Nicotine increases brain functional network efficiency

Abstract

Despite the use of cholinergic therapies in Alzheimer's disease and the development of cholinergic strategies for schizophrenia, relatively little is known about how the system modulates the connectivity and structure of large-scale brain networks. To better understand how nicotinic cholinergic systems alter these networks, this study examined the effects of nicotine on measures of whole-brain network communication efficiency. Resting state fMRI was acquired from fifteen healthy subjects before and after the application of nicotine or placebo transdermal patches in a single blind, crossover design. Data, which were previously examined for default network activity, were analyzed with network topology techniques to measure changes in the communication efficiency of whole-brain networks. Nicotine significantly increased local efficiency, a parameter that estimates the network's tolerance to local errors in communication. Nicotine also significantly enhanced the regional efficiency of limbic and paralimbic areas of the brain, areas which are especially altered in diseases such as Alzheimer's disease and schizophrenia. These changes in network topology may be one mechanism by which cholinergic therapies improve brain function.

Figure 1

Network topology examples, a) Random networks, with randomly determined connections, have high global efficiency, as shown by the many connections crossing the center of the graph, and low local efficiency, as shown by the relatively few connections between nearby nodes. Lattice networks show the reverse pattern. Small-world networks are intermediate, with properties of both, b) Regional efficiency. Both networks have an identical number of nodes and connections, with only small differences in their topology. Left: The regional efficiency of node ‘n’ is low due to the many connections separating it from the nodes marked with asterisks. Communication between n and these nodes will be inefficient. Right: The regional efficiency of n is improved by redistributing its connections. All nodes are now relatively close to node n, allowing for efficient communication between it and all other areas of the network.

Figure 2

Mean SPN for the post-placebo condition as an example of a whole-brain network. Glass-brain layout, circles represent nodes and their size proportional to nodal degree. No significant differences between experimental conditions were observed.

Figure 3

Small-world properties of brain networks. For all conditions and costs, local and global efficiencies were intermediate between simulated random and lattice networks.

Figure 4

Nicotine-associated changes in global and local efficiencies. Nicotine, compared to placebo, significantly increased local efficiency. A trend towards increased global efficiency also was observed.

Figure 5

Changes in regional efficiencies for each AAL region, averaged across hemispheres. Limbic and paralimbic areas showed increased regional efficiencies in response to nicotine as compared to placebo.