Dose-Dependent Effects of Intranasal Insulin on Resting-State Brain Activity

Dose-Dependent Effects of Intranasal Insulin on Resting-State Brain Activity Abstract Context Insulin action in the human brain influences eating behavior, cognition, and whole-body metabolism. Studies investigating brain insulin rely on intranasal application. Objective To investigate effects of three doses of insulin and placebo as nasal sprays on the central and autonomous nervous system and analyze absorption of insulin into the bloodstream. Design, Participants, and Methods Nine healthy men received placebo or 40 U, 80 U, and 160 U insulin spray in randomized order. Before and after spray, brain activity was assessed by functional magnetic resonance imaging, and heart rate variability (HRV) was assessed from electrocardiogram. Plasma insulin, C-peptide, and glucose were measured regularly. Setting General community. Results Nasal insulin administration dose-dependently modulated regional brain activity and the normalized high-frequency component of the HRV. Post hoc analyses revealed that only 160 U insulin showed a considerable difference from placebo. Dose-dependent spillover of nasal insulin into the bloodstream was detected. The brain response was not correlated with this temporary rise in circulating insulin. Conclusions Nasal insulin dose-dependently modulated regional brain activity with the strongest effects after 160 U. However, this dose was accompanied by a transient increase in circulating insulin concentrations due to a spillover into circulation. Our current results may serve as a basis for future studies with nasal insulin to untangle brain insulin effects in health and disease. Research over the last several years identified the human brain as an insulin-sensitive organ (1, 2). In response to the hormone, the central nervous system regulates various functions as the response to food cues, reward processes, and memory (1). Furthermore, insulin effects in the brain influence the rest of the body, modulating peripheral insulin sensitivity, thermogenesis, as well as liver and lipid metabolism (2, 3). However, a substantial number of individuals are brain insulin resistant and therefore lack these effects of the peptide. The best-characterized factor associated with brain insulin resistance is body weight with reduced insulin effects in overweight and obese persons. Beyond this, a number of other factors have been identified thus far (1, 2). However, it is still not known whether factors associated with brain insulin resistance represent cause or consequence thereof. To assess brain-specific insulin effects in humans, most research applied insulin as a nasal spray (2). Research in animals demonstrated that peptide transporters near the olfactory bulb rapidly transport a limited number of peptides (including insulin) into the cerebrospinal fluid (CSF), from which they further reach brain cells (4). In humans, a rise in CSF insulin concentrations was detected as early as 10 minutes after administration of 40 U of the peptide as nasal spray (5). Functional consequences (e.g., effects on brain activity) occur in a comparable time frame (6). The earliest experiments mostly used a dose of 160 U to investigate the consequences of intranasal insulin application on eating behavior and body weight; studies investigating memory function in cognitively impaired patients, in contrast, used a lower dose of insulin (7, 8). An acute dose of 160 U has been shown to stimulate regional brain activity in a number of brain areas (9). Although initial experiments with a lower insulin dose suggested that none of the peptide reached the circulation after administration as a nasal spray (5), it recently became evident that small amounts of human insulin are indeed absorbed into the bloodstream and are detectable in venous blood (10–12). After application of 160 U of insulin as a nasal spray, ∼0.1 U enter the bloodstream (12, 13). Despite having no major effects on blood glucose (e.g., not inducing hypoglycemia), the small rise in circulating insulin may induce effects in peripheral tissues, possibly influencing the interpretation of metabolic effects of nasal insulin. Furthermore, experiments with chronic administration of human insulin as nasal spray suggested that high doses might not necessarily have the strongest effects. Indeed, in a 4-month trial in cognitive-impaired persons, the lower dose of 20 U of the peptide had more profound effects on memory than 40 U (14). Even though the intranasal administration has already been used in a number of studies in humans (6, 7, 15–21), there is still no systematic comparison of the acute effect of different nasal insulin doses on central and autonomous nervous system activity. We therefore investigated the effects of placebo and three different doses of human insulin as nasal spray. Before and after spray, brain activity was assessed by functional magnetic resonance imaging (fMRI) while simultaneously recording electrocardiograms (ECGs) to assess heart rate variability (HRV). Frequent blood sampling was used to analyze the absorption of insulin into the bloodstream. Material and Methods Participants Nine healthy men participated in the study [body mass index (BMI) 20 to 26 kg/m2; age 23 to 30 years] (Table 1). Informed written consent was obtained from all subjects, and the local ethics committee approved the protocol. To ensure that participants were healthy and did not suffer from psychiatric, neurologic, or metabolic diseases, they underwent a thorough medical examination. Table 1. Characteristics of Nine Male Participants Variable  Mean ± SD (Range)  Age (y)  26.56 ± 2.78 (23–30)  BMI (kg/m2)  23.44 ± 2.01 (20.09–26.01)  Body fat content (%)  18.9 ± 2.4 (16.6–21.9)  Waist-to-hip ratio  0.91 ± 0.07 (0.81–0.99)  Hemoglobin A1C (%)  5.2 ± 0.1 (5.1–5.4)  HOMA-IR (mean over all 4 measurement days)  2.25 ± 0.41 (1.51–2.91)  Variable  Mean ± SD (Range)  Age (y)  26.56 ± 2.78 (23–30)  BMI (kg/m2)  23.44 ± 2.01 (20.09–26.01)  Body fat content (%)  18.9 ± 2.4 (16.6–21.9)  Waist-to-hip ratio  0.91 ± 0.07 (0.81–0.99)  Hemoglobin A1C (%)  5.2 ± 0.1 (5.1–5.4)  HOMA-IR (mean over all 4 measurement days)  2.25 ± 0.41 (1.51–2.91)  Abbreviation: HOMA-IR, homeostatic model assessment of insulin resistance. View Large Study design Volunteers received placebo or 40 U, 80 U, and 160 U insulin as nasal spray in randomized order (on 4 separate days with 7 to 14 days’ time lag) with repetitive MRI and ECG measurements (Fig. 1). Participants were blinded to the condition. Experiments were conducted after an overnight fast and started at 7:00 am with a baseline MRI measurement including two blood oxygenation level–dependent (BOLD) resting-state functional MRI (rsfMRI1) and a cerebral blood flow (CBF) measurement (CBF1). After the basal measurement, the respective nasal spray was administered. After 15 and 30 minutes, second and third MRI measurements were performed (rsfMRI2/CBF2 and rsfMRI3/CBF3). During all three MRI measurements, ECG was recorded. The subjective feeling of hunger was rated at two time points (before spray application 60 minutes after intranasal spray) on a visual analog scale from 0 to 10 (0, not hungry at all; 10, very hungry). Figure 1. View largeDownload slide Schematic overview of study design. CBF and rsfMRI for subcortical regions and whole-brain were acquired at different time points before and after intranasal placebo and insulin sprays. ECG was recorded during the entire MRI acquisition. Blood samples, indicated by red symbols, were taken before and 5, 10, 15, 30, 60, 90, and 120 minutes after intranasal placebo and insulin sprays. Figure 1. View largeDownload slide Schematic overview of study design. CBF and rsfMRI for subcortical regions and whole-brain were acquired at different time points before and after intranasal placebo and insulin sprays. ECG was recorded during the entire MRI acquisition. Blood samples, indicated by red symbols, were taken before and 5, 10, 15, 30, 60, 90, and 120 minutes after intranasal placebo and insulin sprays. Participants rated 100 food pictures in two separate blocks according to explicit “liking” (how much do you like the food item in general) and “wanting” (how much would you like to eat the food item right now) on each study day. Application intranasal insulin/placebo On each day, participants received 1.6 mL of nasal spray. It contained either placebo or 40, 80, or 160 U of human insulin (Insulin Actrapid; Novo Nordisk, Bagsvaerd, Denmark). Under supervision, the spray was administered over 4 minutes with two puffs per nostril every minute. Blood measurements Venous blood samples were obtained immediately before as well as 5, 10, 15, 20, 30, 60, 90, and 120 minutes after spray application. Plasma glucose concentrations were measured using the glucose-oxidase method (Yellow Springs Instruments, Yellow Springs, OH). Serum insulin, C-peptide, cortisol, luteinizing hormone (LH), follicle-stimulating hormone, testosterone, and thyroid-stimulating hormone were measured by chemiluminescence assays on the ADVIA Centaur XPT and adrenocorticotropic hormone by an automated solid-phase, chemiluminescent immunoassay on the Immulite XPT analyzer (both from Siemens Healthineers, Eschborn, Germany). Whole-brain fMRI measurement Data acquisition Scanning was conducted at a 3T whole-body Siemens scanner (Magnetom Prisma; Siemens, Erlangen, Germany) with a 20-channel coil. Three different types of functional data sets were recorded each day before and after nasal spray application. In addition, high-resolution T1-weighted anatomical images were obtained. To acquire CBF maps, pseudocontinuous arterial spin labeling was performed using a two-dimensional echo planar imaging (EPI) readout sequence (22) with background suppression. A total of 52 images were acquired with the following parameters: 16 slices; slice thickness 3 mm; 1.5-mm gap; repetition time 4500 ms; echo time 23 ms; field of view (FOV)read 192 mm2; FOVphase 100%; matrix 64 × 64;, flip angle 90°; voxel size 3 × 3 × 3 mm3; bandwidth 2004 px/Hz; and tag gradient strength 7.0 mT/m. The first image volume prior to the preparation scans was used for calibration (M0). Whole-brain BOLD data were collected by using EPI sequences, as recently reported (21). To acquire BOLD activation of subcortical regions at a higher resolution, an EPI sequence with ZOOMit was used. It uses dynamic excitation pulses to achieve selective FOV (zoomed) images, without aliasing artifacts. The following parameters were used: repetition time 3 s; echo time 34 ms; FOVread 192 mm2; FOVphase 33.3%; matrix 96 × 32 × 36; flip angle 90°; voxel size 2 × 2 × 2.5 mm3; and slice thickness 2 mm. Images were acquired in ascending order. Each brain volume comprised 36 axial slices, and each functional run contained 60 image volumes, resulting in a total scan time of 3:06 minutes. Before intranasal insulin application and after 30 minutes, all fMRI measurements were performed as displayed in Fig. 1. Due to time constraints, only CBF and BOLD of subcortical regions were performed after 15 minutes. Pseudocontinuous arterial spin labeling preprocessing Preprocessing was performed using FSL with the following tools (5.0.9). First, the images were realigned using mcflirt. The resulting four-dimensional ASL data were processed using oxford_asl. For CBF quantification, a single-compartment standard kinetic model was used (23). Perfusion images (ΔM) were obtained by pairwise subtraction of tag and