemisphere groups using categorized LIs was not significant (χ2 = 0.35, P > 0.05). Among patients with both early brain injury and a left seizure focus, 6 out of 12 (50%) had atypical language, while among patients with both early brain injury and a right hemisphere focus, 3 out of 8 (37.5%) had atypical language. This difference did not reach statistical significance (P = 0.67, Fisher's Exact Test).

Discussion

These observations are the first direct measurements of language lateralization in a large sample of neurologically normal, right-handed subjects. Language LIs for these subjects ranged along a continuum from left hemisphere dominance to relatively symmetric representation, and there was no obvious dichotomous or bimodal clustering of scores (Fig. 1). Atypical patterns were uncommon, with only 6% of normal subjects showing relatively symmetric activation and none showing clear right hemisphere dominance. Although right hemisphere dominance was not observed in the normals, all normal subjects had some activation in the right hemisphere by the linguistic task used here. Atypical dominance patterns were more frequent among epilepsy patients, comprising 22% of the total in this group. A weak relationship between age and LI was observed in the normal group, with older subjects showing more symmetric LIs. None of the other variables examined, including sex, degree of right-handedness, familial sinistrality, education and level of task performance, were related to language lateralization in the normal sample. Early age of brain injury and weaker right hand dominance were associated with an atypical LI in the epilepsy group.

Lateralization and task used in fMRI

It is likely that LI, as computed in this study, would change significantly if either the target or contrast task were changed. Identification of a cognitive system using functional neuroimaging typically requires a comparison between activation produced by a target task and one designed to control for non-specific neural processes. The degree to which the target and control tasks accomplish their intended purpose is critical to understanding the activation patterns. The linguistic and contrast tasks in the current study were selected to activate the neural networks involved specifically in speech perception and lexical-semantic processing. The activation patterns produced by these tasks have been studied in relatively large samples of neurologically normal humans and, in general, identify regions of the left hemisphere that have been linked to language processing through studies of aphasic patients with brain lesions (Binder et al., 1997). In addition, LIs computed from these tasks were strongly correlated with a similar index generated from IAP testing (Binder et al., 1996). Considering that the incidence of atypical language dominance in our epilepsy patients falls closely in line with previous IAP studies (see Table 4), these findings support the current fMRI results for tasks which on average yield language lateralization results comparable with IAP. Still, the IAP is multifaceted, and it is likely that the fMRI LI does not correlate equally well with all aspects of the IAP examination. Rare patients have been reported with interhemispheric dissociation of component linguistic processes, e.g. left hemisphere comprehension and right hemisphere vocal speech (Kurthen et al., 1994; Risse et al., 1997). If such cases exist, it is our suspicion that the fMRI LI used here will correlate more closely with the language comprehension aspects of the IAP than with the vocal speech components.

Right hemisphere language representation in normal and epilepsy subjects

The results suggest that right hemisphere language dominance is less common in normal subjects than can be inferred from IAP studies in right-handed subjects without early brain injury. Estimates from such studies of the incidence of right hemisphere language dominance fall in the 3–5% range (see Table 4), with the findings of Rasmussen and Milner of 4% being the most widely quoted (Rasmussen and Milner, 1977). Similarly, using fMRI in our own epilepsy sample and applying the most stringent exclusion criteria for late brain injury (brain injury ≥5 years), an estimate of 4% is obtained. Still, fMRI in 100 right-handed, neurologically normal individuals failed to identify a single person with clear right hemisphere dominance. In this regard our results are more consistent with adult-onset lesion studies, which suggest that right hemisphere language dominance is present in <2% of normal dextrals (Zangwill, 1967; Gloning et al., 1969; Mastronardi et al., 1994). The most obvious explanation for the discrepancy between normal and epilepsy group estimates is that efforts to exclude all the factors that influence language lateralization in the epileptic patients (e.g. eliminating patients with language reorganization after early brain injury) are incompletely successful and, thus, an estimate of crossed dominance from this population is inherently biased.

