Changes in glucose metabolism due to aging and gender-related differences in the healthy human brain

Abstract

Using [¹⁸F]fluoro-deoxy-glucose-PET, we studied relative metabolic changes due to age- and gender-related differences in the brain of 126 healthy subjects from their twenties to seventies. We used a data-extraction technique, the three-dimensional stereotactic surface projections (3D-SSP) method, to measure metabolic changes with fewer effects of regional anatomic variances. Simple regression analysis revealed significant age-related increases in relative metabolic values in the parahippocampal and amygdala regions in both sexes in their twenties to forties, and significant age-related decreases in both sexes in their fifties to seventies. Relative values in the frontal lobe showed significant age-related decreases in both sexes in their twenties to forties, but these effects were not seen in subjects in their fifties to seventies. Significant gender differences in correlation coefficients of relative values with age were shown in the parahippocampal, primary sensorimotor, temporal, thalamus and vermis regions in subjects in their 20s to 40s, but disappeared in subjects in their twenties to forties, but were not apparent in subjects in their fifties to seventies except in the vermis. Males in their twenties to sixties and females in their fifties showed significant laterality in relative values in the temporal lobes. Our study demonstrated age- and gender-related differences in glucose metabolism in healthy subjects.

Introduction

Understanding age-related and gender-based differences in the function of the brain in healthy subjects is indispensable for investigating psychiatric diseases, which sometimes show significant relationships to age and gender (Buchsbaum and Hazlett, 1997, Flor-Henry, 1978). Studies with [¹⁸F]fluoro-deoxy-glucose positron emission tomography (FDG-PET) showed that normal aging is associated with specific patterns of regional cerebral glucose metabolism (Kuhl et al., 1982, Salmon et al., 1991, Loessner et al., 1995, Moeller et al., 1996, Petit-Taboué et al., 1998, Hazlett et al., 1998, Hazlett et al., 2000, Ivançević et al., 2000, Willis et al., 2002). However, it was reported that statistically significant age reductions in regional brain glucose metabolism were not detectable in healthy subjects, after correction for brain atrophy (Ibáñez et al., 2004). A study of normal aging in Japanese subjects also found that the reduction in FDG uptake with advancing age, detected without a correction method for parietal volume effects, could be accounted for largely by age-related cerebral volume loss in perisylvian and medial frontal areas (Yanase et al., 2005). These discrepancies suggested the importance of precise regional measurement and correction for brain atrophy using magnetic resonance imaging (MRI) (de Leon et al., 1987).

In addition, it was reported that functional and morphological changes were affected by gender differences (Coffey et al., 1998). There have been a number of studies of gender differences in brain metabolism, but they were also not consistent (Volkow et al., 1997). FDG-PET studies showed higher global cerebral glucose metabolic rates in females than in males (Baxter et al., 1987, Yoshii et al., 1988, Andreason et al., 1994). Female subjects had a lower glucose metabolic rate than male subjects in the caudate, but an equal rate in the putamen (Brickman et al., 2003). Significant and reproducible gender differences were shown in the cerebellar regions (Volkow et al., 1997). On the other hand, no significant differences were found between young males and females for cerebral glucose metabolism (Miura et al., 1990). Using volumetric MRI and the regional rate of glucose metabolism assessed by PET, a study of gender-related differences and effects of aging showed that metabolic findings cannot be fully explained by sex-related differences in brain atrophy, although this probably accounted for some results (Murphy et al., 1996). Moreover, since glucose metabolism is sensitive to cognitive activity, alterations in age-related cognitive activity may affect the measurement of regional metabolism (Buchsbaum and Hazlett, 1997). Differences in language and culture may also affect age- and gender-related metabolism.

These previous investigations of age- and gender-based differences in brain glucose metabolism using FDG-PET had methodological limitations. In this study we employed a data-extraction technique, the three-dimensional stereotactic surface projections (3D-SSP) method (software library NEUROSTAT; Department of Internal Medicine, University of Michigan, Ann Arbor, MI, USA) (Minoshima et al., 1995, Burdette et al., 1996, Minoshima et al., 1998, Minoshima, 2003). This method enables improved region identification, quantitative and visual interpretation, and intersubject comparison for groups of subjects in PET examination. It was reported that quantitative accuracy in the 3D-SSP method was less biased by the size of regions or by cortical atrophy than conventional volume of interest (VOI) analysis and Statistical Parametric Mapping (SPM) (Ishii et al., 2001).

The 3D-SSP images and relative glucose metabolic values were acquired in 126 healthy male and female subjects from their twenties to seventies. Previous studies have suggested that some age-related changes in brain structure and metabolism were not linear over the adult age range, but showed accelerated changes in the elderly (Jernigan et al., 1991, Brickman et al., 2003). Therefore, we divided subjects in their twenties to seventies into two groups (subjects in their twenties to forties, and subjects in the fifties to seventies) and examined these two groups. The purpose of this study was to use FDG-PET to investigate age- and gender-related differences in metabolic activity.