control images. A voxel-wise calibration was performed, thereby correcting possible radiofrequency coil inhomogeneity (23). In a next step, the T1-image was coregistered with the M0-image, and the resulting transformation parameters were applied to the CBF maps. Finally, the CBF maps were normalized in MNI space using Statistical Parametric Mapping 12 (SPM12). rsfMRI data preprocessing We used the Data Processing Assistant for Resting-State fMRI (24) to analyze the resting state fMRI data, which is based on SPM12 and the Resting-State fMRI Data Analysis Toolkit (25) (http://www.restfmri.net). The whole-brain functional images were normalized to voxel size, 3 × 3 × 3 mm3, and the subcortical functional images to 2 × 2 × 2 mm3 and then smoothed (full width at half maximum: 6 mm for whole-brain and 4 mm for the subcortical images). Nuisance regression was performed using white matter, CSF, and the six head motion parameters as covariates. To investigate resting-state brain activity, we calculated the fractional amplitude of low-frequency fluctuation (fALFF; 0.01 to 0.08 Hz) of the BOLD signal (26). The regional intensity of spontaneous BOLD fluctuations is quantified by the power spectrum in the low-frequency range (0.009 to 0.08 Hz) and regularized by the power in the whole-frequency range (0 to 0.25 Hz) (26). Statistical analyses CBF. CBF values were extracted of the hypothalamic region of interest based on recent findings. In this study, we identified the hypothalamus to respond with a persistent decrease in activity up to 30 minutes after intranasal insulin application (6, 10, 19, 20). This response was diminished in individuals with unfavorable fat distribution and whole-body insulin resistance. Baseline-corrected CBF maps were computed to quantify the hypothalamic CBF change 15 and 30 minutes after intranasal insulin application (CBF2 − CBF1 and CBF3 − CBF1, respectively). Repeated-measurement analysis of variance (ANOVA; factor insulin dose with four levels: placebo, 40 U, 80 U, and 160 U insulin and factor time with two levels) with homeostatic model assessment of insulin resistance (HOMA-IR) as covariate was performed in SPSS (version 20; IBM, Armonk, NY) (P < 0.05). rsfMRI. fALFF maps of the subcortical regions were baseline corrected (rsfMRI2 − rsfMRI1 and rsfMRI3 − rsfMRI1). Repeated-measurement ANOVA (full factorial model) was performed in SPM12. Relevant clusters were extracted for post hoc analyses in SPSS (version 20; IBM). A statistical threshold of P family-wise error (PFWE) <0.05 voxel-level whole-brain corrected was applied. Additionally, small volume correction was performed for the hypothalamus, an a priori region of interest (6, 10). All whole-brain fALFF maps were baseline corrected (rsfMRI3 − fMRI1). Repeated-measurement ANOVA/full factorial model was performed in SPM12. Relevant clusters were extracted for post hoc analyses in SPSS (version 20; IBM). A statistical threshold of PFWE < 0.05 voxel-level whole-brain corrected was applied. Additionally, small volume correction was performed for the prefrontal cortex (PFC), an a priori region of interest, based on recent findings showing that activity and functional connectivity can be enhanced with intranasal insulin (6, 19, 21). HRV. ECGs were recorded with BIOPAC MP35 (BIOPAC, Goleta, CA) and a sampling rate of 1000 Hz during the whole fMRI recording. For analysis, we extracted 5 minutes of recording during the CBF measurements. The data were analyzed with Kubios (http://kubios.uef.fi). HRV parameters were calculated on individual R-R interval time series in the low-frequency (0.04 to 0.15 Hz) and high-frequency (0.15 to 0.40 Hz) bands. HRV parameters were baseline corrected and analyzed using a repeated-measurement ANOVA in SPSS (version 20; IBM) (P < 0.05). Blood values. The areas under the curve (AUCs) were calculated using the trapezoid rule. Doses were compared by linear regression analyses with dose as a continuous variable. Results Dose-dependent effects of intranasal insulin on brain activity Hypothalamic CBF change Based on our recent finding, we assessed dose-dependent insulin-induced hypothalamic change in CBF as readout for hypothalamic insulin sensitivity (6, 10, 12, 27). We identified a significant main effect of dose [F(3) = 5.36; P = 0.01] (Fig. 2) and a significant interaction between dose and peripheral insulin sensitivity assessed by HOMA-IR [F(3) = 4.43; P = 0.02]. No main effect of time (15 vs 30 minutes postspray) was observed. Post hoc analyses revealed a significant CBF difference between placebo and 160 U of intranasal insulin (P = 0.04, Bonferroni corrected). Furthermore, a significant positive correlation was observed between HOMA-IR and the hypothalamic response to 160 U of insulin 30 minutes after application (r = 0.76; P = 0.017). Figure 2. View largeDownload slide Dose-dependent intranasal insulin effect on hypothalamic CBF. Bar plot shows the extracted CBF values adjusted for HOMA-IR from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between placebo and 160 U of insulin (*P < 0.05, Bonferroni corrected). Figure 2. View largeDownload slide Dose-dependent intranasal insulin effect on hypothalamic CBF. Bar plot shows the extracted CBF values adjusted for HOMA-IR from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between placebo and 160 U of insulin (*P < 0.05, Bonferroni corrected). Subcortical rsfMRI response using fALFF We observed a significant main effect of dose in the left amygdala (PFWE = 0.05) (Fig. 3A and Table 1). A significant linear decrease with insulin dose was observed in the right caudate nucleus (PFWE = 0.04) (Fig. 3B and Table 1) and the hypothalamus (PFWE = 0.015, small volume corrected) (Fig. 3C and Table 1). Because no main effect of time or interaction between time and condition were observed (PFWE > 0.05), post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. Post hoc analyses showed a significant fALFF difference in the hypothalamus between placebo and 160 U insulin sprays (P = 0.003) and between placebo and 80 U (P = 0.014). For the left amygdala, we observed a significant difference between placebo and 160 U (P < 0.001), between 40 U and 160 U (P < 0.001), and between 80 U and 160 U (P = 0.004) and a statistical trend for difference between placebo and 80 U (P = 0.07). For the caudate response, we observed a significant difference between placebo and 160 U (P = 0.003) and between 40 U and 160 U (P = 0.003). No interactions were observed with peripheral insulin sensitivity assessed as HOMA-IR. Figure 3. View largeDownload slide Dose-dependent intranasal insulin effect on rsfMRI response. Bar plots show the extracted mean z-values of sizeable fALFF clusters from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between conditions as indicated by asterisks (*P < 0.05; **P < 0.005, Bonferroni corrected). (A) Main effect of insulin dose in the left amygdala displayed by yellow color-coded F-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (B) Linear decrease of insulin dose in the right caudate displayed by yellow color-coded T-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (C) Linear decrease of insulin dose in the hypothalamus displayed by yellow color-coded T-value map (Psvc <0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (D) Linear increase of insulin dose in the PFC displayed by yellow color-coded T-value map (Psvc < 0.05, FWE corrected). Bar plot shows the extracted mean z-values of PFC fALFF from before to 30 minutes after nasal spray application for the different conditions. Figure 3. View largeDownload slide Dose-dependent intranasal insulin effect on rsfMRI response. Bar plots show the extracted mean z-values of sizeable fALFF clusters from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between conditions as indicated by asterisks (*P < 0.05; **P < 0.005, Bonferroni corrected). (A) Main effect of insulin dose in the left amygdala displayed by yellow color-coded F-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (B) Linear decrease of insulin dose in the right caudate displayed by yellow color-coded T-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (C) Linear decrease of insulin dose in the hypothalamus displayed by yellow color-coded T-value map (Psvc <0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (D) Linear increase of insulin dose in the PFC displayed by yellow color-coded T-value map (Psvc < 0.05, FWE corrected). Bar plot shows the extracted mean z-values of PFC fALFF from before to 30 minutes after nasal spray application for the different conditions. Whole-brain rsfMRI using fALFF We observed a significant linear increase in the lateral PFC (PFWE = 0.02 small volume corrected) (Fig. 3D and Table 1) with increasing insulin doses. Post hoc analyses showed a significant difference between placebo and 160 U (P < 0.001), between placebo and 80 U (P = 0.015), between 40 U and 160 U (P = 0.001), and between 40 U and 80 U (P = 0.045). No interactions were observed with HOMA-IR. Whole-brain and subcortical rsfMRI results are summarized in Table 2. Table 2. rsfMRI Response to Different Concentrations of Insulin Brain Regions  MNI Coordinate (x, y, z)  P Value, FWE Corrected  z-Value  Post Hoc Analyses (P < 0.05, Bonferroni Corrected)  Main effect of insulin dose (F contrast)           Amygdala  −20, 2, −18  0.05  4.87  PBO < 160 U; 40 U < 160 U; 80 U < 160 U  Linear decrease to insulin dose (T contrast)           Caudate  18, 6, 18  0.04  4.91  PBO > 160 U; 40 U > 160 U   Hypothalamus  −8, −2, −8  0.01a  3.58  PBO > 80 U; P > 160 U  Linear increase to insulin dose (T contrast)           Superior frontal gyrus  18, 39, 45  0.02a  3.92  PBO < 80 U; P < 160 U; 40 U < 80 U; 40 U < 160 U  Brain Regions  MNI Coordinate (x, y, z)  P Value, FWE Corrected  z-Value  Post Hoc Analyses (P < 0.05, Bonferroni Corrected)  Main effect of insulin dose (F contrast)           Amygdala  −20, 2, −18  0.05  4.87  PBO < 160 U; 40 U < 160 U; 80 U < 160 U  Linear decrease to insulin dose (T contrast)           Caudate  18, 6, 18  0.04  4.91  PBO > 160 U; 40 U > 160 U   Hypothalamus  −8, −2, −8  0.01a  3.58  PBO > 80 U; P > 160 U  Linear increase to insulin dose (T contrast)           Superior frontal gyrus  18, 39, 45  0.02a  3.92  PBO < 80 U; P < 160 U; 40 U < 80 U; 40 U < 160 U  Data from full factorial model investigating the effect of insulin dose [placebo (PBO); 40 U, 80 U, and 160 U of insulin]. A statistical threshold of P < 0.05, FWE corrected, was used. a Small volume corrected. View Large All post hoc results were Bonferroni corrected. Behavioral results We acquired subjective feeling of hunger on a visual analog scale before and 60 minutes after nasal spray. We observed no acute effect of insulin dose on hunger and no interaction between insulin dose and time (P > 0.05). There was a significant effect of time on hunger [F(1) = 1.64; P = 0.009]. Furthermore, subjects rated liking and wanting for high-caloric sweet and savory foods 60 minutes after nasal spray. No main effect of dose was observed. However, exploratory correlation analysis revealed a significant positive association between the caudate response to 160 U insulin and liking for sweet foods (r = 0.763; P = 0.008). HRV To assess possible dose-dependent effects on the autonomous nervous system, we recorded ECG during each MRI measurement. We detected a slight decrease in absolute heart rate after all of the tested nasal sprays including placebo [main effect of time: F(2) = 6.4; P = 0.016]. However, there was a significant