One practical interpretation of these findings is that language mapping may not be necessary in right-handed patients prior to right hemisphere surgery when the illness prompting surgery is acquired after childhood. However, 6% of normal subjects and up to 13% of `late onset' epilepsy patients had 40% or more of their language activation in the right hemisphere (i.e. had LIs ≤20), which suggests a significant potential risk for language deficits following right hemisphere surgery. One critical issue is whether bihemispheric representation of language implies that either side can mediate language functions independently, or rather that both hemispheres are necessary (Risse et al., 1997). The previously mentioned correlation between fMRI and IAP laterality indices (Binder et al., 1996) indicates a direct linear relationship between the relative extent of right hemisphere language activation measured by fMRI and the severity of language deficits induced by right hemisphere anaesthetization. This relationship suggests that right hemisphere language-related activation observed by fMRI may have predictive value in determining the risk of postoperative language deficits from right hemisphere surgery. Given these findings, the safety of fMRI mapping recommends its general use prior to brain surgery in sensitive cortical regions, particularly in individuals with brain disease acquired early in life. Ultimately, however, the cost of the procedure will need to be weighed against the likelihood of finding significant right hemisphere language activation in a given individual. While this study provides a reasonable estimate of this likelihood in normal, young, right-handed adults, future studies will need to be conducted on samples of older, left-handed and ambidextrous normal subjects, as well as on larger samples of patients with brain disease acquired at different ages.

The presence of right hemisphere language-related activation in nearly all subjects, the graded variation in language lateralization demonstrated in Fig. 1 and the linear relationship between LI and IAP language lateralization all argue for the view that language lateralization is a continuously variable phenomenon. Use of categorical designations such as `left dominant', `right dominant' and `bilateral' has been ubiquitous and no doubt of value in some clinical settings, but the utility of this approach for understanding the neurobiology of language organization is far less certain. Several authors have proposed, for example, that much of the variability in frequency of atypical dominance observed in different IAP studies (Table 4) is due simply to differences in categorization methods (Loring et al., 1990; Snyder et al., 1990; Kurthen et al., 1994; Risse et al., 1997). For example, while most IAP studies show an incidence rate of atypical dominance (combined right and bilateral language) in the 7–13% range, the data obtained by Zatorre (Zatorre, 1989) imply an incidence rate of 37%. This appears to be due to a categorization rule that identified a patient as having bilateral language representation when language was disrupted in any way on both hemispheric injections, without regard for relative severity of the disruption. Such a categorization method has the effect of increasing the estimate of bilateral language representation and decreasing the estimate of left dominance, an effect identified in other studies as well (Loring et al., 1990; Risse et al., 1997). Similarly, in the current fMRI study, incidence rates for atypical dominance are a function of the cut-off points used for classifying LI. If language lateralization is truly a continuous variable with a unimodal distribution, such categorization is inevitably arbitrary and leads to loss of information. This information could also be critically valuable in clinical applications that depend on highly precise estimates of risk. Using the LI from an fMRI test, for example, it may eventually be possible to predict not just the risk associated with being in the `left dominant' category, but the much more precise risk associated with each particular LI.

Factors related to atypical language lateralization in normal subjects

Variability of language lateralization in normal subjects was not strongly related to sex, degree of right-handedness, presence of left-handed family members, education or task performance. The range of some of these variables was relatively restricted, which may have lessened the ability to detect significant relationships. For example, all subjects in this study were right-handed and had EHI laterality quotients in the range +50 to +100. Had left-handed and ambidextrous subjects been included, a significant relationship between LI and EHI scores might have emerged. Because the normal subject sample was drawn mainly from local universities and the medical college campus, levels of education and task performance also tended to be relatively homogeneous. Thus, while no significant effects of these variables were found over the ranges studied, future investigations using less homogeneous samples would be useful to conclusively rule out the possibility that these variables have a small but measurable relationship to language lateralization.