effect of insulin dose on the normalized high frequency band [F(3) = 3.56; P = 0.047] and a significant insulin dose by time interaction [F(3) = 10.52; P = 0.001]. Post hoc analyses revealed a significant difference between placebo and 160 U response (P = 0.026) and between time point 15 and 30 minutes (P = 0.03). On the normalized low-frequency band, we observed a statistical trend for main effect of dose [F(3) = 3.30; P = 0.058]. Post hoc analysis revealed no noteworthy differences between conditions. To explore brain-peripheral interactions, we correlated the change of the high-frequency band with the brain response to placebo and insulin as recently reported (10). The hypothalamic CBF response correlated positively with the high-frequency change after 160 U of insulin adjusted for BMI (radj = 0.940; P = 0.018). No such correlation was observed after any other insulin dose. Blood data After insulin spray administration, there was a dose-dependent increase in circulating insulin concentrations (PAUC 0–30min = 0.01) of 6.9 ± 19.3 pmol/L after 40 U, 17.2 ± 11.8 pmol/L after 80 U, and 30.9 ± 29.8 pmol/L after 160 U, whereas insulin levels slightly decreased and nonsignificantly after placebo spray by 20.9 ± 23.7 pmol/L. Insulin concentrations reached their peak at 15 minutes post–insulin spray and returned to prespray levels another 15 minutes later (Fig. 4A and 4B). Figure 4. View largeDownload slide Effect of intranasal insulin on peripheral metabolism. At 0 minutes, nasal spray was administered. Filled circles represent measurements after administration of placebo spray, open circles after application of 40 U of nasal insulin, filled triangles after 80 U intranasal insulin, and open triangles after 160 U of insulin spray. (A) Serum insulin concentrations. (B) Absolute change in plasma insulin concentration from baseline (0 minutes). (C) Plasma glucose levels. (D) Serum C-peptide concentrations. Data are means ± SEM. Differences between insulin spray concentrations were addressed by comparisons of AUCs for the indicated time intervals using insulin dose as a continuous variable. Figure 4. View largeDownload slide Effect of intranasal insulin on peripheral metabolism. At 0 minutes, nasal spray was administered. Filled circles represent measurements after administration of placebo spray, open circles after application of 40 U of nasal insulin, filled triangles after 80 U intranasal insulin, and open triangles after 160 U of insulin spray. (A) Serum insulin concentrations. (B) Absolute change in plasma insulin concentration from baseline (0 minutes). (C) Plasma glucose levels. (D) Serum C-peptide concentrations. Data are means ± SEM. Differences between insulin spray concentrations were addressed by comparisons of AUCs for the indicated time intervals using insulin dose as a continuous variable. No correlations between increase in plasma insulin (quantified as both incremental AUC 0 to 30 minutes and peak increase) and any of the brain responses described above were detected (P > 0.05). Neither serum C-peptide concentrations nor plasma glucose levels were different between the days (PAUC 0–120 = 0.9 and PAUC 0–120 = 0.6, respectively; Fig. 4C and 4D). Furthermore, no differences between days were detected for LH, follicle-stimulating hormone, and testosterone, adrenocorticotropic hormone and cortisol, and thyroid-stimulating hormone concentrations (all PAUC 0–120 ≥ 0.6; Supplemental Figs. 1 and 2). Discussion In the current study, we investigated the effect of different insulin doses applied intranasally on the central and autonomous nervous system as well as peripheral metabolism. We detected dose-dependent effects of intranasal insulin on brain activity and regional blood flow. The hypothalamus, amygdala, caudate nucleus, and the lateral PFC showed a most prominent effect for 160 U human insulin compared with placebo. Furthermore, we identified a dose-dependent increase in circulating insulin concentrations as well as an increase in high-frequency band activity of the autonomous nervous system. We were able to replicate previous findings on insulin action in specific human brain areas (9). However, high doses of insulin nasal spray (e.g., at least 80 U) were necessary to acutely introduce detectable changes in brain activity in these regions. Although much lower insulin doses such as 20 U may be beneficial to study chronic effects on complex brain functions (14), acute quantification of regional insulin effects by fMRI seems to require higher doses (e.g., the frequently applied 160 U). As we did not detect noteworthy effects of 20 and 40 U of nasal insulin, our results raise the question of the underlying mechanisms of such chronic effects of low-dose nasal insulin. These may include repeated subthreshold activations of specific regions, but could also be related to changes in the milieu of the brain that might arise under chronic exposure. When addressing effects of nasal insulin on metabolic function, higher insulin doses have a potential disadvantage by temporarily increasing circulating insulin. We precisely quantified this in a dose-dependent manner in this study. As the excursion of insulin is not accompanied by changes in C-peptide, which would indicate endogenous origin, the rise in insulin can only be caused by a spillover of spray into the bloodstream. This spillover has been detected previously after 160 U of nasal insulin and corresponds to an intravenous insulin dose of ∼2.5 mU/kg body weight (12) or an absolute dose of 0.1 U (13) for this nasal insulin dose. Furthermore, the spillover seems to be different between human insulin, as used in our current study, and rapid-acting insulin analogs. Although application of 40 U human insulin caused only minor changes in plasma insulin, 40 U of the rapid-acting insulin lispro introduced marked rises in circulating levels in a recent study (28). Furthermore, the kinetics of intranasal human insulin and the insulin analog lispro seem to be quite different. A delayed absorption of insulin lispro reached peak levels in the blood at least 15 minutes later than human insulin (28). By this time, plasma insulin levels were already back to baseline after human insulin spray. The magnitude of the insulin spillover into circulation was not related to any of the detected brain effects in our study. Hence, penetration of the peptide directly into the brain and not the transport via the bloodstream seems to be the major mode of action of nasal insulin in the CNS. Although not routinely done in the past (15, 20, 29), our current results on insulin spillover indicate that this phenomenon should be mimicked by intravenous insulin application when studying peripheral tissues. Of note, we did so in one recent experiment: In this study, insulin delivery to the brain via insulin nasal spray improved peripheral insulin sensitivity by suppressing endogenous glucose production and stimulating glucose uptake into peripheral tissues independent of insulin spillover (12). The insulin-induced change in HRV substantiates previous results on the role of the autonomous nervous system (20). Intranasal insulin specifically induced change in the high-frequency band, which represents mainly the parasympathetic branch of the autonomous nervous system. This is well in line with animal data showing that central insulin action is transmitted via the major parasympathetic nerve (i.e., the vagus nerve) to peripheral organs (30). Just as in animals, the hypothalamus seems to contribute to this response, as it plays a pivotal role for homeostatic regulation and integrating metabolic signals. Insulin reactivity of the hypothalamus has been shown to be compromised in obese individuals (6, 21, 27). Previous results in humans indicated that parasympathetic outflows from the hypothalamus contribute to the modulation of peripheral insulin sensitivity (20). For other aspects of peripheral metabolism (e.g., lipolysis or liver metabolism), this has not been investigated yet. Our current results indicate that future studies aiming to monitor the autonomous nervous system should apply higher doses of insulin nasal spray, as we only detected notable effects after 160 U. Another possibility of brain-derived modulation of peripheral metabolism is endocrine signals. Potential mechanisms involve the hypothalamic–pituitary–adrenal axis, the hypothalamic–pituitary–gonadal axis, or modulation of thyroid function. Just as previously reported in a larger group for cortisol (10), we detected no effects of nasal insulin on the hypothalamic–pituitary–adrenal axis. Furthermore, the pituitary-gonadal and thyroid axes were unaffected by nasal insulin. However, we cannot exclude that chronic nasal insulin administration may have effect on endocrine functions (especially gonadal), as brain-specific knockout of the insulin receptor impaired LH regulation and thereby reproductive function (31) in rodents. As we did not detect acute effects of nasal insulin on the assessed endocrine functions, it is most likely that the autonomous nervous system is the key transducer for acute peripheral effects. Besides the hypothalamus, a dose-dependent effect of insulin was found in regions recently identified as insulin sensitive with strong relevance for food intake and body weight regulation. The amygdala and lateral PFC showed a marked increase for 160 U of insulin, whereas the caudate nucleus (striatal region) revealed a noteworthy decrease in resting-state activity. The hypothalamus plays a vital role in whole-body energy homeostasis. In animal studies, insulin signaling is similar in the amygdala and the hypothalamus, including obesity-related dysregulation (32). Furthermore, both are embedded in the striato-prefrontal circuitry (33, 34), which is particularly sensitive to increasing peripheral and central insulin levels (for a recent review, see 1, 9, 21). Furthermore, as seen in the current study, intranasal insulin influences neural activity in the striatum and PFC (6, 10, 20, 35–37). Moreover, animal models have shown multiple interactions between homeostatic and reward brain circuits (38), in particular that insulin can amplify dopamine release in the striatum (39). Hence, the marked insulin-induced hypothalamic decrease can potentially increase satiety, whereas the decrease in caudate activity and increase in amygdala and PFC activity to insulin may attenuate the rewarding properties of food and motivation for food consumption. Taken together, we detected dose-dependent effects of intranasal insulin on regional brain activity and parasympathetic tone. Although no acute effects of 40 U of nasal insulin were observed, 160 U of the peptide had the strongest effects. However, this dose was accompanied by a transient increase in circulating insulin concentrations due to a spillover into circulation. This is no major obstacle for the assessment of regional brain effects by imaging but should be mimicked in studies on peripheral metabolism. Taking this into account, intranasal insulin application is currently the best available tool to dissect central from peripheral insulin effects. Our current results can be the basis for the design of future studies with nasal insulin administration to disentangle brain insulin effects in health and disease. Abbreviations: ANOVA analysis of variance AUC area under the curve BMI body mass index BOLD blood oxygenation level–dependent CBF cerebral blood flow CSF cerebrospinal fluid ECG electrocardiogram EPI echo planar imaging fALFF fractional amplitude of low-frequency fluctuation fMRI functional magnetic resonance imaging FOV field of view HOMA-IR homeostatic model assessment of insulin resistance HRV heart rate variability LH luteinizing hormone MRI magnetic resonance imaging PFC prefrontal cortex PFWE P family-wise error rsfMRI resting-state functional magnetic resonance imaging SPM12 Statistical Parametric Mapping 12. Acknowledgments We thank all study participants for cooperation on this project and acknowledge the excellent technical assistance of Dr. Dorothea Baumann, Maike Borutta, Anja Dessecker, Dr. Louise Fritsche, Christoph Gassenmaier, Ellen Kollmar, Dr. Steffen Reichert, and Andreas Vosseler (all from University of Tübingen, Tübingen, Germany). Financial Support: This study was supported in part by Grant 01GI0925 from the Federal Ministry of Education and Research to the German Center for Diabetes Research, Helmholtz Alliance ICEMED: Imaging and Curing Environmental Metabolic Diseases; a European Association for the Study of Diabetes Rising Star award supported by Novo Nordisk (to M.H.); and by the European Union Seventh Framework Programme (FP7/2007–2013) under Grant Agreement 607310 (Nudge-it). Disclosure Summary: The authors have nothing to disclose. References 1. Kullmann S, Heni M, Hallschmid M, Fritsche A, Preissl H, Häring HU. Brain Insulin Resistance at the Crossroads of Metabolic and Cognitive Disorders in Humans. Physiol Rev . 2016; 96( 4): 1169– 1209. Google Scholar CrossRef Search ADS PubMed  2. Heni M, Kullmann S, Preissl H, Fritsche A, Häring HU. Impaired insulin action in the human brain: causes and metabolic consequences. Nat Rev Endocrinol . 2015; 11( 12): 701– 711. Google Scholar CrossRef Search ADS PubMed  3. Kleinridders A, Ferris HA, Cai W, Kahn CR. Insulin action in brain regulates systemic metabolism and brain function. Diabetes . 2014; 63( 7): 2232– 2243. Google Scholar CrossRef Search ADS PubMed  4. Lochhead JJ, Thorne RG. Intranasal delivery of biologics to the central nervous system. Adv Drug Deliv Rev . 2012; 64( 7): 614– 628. Google Scholar CrossRef Search ADS PubMed  5. Born J, Lange T, Kern W, McGregor GP, Bickel U, Fehm HL. Sniffing neuropeptides: a transnasal approach to the human brain. Nat Neurosci . 2002; 5( 6): 514– 516. Google Scholar CrossRef Search ADS PubMed  6. Kullmann S, Heni M, Veit R, Scheffler K, Machann J, Häring HU, Fritsche A, Preissl H. Selective insulin resistance in homeostatic and cognitive control brain areas in overweight and obese adults. Diabetes Care . 2015; 38( 6): 1044– 1050. Google Scholar CrossRef Search ADS PubMed  7. Benedict C, Kern W, Schultes B, Born J, Hallschmid M. Differential sensitivity of men and women to anorexigenic and memory-improving effects of intranasal insulin. J Clin Endocrinol Metab . 2008; 93( 4): 1339– 1344. Google Scholar CrossRef Search ADS PubMed  8. Reger MA, Watson GS, Frey WH II, Baker LD, Cholerton B, Keeling ML, Belongia DA, Fishel MA, Plymate SR, Schellenberg GD, Cherrier MM, Craft S. Effects of intranasal insulin on cognition in memory-impaired older adults: modulation by APOE genotype. Neurobiol Aging . 2006; 27( 3): 451– 458. Google Scholar CrossRef Search ADS PubMed  9. Kullmann S, Heni M, Fritsche A, Preissl H. Insulin action in the human brain: evidence from neuroimaging studies. J Neuroendocrinol . 2015; 27( 6): 419– 423. Google Scholar CrossRef Search ADS PubMed  10. Heni M, Kullmann S, Ketterer C, Guthoff M, Linder K, Wagner R, Stingl KT, Veit R, Staiger H, Häring HU, Preissl H, Fritsche A. Nasal insulin changes peripheral insulin sensitivity simultaneously with altered activity in homeostatic and reward-related human brain regions. Diabetologia . 2012; 55( 6): 1773– 1782. Google Scholar CrossRef Search ADS PubMed  11. Dash S, Xiao C, Morgantini C, Koulajian K, Lewis GF. Is insulin action in the brain relevant in regulating blood glucose in humans? J Clin Endocrinol Metab . 2015; 100( 7): 2525– 2531. Google Scholar CrossRef Search ADS PubMed  12. Heni M, Wagner R, Kullmann S, Gancheva S, Roden M, Peter A, Stefan N, Preissl H, Häring HU, Fritsche A. Hypothalamic and striatal insulin action suppresses endogenous glucose production and may stimulate glucose uptake during hyperinsulinemia in lean but not in overweight men. Diabetes . 2017; 66( 7): 1797– 1806. Google Scholar CrossRef Search ADS PubMed  13. Gancheva S, Koliaki C, Bierwagen A, Nowotny P, Heni M, Fritsche A, Häring HU, Szendroedi J, Roden M. Effects of intranasal insulin on hepatic fat accumulation and energy metabolism in humans. Diabetes . 2015; 64( 6): 1966– 1975. Google Scholar CrossRef Search ADS PubMed  14. Craft S, Baker LD, Montine TJ, Minoshima S, Watson GS, Claxton A, Arbuckle M, Callaghan M, Tsai E, Plymate SR, Green PS, Leverenz J, Cross D, Gerton B. Intranasal insulin therapy for Alzheimer disease and amnestic mild cognitive impairment: a pilot clinical trial. Arch Neurol . 2012; 69( 1): 29– 38. Google Scholar CrossRef Search ADS PubMed  15. Benedict C, Brede S, Schiöth HB, Lehnert H, Schultes B, Born J, Hallschmid M. Intranasal insulin enhances postprandial thermogenesis and lowers postprandial serum insulin levels in healthy men. Diabetes . 2011; 60( 1): 114– 118. Google Scholar CrossRef Search ADS PubMed  16. Benedict C, Hallschmid M, Hatke A, Schultes B, Fehm HL, Born J, Kern W. Intranasal insulin improves memory in humans. Psychoneuroendocrinology . 2004; 29( 10): 1326– 1334. Google Scholar CrossRef Search ADS PubMed  17. Benedict C, Hallschmid M, Schultes B, Born J, Kern W. Intranasal insulin to improve memory function in humans. Neuroendocrinology . 2007; 86( 2): 136– 142. Google Scholar CrossRef Search ADS PubMed  18. Benedict L, Nelson CA, Schunk E, Sullwold K, Seaquist ER. Effect of insulin on the brain activity obtained during visual and memory tasks in healthy human subjects. Neuroendocrinology . 2006; 83( 1): 20– 26. Google Scholar CrossRef Search ADS PubMed  19. Kullmann S, Frank S, Heni M, Ketterer C, Veit R, Häring HU, Fritsche A, Preissl H. Intranasal insulin modulates intrinsic reward and prefrontal circuitry of the human brain in lean women. Neuroendocrinology . 2013; 97( 2): 176– 182. Google Scholar CrossRef Search ADS PubMed  20. Heni M, Wagner R, Kullmann S, Veit R, Mat Husin H, Linder K, Benkendorff C, Peter A, Stefan N, Häring HU, Preissl H, Fritsche A. Central insulin administration improves whole-body insulin sensitivity via hypothalamus and parasympathetic outputs in men. Diabetes . 2014; 63( 12): 4083– 4088. Google Scholar CrossRef Search ADS PubMed  21. Kullmann S, Heni M, Veit R, Scheffler K, Machann J, Häring HU, Fritsche A, Preissl H. Intranasal insulin enhances brain functional connectivity mediating the relationship between adiposity and subjective feeling of hunger. Sci Rep . 2017; 7( 1): 1627. Google Scholar CrossRef Search ADS PubMed  22. Pohmann R, Budde J, Auerbach EJ, Adriany G, Uğurbil K. Theoretical and experimental evaluation of continuous arterial spin labeling techniques. Magn Reson Med . 2010; 63( 2): 438– 446. Google Scholar CrossRef Search ADS PubMed  23. Alsop DC, Detre JA, Golay X, Günther M, Hendrikse J, Hernandez-Garcia L, Lu H, MacIntosh BJ, Parkes LM, Smits M, van Osch MJ, Wang DJ, Wong EC, Zaharchuk G. Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: A consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia. Magn Reson Med . 2015; 73( 1): 102– 116. Google Scholar CrossRef Search ADS PubMed  24. Chao-Gan Y, Yu-Feng Z. DPARSF: a MATLAB toolbox for “pipeline” data analysis of resting-state fMRI. Front Syst Neurosci . 2010; 4: 13. Google Scholar PubMed  25. Song XW, Dong ZY, Long XY, Li SF, Zuo XN, Zhu CZ, He Y, Yan CG, Zang YF. REST: a toolkit for resting-state functional magnetic resonance imaging data processing. PLoS One . 2011; 6( 9): e25031. Google Scholar CrossRef Search ADS PubMed  26. Zou QH, Zhu CZ, Yang Y, Zuo XN, Long XY, Cao QJ, Wang YF, Zang YF. An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: fractional ALFF. J Neurosci Methods . 2008; 172( 1): 137– 141. Google Scholar CrossRef Search ADS PubMed  27. Kullmann S, Fritsche A, Wagner R, Schwab S, Häring HU, Preissl H, Heni M. Hypothalamic insulin responsiveness is associated with pancreatic insulin secretion in humans. Physiol Behav . 2017; 176: 134– 138. Google Scholar CrossRef Search ADS PubMed  28. Dash S, Xiao C, Morgantini C, Koulajian K, Lewis GF. Intranasal insulin suppresses endogenous glucose production in humans compared with placebo in the presence of similar venous insulin concentrations. Diabetes . 2015; 64( 3): 766– 774. Google Scholar CrossRef Search ADS PubMed  29. Iwen KA, Scherer T, Heni M, Sayk F, Wellnitz T, Machleidt F, Preissl H, Häring HU, Fritsche A, Lehnert H, Buettner C, Hallschmid M. Intranasal insulin suppresses systemic but not subcutaneous lipolysis in healthy humans. J Clin Endocrinol Metab . 2014; 99( 2): E246– E251. Google Scholar CrossRef Search ADS PubMed  30. Pocai A, Lam TK, Gutierrez-Juarez R, Obici S, Schwartz GJ, Bryan J, Aguilar-Bryan L, Rossetti L. Hypothalamic K(ATP) channels control hepatic glucose production. Nature . 2005; 434( 7036): 1026– 1031. Google Scholar CrossRef Search ADS PubMed  31. Brüning JC, Gautam D, Burks DJ, Gillette J, Schubert M, Orban PC, Klein R, Krone W, Müller-Wieland D, Kahn CR. Role of brain insulin receptor in control of body weight and reproduction. Science . 2000; 289( 5487): 2122– 2125. Google Scholar CrossRef Search ADS PubMed  32. Areias MF, Prada PO. Mechanisms of insulin resistance in the amygdala: influences on food intake. Behav Brain Res . 2015; 282: 209– 217. Google Scholar CrossRef Search ADS PubMed  33. Kullmann S, Heni M, Linder K, Zipfel S, Häring HU, Veit R, Fritsche A, Preissl H. Resting-state functional connectivity of the human hypothalamus. Hum Brain Mapp . 2014; 35( 12): 6088– 6096. Google Scholar CrossRef Search ADS PubMed  34. Roy AK, Shehzad Z, Margulies DS, Kelly AM, Uddin LQ, Gotimer K, Biswal BB, Castellanos FX, Milham MP. Functional connectivity of the human amygdala using resting state fMRI. Neuroimage . 2009; 45( 2): 614– 626. Google Scholar CrossRef Search ADS PubMed  35. Stingl KT, Kullmann S, Guthoff M, Heni M, Fritsche A, Preissl H. Insulin modulation of magnetoencephalographic resting state dynamics in lean and obese subjects. Front Syst Neurosci . 2010; 4: 157. Google Scholar CrossRef Search ADS PubMed  36. Guthoff M, Grichisch Y, Canova C, Tschritter O, Veit R, Hallschmid M, Häring HU, Preissl H, Hennige AM, Fritsche A. Insulin modulates food-related activity in the central nervous system. J Clin Endocrinol Metab . 2010; 95( 2): 748– 755. Google Scholar CrossRef Search ADS PubMed  37. Heni M, Kullmann S, Ahlqvist E, Wagner R, Machicao F, Staiger H, Häring HU, Almgren P, Groop LC, Small DM, Fritsche A, Preissl H. Interaction between the obesity-risk gene FTO and the dopamine D2 receptor gene ANKK1/TaqIA on insulin sensitivity. Diabetologia . 2016; 59( 12): 2622– 2631. Google Scholar CrossRef Search ADS PubMed  38. Ferrario CR, Labouèbe G, Liu S, Nieh EH, Routh VH, Xu S, O’Connor EC. Homeostasis meets motivation in the battle to control food intake. J Neurosci . 2016; 36( 45): 11469– 11481. Google Scholar CrossRef Search ADS PubMed  39. Stouffer MA, Woods CA, Patel JC, Lee CR, Witkovsky P, Bao L, Machold RP, Jones KT, de Vaca SC, Reith ME, Carr KD, Rice ME. Insulin enhances striatal dopamine release by activating cholinergic interneurons and thereby signals reward. Nat Commun . 2015; 6: 8543. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Loading next page...