The lack of sex effects on language laterality observed here is consistent with recent PET studies of language processing (Buckner et al., 1995; Price et al., 1996a). In contrast, Shaywitz and colleagues found, using fMRI, that men had greater left lateralization of activation in the inferior frontal gyrus during orthographic, phonological and semantic language tasks (Shaywitz et al., 1995; Pugh et al., 1996). These discrepancies could conceivably be due to any of several differences in methodology among the different studies, which included differences in task demands, image acquisition techniques, data analysis methods and statistical thresholds. Because the LI used in the present study is based on activation over the hemisphere as a whole, the current analysis may not detect gender differences affecting more focal regions such as the inferior frontal gyrus. For this reason, we recently performed several additional voxel-wise and region-of-interest analyses of our normal subject data, including an analysis using regions of interest identical to those used by Shaywitz and colleagues (Shaywitz et al., 1995). These additional analyses showed no statistically significant differences between women and men in language-related activation at a voxel level, and no sex differences in lateralization of language-related activation for any region of interest (Frost et al., 1999). Future studies will be needed to explain the differing results, but the preponderance of current evidence suggests that sex differences in large-scale language organization, if present, are probably small in comparison to the degree of similarity of this brain system in men and women.

An unexpected observation was that LI in the normal group was negatively correlated with age, indicating that older subjects tended to have more symmetric activation. Although relatively small, this effect also proved significant in a subsequent rank-order correlation test. While unexpected, this finding is not surprising in that other investigators have shown age-related changes in the patterns of functional activity in visual perception (Grady et al., 1994) and memory (Grady et al., 1995). These changes in activation patterns were interpreted as adaptations that compensate for age-related changes in primary memory and perceptual systems. Because the age range of our normal subjects was relatively limited, the magnitude of age-related effects may be underestimated. Future studies of normal subjects will need to include a wider age range to obtain a more reliable estimate of this effect.

Factors related to atypical language lateralization in epilepsy

Language lateralization was not strongly related to sex, education or task performance in the epilepsy patients, corroborating similar negative findings in the normal group. Task performance was somewhat more variable in the epilepsy group, suggesting that any relationship that may exist between this variable and LI is weak and would only be observable over a wide range of variation or in a very large subject sample. Age and full scale IQ also showed no significant relationship to LI over the ranges studied. The lack of a detectable age effect in the epilepsy group could be explained by the fact that there were fewer subjects, allowing other more potent variables, namely handedness and brain injury, to mask this weak relationship. In particular, there was a weak positive correlation between age and age of onset of intractable seizures (r = 0.27, P = 0.06), indicating that patients with onset of intractable seizures at a younger age tended to present to our programme at a younger age. This selection bias would have produced an effect on LI that was opposite in direction to the age effect observed in normal subjects. Because the effect on LI of age of intractable seizure onset was much stronger than the age effect observed in normal subjects, this latter relationship may have been unobservable in the epilepsy group.

In contrast to the normal group, the epilepsy group showed a relatively strong correlation (r = 0.54; P < 0.001) between language lateralization and degree of right-handedness on the EHI, despite the fact that the normal and epilepsy groups had very similar mean EHI scores and EHI score variance. This finding concurs with previous studies demonstrating a close relationship between left-handedness and atypical language dominance in epilepsy patients (Rasmussen and Milner, 1977; Mateer and Dodrill, 1983; Rausch and Walsh, 1984; Rey et al., 1988; Woods et al., 1988; Loring et al., 1990; Strauss et al., 1990; Satz et al., 1994; Helmstaedter et al., 1997; Risse et al., 1997) and suggests that pathological processes in epilepsy exert a common set of influences on the development of both manual and language lateralization. In contrast, normal subjects do not show this relationship, presumably because handedness variation within this range in the normal population is due to non-pathological (e.g. genetic) causes that do not affect language lateralization.