 
/lp/ou_press/dose-dependent-effects-of-intranasal-insulin-on-resting-state-brain-GkNWwYDNPL
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-01976
Publisher site
See Article on Publisher Site

Abstract

Abstract Context Insulin action in the human brain influences eating behavior, cognition, and whole-body metabolism. Studies investigating brain insulin rely on intranasal application. Objective To investigate effects of three doses of insulin and placebo as nasal sprays on the central and autonomous nervous system and analyze absorption of insulin into the bloodstream. Design, Participants, and Methods Nine healthy men received placebo or 40 U, 80 U, and 160 U insulin spray in randomized order. Before and after spray, brain activity was assessed by functional magnetic resonance imaging, and heart rate variability (HRV) was assessed from electrocardiogram. Plasma insulin, C-peptide, and glucose were measured regularly. Setting General community. Results Nasal insulin administration dose-dependently modulated regional brain activity and the normalized high-frequency component of the HRV. Post hoc analyses revealed that only 160 U insulin showed a considerable difference from placebo. Dose-dependent spillover of nasal insulin into the bloodstream was detected. The brain response was not correlated with this temporary rise in circulating insulin. Conclusions Nasal insulin dose-dependently modulated regional brain activity with the strongest effects after 160 U. However, this dose was accompanied by a transient increase in circulating insulin concentrations due to a spillover into circulation. Our current results may serve as a basis for future studies with nasal insulin to untangle brain insulin effects in health and disease. Research over the last several years identified the human brain as an insulin-sensitive organ (1, 2). In response to the hormone, the central nervous system regulates various functions as the response to food cues, reward processes, and memory (1). Furthermore, insulin effects in the brain influence the rest of the body, modulating peripheral insulin sensitivity, thermogenesis, as well as liver and lipid metabolism (2, 3). However, a substantial number of individuals are brain insulin resistant and therefore lack these effects of the peptide. The best-characterized factor associated with brain insulin resistance is body weight with reduced insulin effects in overweight and obese persons. Beyond this, a number of other factors have been identified thus far (1, 2). However, it is still not known whether factors associated with brain insulin resistance represent cause or consequence thereof. To assess brain-specific insulin effects in humans, most research applied insulin as a nasal spray (2). Research in animals demonstrated that peptide transporters near the olfactory bulb rapidly transport a limited number of peptides (including insulin) into the cerebrospinal fluid (CSF), from which they further reach brain cells (4). In humans, a rise in CSF insulin concentrations was detected as early as 10 minutes after administration of 40 U of the peptide as nasal spray (5). Functional consequences (e.g., effects on brain activity) occur in a comparable time frame (6). The earliest experiments mostly used a dose of 160 U to investigate the consequences of intranasal insulin application on eating behavior and body weight; studies investigating memory function in cognitively impaired patients, in contrast, used a lower dose of insulin (7, 8). An acute dose of 160 U has been shown to stimulate regional brain activity in a number of brain areas (9). Although initial experiments with a lower insulin dose suggested that none of the peptide reached the circulation after administration as a nasal spray (5), it recently became evident that small amounts of human insulin are indeed absorbed into the bloodstream and are detectable in venous blood (10–12). After application of 160 U of insulin as a nasal spray, ∼0.1 U enter the bloodstream (12, 13). Despite having no major effects on blood glucose (e.g., not inducing hypoglycemia), the small rise in circulating insulin may induce effects in peripheral tissues, possibly influencing the interpretation of metabolic effects of nasal insulin. Furthermore, experiments with chronic administration of human insulin as nasal spray suggested that high doses might not necessarily have the strongest effects. Indeed, in a 4-month trial in cognitive-impaired persons, the lower dose of 20 U of the peptide had more profound effects on memory than 40 U (14). Even though the intranasal administration has already been used in a number of studies in humans (6, 7, 15–21), there is still no systematic comparison of the acute effect of different nasal insulin doses on central and autonomous nervous system activity. We therefore investigated the effects of placebo and three different doses of human insulin as nasal spray. Before and after spray, brain activity was assessed by functional magnetic resonance imaging (fMRI) while simultaneously recording electrocardiograms (ECGs) to assess heart rate variability (HRV). Frequent blood sampling was used to analyze the absorption of insulin into the bloodstream. Material and Methods Participants Nine healthy men participated in the study [body mass index (BMI) 20 to 26 kg/m2; age 23 to 30 years] (Table 1). Informed written consent was obtained from all subjects, and the local ethics committee approved the protocol. To ensure that participants were healthy and did not suffer from psychiatric, neurologic, or metabolic diseases, they underwent a thorough medical examination. Table 1. Characteristics of Nine Male Participants Variable  Mean ± SD (Range)  Age (y)  26.56 ± 2.78 (23–30)  BMI (kg/m2)  23.44 ± 2.01 (20.09–26.01)  Body fat content (%)  18.9 ± 2.4 (16.6–21.9)  Waist-to-hip ratio  0.91 ± 0.07 (0.81–0.99)  Hemoglobin A1C (%)  5.2 ± 0.1 (5.1–5.4)  HOMA-IR (mean over all 4 measurement days)  2.25 ± 0.41 (1.51–2.91)  Variable  Mean ± SD (Range)  Age (y)  26.56 ± 2.78 (23–30)  BMI (kg/m2)  23.44 ± 2.01 (20.09–26.01)  Body fat content (%)  18.9 ± 2.4 (16.6–21.9)  Waist-to-hip ratio  0.91 ± 0.07 (0.81–0.99)  Hemoglobin A1C (%)  5.2 ± 0.1 (5.1–5.4)  HOMA-IR (mean over all 4 measurement days)  2.25 ± 0.41 (1.51–2.91)  Abbreviation: HOMA-IR, homeostatic model assessment of insulin resistance. View Large Study design Volunteers received placebo or 40 U, 80 U, and 160 U insulin as nasal spray in randomized order (on 4 separate days with 7 to 14 days’ time lag) with repetitive MRI and ECG measurements (Fig. 1). Participants were blinded to the condition. Experiments were conducted after an overnight fast and started at 7:00 am with a baseline MRI measurement including two blood oxygenation level–dependent (BOLD) resting-state functional MRI (rsfMRI1) and a cerebral blood flow (CBF) measurement (CBF1). After the basal measurement, the respective nasal spray was administered. After 15 and 30 minutes, second and third MRI measurements were performed (rsfMRI2/CBF2 and rsfMRI3/CBF3). During all three MRI measurements, ECG was recorded. The subjective feeling of hunger was rated at two time points (before spray application 60 minutes after intranasal spray) on a visual analog scale from 0 to 10 (0, not hungry at all; 10, very hungry). Figure 1. View largeDownload slide Schematic overview of study design. CBF and rsfMRI for subcortical regions and whole-brain were acquired at different time points before and after intranasal placebo and insulin sprays. ECG was recorded during the entire MRI acquisition. Blood samples, indicated by red symbols, were taken before and 5, 10, 15, 30, 60, 90, and 120 minutes after intranasal placebo and insulin sprays. Figure 1. View largeDownload slide Schematic overview of study design. CBF and rsfMRI for subcortical regions and whole-brain were acquired at different time points before and after intranasal placebo and insulin sprays. ECG was recorded during the entire MRI acquisition. Blood samples, indicated by red symbols, were taken before and 5, 10, 15, 30, 60, 90, and 120 minutes after intranasal placebo and insulin sprays. Participants rated 100 food pictures in two separate blocks according to explicit “liking” (how much do you like the food item in general) and “wanting” (how much would you like to eat the food item right now) on each study day. Application intranasal insulin/placebo On each day, participants received 1.6 mL of nasal spray. It contained either placebo or 40, 80, or 160 U of human insulin (Insulin Actrapid; Novo Nordisk, Bagsvaerd, Denmark). Under supervision, the spray was administered over 4 minutes with two puffs per nostril every minute. Blood measurements Venous blood samples were obtained immediately before as well as 5, 10, 15, 20, 30, 60, 90, and 120 minutes after spray application. Plasma glucose concentrations were measured using the glucose-oxidase method (Yellow Springs Instruments, Yellow Springs, OH). Serum insulin, C-peptide, cortisol, luteinizing hormone (LH), follicle-stimulating hormone, testosterone, and thyroid-stimulating hormone were measured by chemiluminescence assays on the ADVIA Centaur XPT and adrenocorticotropic hormone by an automated solid-phase, chemiluminescent immunoassay on the Immulite XPT analyzer (both from Siemens Healthineers, Eschborn, Germany). Whole-brain fMRI measurement Data acquisition Scanning was conducted at a 3T whole-body Siemens scanner (Magnetom Prisma; Siemens, Erlangen, Germany) with a 20-channel coil. Three different types of functional data sets were recorded each day before and after nasal spray application. In addition, high-resolution T1-weighted anatomical images were obtained. To acquire CBF maps, pseudocontinuous arterial spin labeling was performed using a two-dimensional echo planar imaging (EPI) readout sequence (22) with background suppression. A total of 52 images were acquired with the following parameters: 16 slices; slice thickness 3 mm; 1.5-mm gap; repetition time 4500 ms; echo time 23 ms; field of view (FOV)read 192 mm2; FOVphase 100%; matrix 64 × 64;, flip angle 90°; voxel size 3 × 3 × 3 mm3; bandwidth 2004 px/Hz; and tag gradient strength 7.0 mT/m. The first image volume prior to the preparation scans was used for calibration (M0). Whole-brain BOLD data were collected by using EPI sequences, as recently reported (21). To acquire BOLD activation of subcortical regions at a higher resolution, an EPI sequence with ZOOMit was used. It uses dynamic excitation pulses to achieve selective FOV (zoomed) images, without aliasing artifacts. The following parameters were used: repetition time 3 s; echo time 34 ms; FOVread 192 mm2; FOVphase 33.3%; matrix 96 × 32 × 36; flip angle 90°; voxel size 2 × 2 × 2.5 mm3; and slice thickness 2 mm. Images were acquired in ascending order. Each brain volume comprised 36 axial slices, and each functional run contained 60 image volumes, resulting in a total scan time of 3:06 minutes. Before intranasal insulin application and after 30 minutes, all fMRI measurements were performed as displayed in Fig. 1. Due to time constraints, only CBF and BOLD of subcortical regions were performed after 15 minutes. Pseudocontinuous arterial spin labeling preprocessing Preprocessing was performed using FSL with the following tools (5.0.9). First, the images were realigned using mcflirt. The resulting four-dimensional ASL data were processed using oxford_asl. For CBF quantification, a single-compartment standard kinetic model was used (23). Perfusion images (ΔM) were obtained by pairwise subtraction of tag and control images. A voxel-wise calibration was performed, thereby correcting possible radiofrequency coil inhomogeneity (23). In a next step, the T1-image was coregistered with the M0-image, and the resulting transformation parameters were applied to the CBF maps. Finally, the CBF maps were normalized in MNI space using Statistical Parametric Mapping 12 (SPM12). rsfMRI data preprocessing We used the Data Processing Assistant for Resting-State fMRI (24) to analyze the resting state fMRI data, which is based on SPM12 and the Resting-State fMRI Data Analysis Toolkit (25) (http://www.restfmri.net). The whole-brain functional images were normalized to voxel size, 3 × 3 × 3 mm3, and the subcortical functional images to 2 × 2 × 2 mm3 and then smoothed (full width at half maximum: 6 mm for whole-brain and 4 mm for the subcortical images). Nuisance regression was performed using white matter, CSF, and the six head motion parameters as covariates. To investigate resting-state brain activity, we calculated the fractional amplitude of low-frequency fluctuation (fALFF; 0.01 to 0.08 Hz) of the BOLD signal (26). The regional intensity of spontaneous BOLD fluctuations is quantified by the power spectrum in the low-frequency range (0.009 to 0.08 Hz) and regularized by the power in the whole-frequency range (0 to 0.25 Hz) (26). Statistical analyses CBF. CBF values were extracted of the hypothalamic region of interest based on recent findings. In this study, we identified the hypothalamus to respond with a persistent decrease in activity up to 30 minutes after intranasal insulin application (6, 10, 19, 20). This response was diminished in individuals with unfavorable fat distribution and whole-body insulin resistance. Baseline-corrected CBF maps were computed to quantify the hypothalamic CBF change 15 and 30 minutes after intranasal insulin application (CBF2 − CBF1 and CBF3 − CBF1, respectively). Repeated-measurement analysis of variance (ANOVA; factor insulin dose with four levels: placebo, 40 U, 80 U, and 160 U insulin and factor time with two levels) with homeostatic model assessment of insulin resistance (HOMA-IR) as covariate was performed in SPSS (version 20; IBM, Armonk, NY) (P < 0.05). rsfMRI. fALFF maps of the subcortical regions were baseline corrected (rsfMRI2 − rsfMRI1 and rsfMRI3 − rsfMRI1). Repeated-measurement ANOVA (full factorial model) was performed in SPM12. Relevant clusters were extracted for post hoc analyses in SPSS (version 20; IBM). A statistical threshold of P family-wise error (PFWE) <0.05 voxel-level whole-brain corrected was applied. Additionally, small volume correction was performed for the hypothalamus, an a priori region of interest (6, 10). All whole-brain fALFF maps were baseline corrected (rsfMRI3 − fMRI1). Repeated-measurement ANOVA/full factorial model was performed in SPM12. Relevant clusters were extracted for post hoc analyses in SPSS (version 20; IBM). A statistical threshold of PFWE < 0.05 voxel-level whole-brain corrected was applied. Additionally, small volume correction was performed for the prefrontal cortex (PFC), an a priori region of interest, based on recent findings showing that activity and functional connectivity can be enhanced with intranasal insulin (6, 19, 21). HRV. ECGs were recorded with BIOPAC MP35 (BIOPAC, Goleta, CA) and a sampling rate of 1000 Hz during the whole fMRI recording. For analysis, we extracted 5 minutes of recording during the CBF measurements. The data were analyzed with Kubios (http://kubios.uef.fi). HRV parameters were calculated on individual R-R interval time series in the low-frequency (0.04 to 0.15 Hz) and high-frequency (0.15 to 0.40 Hz) bands. HRV parameters were baseline corrected and analyzed using a repeated-measurement ANOVA in SPSS (version 20; IBM) (P < 0.05). Blood values. The areas under the curve (AUCs) were calculated using the trapezoid rule. Doses were compared by linear regression analyses with dose as a continuous variable. Results Dose-dependent effects of intranasal insulin on brain activity Hypothalamic CBF change Based on our recent finding, we assessed dose-dependent insulin-induced hypothalamic change in CBF as readout for hypothalamic insulin sensitivity (6, 10, 12, 27). We identified a significant main effect of dose [F(3) = 5.36; P = 0.01] (Fig. 2) and a significant interaction between dose and peripheral insulin sensitivity assessed by HOMA-IR [F(3) = 4.43; P = 0.02]. No main effect of time (15 vs 30 minutes postspray) was observed. Post hoc analyses revealed a significant CBF difference between placebo and 160 U of intranasal insulin (P = 0.04, Bonferroni corrected). Furthermore, a significant positive correlation was observed between HOMA-IR and the hypothalamic response to 160 U of insulin 30 minutes after application (r = 0.76; P = 0.017). Figure 2. View largeDownload slide Dose-dependent intranasal insulin effect on hypothalamic CBF. Bar plot shows the extracted CBF values adjusted for HOMA-IR from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between placebo and 160 U of insulin (*P < 0.05, Bonferroni corrected). Figure 2. View largeDownload slide Dose-dependent intranasal insulin effect on hypothalamic CBF. Bar plot shows the extracted CBF values adjusted for HOMA-IR from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between placebo and 160 U of insulin (*P < 0.05, Bonferroni corrected). Subcortical rsfMRI response using fALFF We observed a significant main effect of dose in the left amygdala (PFWE = 0.05) (Fig. 3A and Table 1). A significant linear decrease with insulin dose was observed in the right caudate nucleus (PFWE = 0.04) (Fig. 3B and Table 1) and the hypothalamus (PFWE = 0.015, small volume corrected) (Fig. 3C and Table 1). Because no main effect of time or interaction between time and condition were observed (PFWE > 0.05), post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. Post hoc analyses showed a significant fALFF difference in the hypothalamus between placebo and 160 U insulin sprays (P = 0.003) and between placebo and 80 U (P = 0.014). For the left amygdala, we observed a significant difference between placebo and 160 U (P < 0.001), between 40 U and 160 U (P < 0.001), and between 80 U and 160 U (P = 0.004) and a statistical trend for difference between placebo and 80 U (P = 0.07). For the caudate response, we observed a significant difference between placebo and 160 U (P = 0.003) and between 40 U and 160 U (P = 0.003). No interactions were observed with peripheral insulin sensitivity assessed as HOMA-IR. Figure 3. View largeDownload slide Dose-dependent intranasal insulin effect on rsfMRI response. Bar plots show the extracted mean z-values of sizeable fALFF clusters from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between conditions as indicated by asterisks (*P < 0.05; **P < 0.005, Bonferroni corrected). (A) Main effect of insulin dose in the left amygdala displayed by yellow color-coded F-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (B) Linear decrease of insulin dose in the right caudate displayed by yellow color-coded T-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (C) Linear decrease of insulin dose in the hypothalamus displayed by yellow color-coded T-value map (Psvc <0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (D) Linear increase of insulin dose in the PFC displayed by yellow color-coded T-value map (Psvc < 0.05, FWE corrected). Bar plot shows the extracted mean z-values of PFC fALFF from before to 30 minutes after nasal spray application for the different conditions. Figure 3. View largeDownload slide Dose-dependent intranasal insulin effect on rsfMRI response. Bar plots show the extracted mean z-values of sizeable fALFF clusters from before to after nasal spray application of placebo and 40 U, 80 U, and 160 U insulin. Post hoc analyses showed significant differences between conditions as indicated by asterisks (*P < 0.05; **P < 0.005, Bonferroni corrected). (A) Main effect of insulin dose in the left amygdala displayed by yellow color-coded F-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (B) Linear decrease of insulin dose in the right caudate displayed by yellow color-coded T-value map (P < 0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (C) Linear decrease of insulin dose in the hypothalamus displayed by yellow color-coded T-value map (Psvc <0.05, FWE corrected). Because no main effects of time or interaction between time and condition were observed, post hoc analyses were performed on mean response 15 and 30 minutes after nasal spray. (D) Linear increase of insulin dose in the PFC displayed by yellow color-coded T-value map (Psvc < 0.05, FWE corrected). Bar plot shows the extracted mean z-values of PFC fALFF from before to 30 minutes after nasal spray application for the different conditions. Whole-brain rsfMRI using fALFF We observed a significant linear increase in the lateral PFC (PFWE = 0.02 small volume corrected) (Fig. 3D and Table 1) with increasing insulin doses. Post hoc analyses showed a significant difference between placebo and 160 U (P < 0.001), between placebo and 80 U (P = 0.015), between 40 U and 160 U (P = 0.001), and between 40 U and 80 U (P = 0.045). No interactions were observed with HOMA-IR. Whole-brain and subcortical rsfMRI results are summarized in Table 2. Table 2. rsfMRI Response to Different Concentrations of Insulin Brain Regions  MNI Coordinate (x, y, z)  P Value, FWE Corrected  z-Value  Post Hoc Analyses (P < 0.05, Bonferroni Corrected)  Main effect of insulin dose (F contrast)           Amygdala  −20, 2, −18  0.05  4.87  PBO < 160 U; 40 U < 160 U; 80 U < 160 U  Linear decrease to insulin dose (T contrast)           Caudate  18, 6, 18  0.04  4.91  PBO > 160 U; 40 U > 160 U   Hypothalamus  −8, −2, −8  0.01a  3.58  PBO > 80 U; P > 160 U  Linear increase to insulin dose (T contrast)           Superior frontal gyrus  18, 39, 45  0.02a  3.92  PBO < 80 U; P < 160 U; 40 U < 80 U; 40 U < 160 U  Brain Regions  MNI Coordinate (x, y, z)  P Value, FWE Corrected  z-Value  Post Hoc Analyses (P < 0.05, Bonferroni Corrected)  Main effect of insulin dose (F contrast)           Amygdala  −20, 2, −18  0.05  4.87  PBO < 160 U; 40 U < 160 U; 80 U < 160 U  Linear decrease to insulin dose (T contrast)           Caudate  18, 6, 18  0.04  4.91  PBO > 160 U; 40 U > 160 U   Hypothalamus  −8, −2, −8  0.01a  3.58  PBO > 80 U; P > 160 U  Linear increase to insulin dose (T contrast)           Superior frontal gyrus  18, 39, 45  0.02a  3.92  PBO < 80 U; P < 160 U; 40 U < 80 U; 40 U < 160 U  Data from full factorial model investigating the effect of insulin dose [placebo (PBO); 40 U, 80 U, and 160 U of insulin]. A statistical threshold of P < 0.05, FWE corrected, was used. a Small volume corrected. View Large All post hoc results were Bonferroni corrected. Behavioral results We acquired subjective feeling of hunger on a visual analog scale before and 60 minutes after nasal spray. We observed no acute effect of insulin dose on hunger and no interaction between insulin dose and time (P > 0.05). There was a significant effect of time on hunger [F(1) = 1.64; P = 0.009]. Furthermore, subjects rated liking and wanting for high-caloric sweet and savory foods 60 minutes after nasal spray. No main effect of dose was observed. However, exploratory correlation analysis revealed a significant positive association between the caudate response to 160 U insulin and liking for sweet foods (r = 0.763; P = 0.008). HRV To assess possible dose-dependent effects on the autonomous nervous system, we recorded ECG during each MRI measurement. We detected a slight decrease in absolute heart rate after