The other significant predictors of LI in the epilepsy group were age of onset of intractable seizures (r = 0.50) and age of suspected brain injury (r = 0.39). Together with degree of right-handedness, the intractable seizures variable accounted for 40% and the brain injury variable 39% of the variance in LI. Consistent with several previous studies (Rasmussen and Milner, 1977; Mateer and Dodrill, 1983; Strauss and Wada, 1983), we found that patients with evidence of earlier brain injury tended to have more right hemisphere language representation. The relationship between LI and age of onset of intractable seizures was slightly stronger, probably reflecting the fact that recurrent seizures represent a more definitive marker of permanent brain dysfunction than might a febrile seizure or an early neurological event that could have only transient sequelae. Visual inspection of the scatterplots relating these variables to LI did not show any obvious non-linearity in the relationship, nor was the effect any stronger when age of brain injury and intractable seizures were treated as categorical variables (P < 0.05) than as a continuous variable in a linear regression analysis (P < 0.001). These results imply that there may not be an absolute `critical period' during which brain injury can influence language organization and after which no such influence is possible. Instead, the roughly linear relationship between these variables suggests that later brain injury can still affect language lateralization, although less profoundly.

Given prior findings that reorganization of language is most marked after early left hemisphere injury (Strauss et al., 1990; Satz et al., 1994), it was somewhat surprising that there was no significant relationship between the side of seizure focus and LI, especially since a conservative criterion was used to define the seizure focus (i.e. side of surgical resection). There are several possible explanations for this finding. First, the sample size may have been too small to detect differences due to seizure side. The difference in median LI between left and right groups, although not significant, did show a slight trend in the expected direction, i.e. the left seizure focus subgroup had slightly weaker left language dominance than the right seizure subgroup. A second possibility is that the lack of an effect of seizure focus lateralization may be related to exclusion from the study of left-handed and ambidextrous subjects. Patients with `pathological' left-handedness are the most likely to demonstrate atypical language organization (Woods et al., 1988). Exclusion of this subgroup may have removed from our sample most of the observable lateralizing effects due to left hemisphere injury that occur in the epilepsy population. This proposal receives some support from the observed relationship between EHI handedness scores and language LI in the epilepsy group. fMRI studies of left-handed normal and epilepsy subjects should help to test this hypothesis. Finally, this negative finding may reflect a sampling bias because patients with early left brain injury and subsequent shift of language functions to the right hemisphere `commonly show a general cognitive retardation' (Strauss et al., 1992b; see also Woods and Teuber, 1973; Milner, 1974; Strauss et al., 1990). These patients may have been excluded from our study either because of a full scale IQ ≤70 or inadequate task performance levels. Future studies should include epilepsy patients with a broader range of intellectual abilities to elucidate more fully the effects of early left brain injury on language lateralization.

Epilepsy patients differ from normal subjects in a variety of important ways and a direct comparison of LIs for these groups is therefore problematic. Demographic differences, medications and underlying brain disease might all conceivably affect the LI. The difference we observed in frequency of atypical language lateralization between the two groups is not likely to be related to sex, education or task performance differences, since these variables were not strongly correlated with LIs in either group. Age, familial sinistrality and full scale IQ also do not appear to be strongly related to the LI and therefore seem unlikely to account for large group differences. The significant relationships between age of seizure onset, degree of right-handedness and LI in the epilepsy group, however, are consistent with the proposal that the higher incidence of atypical language lateralization in this group is attributable to a subgroup of patients with reorganized language functions secondary to early brain injury. The distribution of language lateralization patterns observed in epilepsy is thus due to superimposition of this subgroup with `pathological reorganization' on the normally occurring language dominance distribution.

Acknowledgments

We wish to thank team members of the Comprehensive Epilepsy Program and Dr Robert Cox of the Biophysics Institute at the Medical College of Wisconsin for technical assistance. This research was supported by a grant from the Charles A. Dana Foundation, National Institute of Neurological Diseases and Stroke Grants RO1 NS33576 and RO1 NS35929, and National Institute of Mental Health Grant PO1 MH51358. Preliminary results from this study were presented at the 25th Annual Meeting of the International Neuropsychological Society, Orlando, Florida in 1997.

References

View Abstract