all of the tested nasal sprays including placebo [main effect of time: F(2) = 6.4; P = 0.016]. However, there was a significant effect of insulin dose on the normalized high frequency band [F(3) = 3.56; P = 0.047] and a significant insulin dose by time interaction [F(3) = 10.52; P = 0.001]. Post hoc analyses revealed a significant difference between placebo and 160 U response (P = 0.026) and between time point 15 and 30 minutes (P = 0.03). On the normalized low-frequency band, we observed a statistical trend for main effect of dose [F(3) = 3.30; P = 0.058]. Post hoc analysis revealed no noteworthy differences between conditions. To explore brain-peripheral interactions, we correlated the change of the high-frequency band with the brain response to placebo and insulin as recently reported (10). The hypothalamic CBF response correlated positively with the high-frequency change after 160 U of insulin adjusted for BMI (radj = 0.940; P = 0.018). No such correlation was observed after any other insulin dose. Blood data After insulin spray administration, there was a dose-dependent increase in circulating insulin concentrations (PAUC 0–30min = 0.01) of 6.9 ± 19.3 pmol/L after 40 U, 17.2 ± 11.8 pmol/L after 80 U, and 30.9 ± 29.8 pmol/L after 160 U, whereas insulin levels slightly decreased and nonsignificantly after placebo spray by 20.9 ± 23.7 pmol/L. Insulin concentrations reached their peak at 15 minutes post–insulin spray and returned to prespray levels another 15 minutes later (Fig. 4A and 4B). Figure 4. View largeDownload slide Effect of intranasal insulin on peripheral metabolism. At 0 minutes, nasal spray was administered. Filled circles represent measurements after administration of placebo spray, open circles after application of 40 U of nasal insulin, filled triangles after 80 U intranasal insulin, and open triangles after 160 U of insulin spray. (A) Serum insulin concentrations. (B) Absolute change in plasma insulin concentration from baseline (0 minutes). (C) Plasma glucose levels. (D) Serum C-peptide concentrations. Data are means ± SEM. Differences between insulin spray concentrations were addressed by comparisons of AUCs for the indicated time intervals using insulin dose as a continuous variable. Figure 4. View largeDownload slide Effect of intranasal insulin on peripheral metabolism. At 0 minutes, nasal spray was administered. Filled circles represent measurements after administration of placebo spray, open circles after application of 40 U of nasal insulin, filled triangles after 80 U intranasal insulin, and open triangles after 160 U of insulin spray. (A) Serum insulin concentrations. (B) Absolute change in plasma insulin concentration from baseline (0 minutes). (C) Plasma glucose levels. (D) Serum C-peptide concentrations. Data are means ± SEM. Differences between insulin spray concentrations were addressed by comparisons of AUCs for the indicated time intervals using insulin dose as a continuous variable. No correlations between increase in plasma insulin (quantified as both incremental AUC 0 to 30 minutes and peak increase) and any of the brain responses described above were detected (P > 0.05). Neither serum C-peptide concentrations nor plasma glucose levels were different between the days (PAUC 0–120 = 0.9 and PAUC 0–120 = 0.6, respectively; Fig. 4C and 4D). Furthermore, no differences between days were detected for LH, follicle-stimulating hormone, and testosterone, adrenocorticotropic hormone and cortisol, and thyroid-stimulating hormone concentrations (all PAUC 0–120 ≥ 0.6; Supplemental Figs. 1 and 2). Discussion In the current study, we investigated the effect of different insulin doses applied intranasally on the central and autonomous nervous system as well as peripheral metabolism. We detected dose-dependent effects of intranasal insulin on brain activity and regional blood flow. The hypothalamus, amygdala, caudate nucleus, and the lateral PFC showed a most prominent effect for 160 U human insulin compared with placebo. Furthermore, we identified a dose-dependent increase in circulating insulin concentrations as well as an increase in high-frequency band activity of the autonomous nervous system. We were able to replicate previous findings on insulin action in specific human brain areas (9). However, high doses of insulin nasal spray (e.g., at least 80 U) were necessary to acutely introduce detectable changes in brain activity in these regions. Although much lower insulin doses such as 20 U may be beneficial to study chronic effects on complex brain functions (14), acute quantification of regional insulin effects by fMRI seems to require higher doses (e.g., the frequently applied 160 U). As we did not detect noteworthy effects of 20 and 40 U of nasal insulin, our results raise the question of the underlying mechanisms of such chronic effects of low-dose nasal insulin. These may include repeated subthreshold activations of specific regions, but could also be related to changes in the milieu of the brain that might arise under chronic exposure. When addressing effects of nasal insulin on metabolic function, higher insulin doses have a potential disadvantage by temporarily increasing circulating insulin. We precisely quantified this in a dose-dependent manner in this study. As the excursion of insulin is not accompanied by changes in C-peptide, which would indicate endogenous origin, the rise in insulin can only be caused by a spillover of spray into the bloodstream. This spillover has been detected previously after 160 U of nasal insulin and corresponds to an intravenous insulin dose of ∼2.5 mU/kg body weight (12) or an absolute dose of 0.1 U (13) for this nasal insulin dose. Furthermore, the spillover seems to be different between human insulin, as used in our current study, and rapid-acting insulin analogs. Although application of 40 U human insulin caused only minor changes in plasma insulin, 40 U of the rapid-acting insulin lispro introduced marked rises in circulating levels in a recent study (28). Furthermore, the kinetics of intranasal human insulin and the insulin analog lispro seem to be quite different. A delayed absorption of insulin lispro reached peak levels in the blood at least 15 minutes later than human insulin (28). By this time, plasma insulin levels were already back to baseline after human insulin spray. The magnitude of the insulin spillover into circulation was not related to any of the detected brain effects in our study. Hence, penetration of the peptide directly into the brain and not the transport via the bloodstream seems to be the major mode of action of nasal insulin in the CNS. Although not routinely done in the past (15, 20, 29), our current results on insulin spillover indicate that this phenomenon should be mimicked by intravenous insulin application when studying peripheral tissues. Of note, we did so in one recent experiment: In this study, insulin delivery to the brain via insulin nasal spray improved peripheral insulin sensitivity by suppressing endogenous glucose production and stimulating glucose uptake into peripheral tissues independent of insulin spillover (12). The insulin-induced change in HRV substantiates previous results on the role of the autonomous nervous system (20). Intranasal insulin specifically induced change in the high-frequency band, which represents mainly the parasympathetic branch of the autonomous nervous system. This is well in line with animal data showing that central insulin action is transmitted via the major parasympathetic nerve (i.e., the vagus nerve) to peripheral organs (30). Just as in animals, the hypothalamus seems to contribute to this response, as it plays a pivotal role for homeostatic regulation and integrating metabolic signals. Insulin reactivity of the hypothalamus has been shown to be compromised in obese individuals (6, 21, 27). Previous results in humans indicated that parasympathetic outflows from the hypothalamus contribute to the modulation of peripheral insulin sensitivity (20). For other aspects of peripheral metabolism (e.g., lipolysis or liver metabolism), this has not been investigated yet. Our current results indicate that future studies aiming to monitor the autonomous nervous system should apply higher doses of insulin nasal spray, as we only detected notable effects after 160 U. Another possibility of brain-derived modulation of peripheral metabolism is endocrine signals. Potential mechanisms involve the hypothalamic–pituitary–adrenal axis, the hypothalamic–pituitary–gonadal axis, or modulation of thyroid function. Just as previously reported in a larger group for cortisol (10), we detected no effects of nasal insulin on the hypothalamic–pituitary–adrenal axis. Furthermore, the pituitary-gonadal and thyroid axes were unaffected by nasal insulin. However, we cannot exclude that chronic nasal insulin administration may have effect on endocrine functions (especially gonadal), as brain-specific knockout of the insulin receptor impaired LH regulation and thereby reproductive function (31) in rodents. As we did not detect acute effects of nasal insulin on the assessed endocrine functions, it is most likely that the autonomous nervous system is the key transducer for acute peripheral effects. Besides the hypothalamus, a dose-dependent effect of insulin was found in regions recently identified as insulin sensitive with strong relevance for food intake and body weight regulation. The amygdala and lateral PFC showed a marked increase for 160 U of insulin, whereas the caudate nucleus (striatal region) revealed a noteworthy decrease in resting-state activity. The hypothalamus plays a vital role in whole-body energy homeostasis. In animal studies, insulin signaling is similar in the amygdala and the hypothalamus, including obesity-related dysregulation (32). Furthermore, both are embedded in the striato-prefrontal circuitry (33, 34), which is particularly sensitive to increasing peripheral and central insulin levels (for a recent review, see 1, 9, 21). Furthermore, as seen in the current study, intranasal insulin influences neural activity in the striatum and PFC (6, 10, 20, 35–37). Moreover, animal models have shown multiple interactions between homeostatic and reward brain circuits (38), in particular that insulin can amplify dopamine release in the striatum (39). Hence, the marked insulin-induced hypothalamic decrease can potentially increase satiety, whereas the decrease in caudate activity and increase in amygdala and PFC activity to insulin may attenuate the rewarding properties of food and motivation for food consumption. Taken together, we detected dose-dependent effects of intranasal insulin on regional brain activity and parasympathetic tone. Although no acute effects of 40 U of nasal insulin were observed, 160 U of the peptide had the strongest effects. However, this dose was accompanied by a transient increase in circulating insulin concentrations due to a spillover into circulation. This is no major obstacle for the assessment of regional brain effects by imaging but should be mimicked in studies on peripheral metabolism. Taking this into account, intranasal insulin application is currently the best available tool to dissect central from peripheral insulin effects. Our current results can be the basis for the design of future studies with nasal insulin administration to disentangle brain insulin effects in health and disease. Abbreviations: ANOVA analysis of variance AUC area under the curve BMI body mass index BOLD blood oxygenation level–dependent CBF cerebral blood flow CSF cerebrospinal fluid ECG electrocardiogram EPI echo planar imaging fALFF fractional amplitude of low-frequency fluctuation fMRI functional magnetic resonance imaging FOV field of view HOMA-IR homeostatic model assessment of insulin resistance HRV heart rate variability LH luteinizing hormone MRI magnetic resonance imaging PFC prefrontal cortex PFWE P family-wise error rsfMRI resting-state functional magnetic resonance imaging SPM12 Statistical Parametric Mapping 12. Acknowledgments We thank all study participants for cooperation on this project and acknowledge the excellent technical assistance of Dr. Dorothea Baumann, Maike Borutta, Anja Dessecker, Dr. Louise Fritsche, Christoph Gassenmaier, Ellen Kollmar, Dr. Steffen Reichert, and Andreas Vosseler (all from University of Tübingen, Tübingen, Germany). Financial Support: This study was supported in part by Grant 01GI0925 from the Federal Ministry of Education and Research to the German Center for Diabetes Research, Helmholtz Alliance ICEMED: Imaging and Curing Environmental Metabolic Diseases; a European Association for the Study of Diabetes Rising Star award supported by Novo Nordisk (to M.H.); and by the European Union Seventh Framework Programme (FP7/2007–2013) under Grant Agreement 607310 (Nudge-it). Disclosure Summary: The authors have nothing to disclose. References 1. Kullmann S, Heni M, Hallschmid M, Fritsche A, Preissl H, Häring HU. Brain Insulin Resistance at the Crossroads of Metabolic and Cognitive Disorders in Humans. Physiol Rev . 2016; 96( 4): 1169– 1209. Google Scholar CrossRef Search ADS PubMed  2. Heni M, Kullmann S, Preissl H, Fritsche A, Häring HU. Impaired insulin action in the human brain: causes and metabolic consequences. Nat Rev Endocrinol . 2015; 11( 12): 701– 711. Google Scholar CrossRef Search ADS PubMed  3. Kleinridders A, Ferris HA, Cai W, Kahn CR. Insulin action in brain regulates systemic metabolism and brain function. Diabetes . 2014; 63( 7): 2232– 2243. Google Scholar CrossRef Search ADS PubMed  4. Lochhead JJ, Thorne RG. Intranasal delivery of biologics to the central nervous system. Adv Drug Deliv Rev . 2012; 64( 7): 614– 628. Google Scholar CrossRef Search ADS PubMed  5. Born J, Lange T, Kern W, McGregor GP, Bickel U, Fehm HL. Sniffing neuropeptides: a transnasal approach to the human brain. Nat Neurosci . 2002; 5( 6): 514– 516. Google Scholar CrossRef Search ADS PubMed  6. Kullmann S, Heni M, Veit R, Scheffler K, Machann J, Häring HU, Fritsche A, Preissl H. Selective insulin resistance in homeostatic and cognitive control brain areas in overweight and obese adults. Diabetes Care . 2015; 38( 6): 1044– 1050. Google Scholar CrossRef Search ADS PubMed  7. Benedict C, Kern W, Schultes B, Born J, Hallschmid M. Differential sensitivity of men and women to anorexigenic and memory-improving effects of intranasal insulin. J Clin Endocrinol Metab . 2008; 93( 4): 1339– 1344. Google Scholar CrossRef Search ADS PubMed  8. Reger MA, Watson GS, Frey WH II, Baker LD, Cholerton B, Keeling ML, Belongia DA, Fishel MA, Plymate SR, Schellenberg GD, Cherrier MM, Craft S. Effects of intranasal insulin on cognition in memory-impaired older adults: modulation by APOE genotype. Neurobiol Aging . 2006; 27( 3): 451– 458. Google Scholar CrossRef Search ADS PubMed  9. Kullmann S, Heni M, Fritsche A, Preissl H. Insulin action in the human brain: evidence from neuroimaging studies. J Neuroendocrinol . 2015; 27( 6): 419– 423. Google Scholar CrossRef Search ADS PubMed  10. Heni M, Kullmann S, Ketterer C, Guthoff M, Linder K, Wagner R, Stingl KT, Veit R, Staiger H, Häring HU, Preissl H, Fritsche A. Nasal insulin changes peripheral insulin sensitivity simultaneously with altered activity in homeostatic and reward-related human brain regions. Diabetologia . 2012; 55( 6): 1773– 1782. Google Scholar CrossRef Search ADS PubMed  11. Dash S, Xiao C, Morgantini C, Koulajian K, Lewis GF. Is insulin action in the brain relevant in regulating blood glucose in humans? J Clin Endocrinol Metab . 2015; 100( 7): 2525– 2531. Google Scholar CrossRef Search ADS PubMed  12. Heni M, Wagner R, Kullmann S, Gancheva S, Roden M, Peter A, Stefan N, Preissl H, Häring HU, Fritsche A. Hypothalamic and striatal insulin action suppresses endogenous glucose production and may stimulate glucose uptake during hyperinsulinemia in lean but not in overweight men. Diabetes . 2017; 66( 7): 1797– 1806. Google Scholar CrossRef Search ADS PubMed  13. Gancheva S, Koliaki C, Bierwagen A, Nowotny P, Heni M, Fritsche A, Häring HU, Szendroedi J, Roden M. Effects of intranasal insulin on hepatic fat accumulation and energy metabolism in humans. Diabetes . 2015; 64( 6): 1966– 1975. Google Scholar CrossRef Search ADS PubMed  14. Craft S, Baker LD, Montine TJ, Minoshima S, Watson GS, Claxton A, Arbuckle M, Callaghan M, Tsai E, Plymate SR, Green PS, Leverenz J, Cross D, Gerton B. Intranasal insulin therapy for Alzheimer disease and amnestic mild cognitive impairment: a pilot clinical trial. Arch Neurol . 2012; 69( 1): 29– 38. Google Scholar CrossRef Search ADS PubMed  15. Benedict C, Brede S, Schiöth HB, Lehnert H, Schultes B, Born J, Hallschmid M. Intranasal insulin enhances postprandial thermogenesis and lowers postprandial serum insulin levels in healthy men. Diabetes . 2011; 60( 1): 114– 118. Google Scholar CrossRef Search ADS PubMed  16. Benedict C, Hallschmid M, Hatke A, Schultes B, Fehm HL, Born J, Kern W. Intranasal insulin improves memory in humans. Psychoneuroendocrinology . 2004; 29( 10): 1326– 1334. Google Scholar CrossRef Search ADS PubMed  17. Benedict C, Hallschmid M, Schultes B, Born J, Kern W. Intranasal insulin to improve memory function in humans. Neuroendocrinology . 2007; 86( 2): 136– 142. Google Scholar CrossRef Search ADS PubMed  18. Benedict L, Nelson CA, Schunk E, Sullwold K, Seaquist ER. Effect of insulin on the brain activity obtained during visual and memory tasks in healthy human subjects. Neuroendocrinology . 2006; 83( 1): 20– 26. Google Scholar CrossRef Search ADS PubMed  19. Kullmann S, Frank S, Heni M, Ketterer C, Veit R, Häring HU, Fritsche A, Preissl H. Intranasal insulin modulates intrinsic reward and prefrontal circuitry of the human brain in lean women. Neuroendocrinology . 2013; 97( 2): 176– 182. Google Scholar CrossRef Search ADS PubMed  20. Heni M, Wagner R, Kullmann S, Veit R, Mat Husin H, Linder K, Benkendorff C, Peter A, Stefan N, Häring HU, Preissl H, Fritsche A. Central insulin administration improves whole-body insulin sensitivity via hypothalamus and parasympathetic outputs in men. Diabetes . 2014; 63( 12): 4083– 4088. Google Scholar CrossRef Search ADS PubMed  21. Kullmann S, Heni M, Veit R, Scheffler K, Machann J, Häring HU, Fritsche A, Preissl H. Intranasal insulin enhances brain functional connectivity mediating the relationship between adiposity and subjective feeling of hunger. Sci Rep . 2017; 7( 1): 1627. Google Scholar CrossRef Search ADS PubMed  22. Pohmann R, Budde J, Auerbach EJ, Adriany G, Uğurbil K. Theoretical and experimental evaluation of continuous arterial spin labeling techniques. Magn Reson Med . 2010; 63( 2): 438– 446. Google Scholar CrossRef Search ADS PubMed  23. Alsop DC, Detre JA, Golay X, Günther M, Hendrikse J, Hernandez-Garcia L, Lu H, MacIntosh BJ, Parkes LM, Smits M, van Osch MJ, Wang DJ, Wong EC, Zaharchuk G. Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: A consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia. Magn Reson Med . 2015; 73( 1): 102– 116. Google Scholar CrossRef Search ADS PubMed  24. Chao-Gan Y, Yu-Feng Z. DPARSF: a MATLAB toolbox for “pipeline” data analysis of resting-state fMRI. Front Syst Neurosci . 2010; 4: 13. Google Scholar PubMed  25. Song XW, Dong ZY, Long XY, Li SF, Zuo XN, Zhu CZ, He Y, Yan CG, Zang YF. REST: a toolkit for resting-state functional magnetic resonance imaging data processing. PLoS One . 2011; 6( 9): e25031. Google Scholar CrossRef Search ADS PubMed  26. Zou QH, Zhu CZ, Yang Y, Zuo XN, Long XY, Cao QJ, Wang YF, Zang YF. An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: fractional ALFF. J Neurosci Methods . 2008; 172( 1): 137– 141. Google Scholar CrossRef Search ADS PubMed  27. Kullmann S, Fritsche A, Wagner R, Schwab S, Häring HU, Preissl H, Heni M. Hypothalamic insulin responsiveness is associated with pancreatic insulin secretion in humans. Physiol Behav . 2017; 176: 134– 138. Google Scholar CrossRef Search ADS PubMed  28. Dash S, Xiao C, Morgantini C, Koulajian K, Lewis GF. Intranasal insulin suppresses endogenous glucose production in humans compared with placebo in the presence of similar venous insulin concentrations. Diabetes . 2015; 64( 3): 766– 774. Google Scholar CrossRef Search ADS PubMed  29. Iwen KA, Scherer T, Heni M, Sayk F, Wellnitz T, Machleidt F, Preissl H, Häring HU, Fritsche A, Lehnert H, Buettner C, Hallschmid M. Intranasal insulin suppresses systemic but not subcutaneous lipolysis in healthy humans. J Clin Endocrinol Metab . 2014; 99( 2): E246– E251. Google Scholar CrossRef Search ADS PubMed  30. Pocai A, Lam TK, Gutierrez-Juarez R, Obici S, Schwartz GJ, Bryan J, Aguilar-Bryan L, Rossetti L. Hypothalamic K(ATP) channels control hepatic glucose production. Nature . 2005; 434( 7036): 1026– 1031. Google Scholar CrossRef Search ADS PubMed  31. Brüning JC, Gautam D, Burks DJ, Gillette J, Schubert M, Orban PC, Klein R, Krone W, Müller-Wieland D, Kahn CR. Role of brain insulin receptor in control of body weight and reproduction. Science . 2000; 289( 5487): 2122– 2125. Google Scholar CrossRef Search ADS PubMed  32. Areias MF, Prada PO. Mechanisms of insulin resistance in the amygdala: influences on food intake. Behav Brain Res . 2015; 282: 209– 217. Google Scholar CrossRef Search ADS PubMed  33. Kullmann S, Heni M, Linder K, Zipfel S, Häring HU, Veit R, Fritsche A, Preissl H. Resting-state functional connectivity of the human hypothalamus. Hum Brain Mapp . 2014; 35( 12): 6088– 6096. Google Scholar CrossRef Search ADS PubMed  34. Roy AK, Shehzad Z, Margulies DS, Kelly AM, Uddin LQ, Gotimer K, Biswal BB, Castellanos FX, Milham MP. Functional connectivity of the human amygdala using resting state fMRI. Neuroimage . 2009; 45( 2): 614– 626. Google Scholar CrossRef Search ADS PubMed  35. Stingl KT, Kullmann S, Guthoff M, Heni M, Fritsche A, Preissl H. Insulin modulation of magnetoencephalographic resting state dynamics in lean and obese subjects. Front Syst Neurosci . 2010; 4: 157. Google Scholar CrossRef Search ADS PubMed  36. Guthoff M, Grichisch Y, Canova C, Tschritter O, Veit R, Hallschmid M, Häring HU, Preissl H, Hennige AM, Fritsche A. Insulin modulates food-related activity in the central nervous system. J Clin Endocrinol Metab . 2010; 95( 2): 748– 755. Google Scholar CrossRef Search ADS PubMed  37. Heni M, Kullmann S, Ahlqvist E, Wagner R, Machicao F, Staiger H, Häring HU, Almgren P, Groop LC, Small DM, Fritsche A, Preissl H. Interaction between the obesity-risk gene FTO and the dopamine D2 receptor gene ANKK1/TaqIA on insulin sensitivity. Diabetologia . 2016; 59( 12): 2622– 2631. Google Scholar CrossRef Search ADS PubMed  38. Ferrario CR, Labouèbe G, Liu S, Nieh EH, Routh VH, Xu S, O’Connor EC. Homeostasis meets motivation in the battle to control food intake. J Neurosci . 2016; 36( 45): 11469– 11481. Google Scholar CrossRef Search ADS PubMed  39. Stouffer MA, Woods CA, Patel JC, Lee CR, Witkovsky P, Bao L, Machold RP, Jones KT, de Vaca SC, Reith ME, Carr KD, Rice ME. Insulin enhances striatal dopamine release by activating cholinergic interneurons and thereby signals reward. Nat Commun . 2015; 6: 8543. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Jan 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial