Hypoglycemia unawareness in Type 1 diabetes suppresses brain responses to 1 hypoglycemia 2 3

28 Background: Amongst non-diabetic individuals, mild glucose decrements alter brain activity in 29 regions linked to reward, motivation and executive control. Whether these effects differ in T1DM 30 patients with and without hypoglycemia awareness remains unclear. 31 Methods: 42 individuals (13 healthy control subjects (HC), 16 T1DM individuals with 32 hypoglycemia awareness (T1DM-Aware) and 13 T1DM individuals with hypoglycemia 33 unawareness (T1DM-Unaware)) underwent BOLD fMRI brain imaging during a 2-step 34 hyperinsulinemic euglycemic (90 mg/dl)-hypoglycemic (60 mg/dl) clamp for assessment of 35 neural responses to mild hypoglycemia. 36 Results: Mild hypoglycemia in HC altered activity in the caudate, insula, prefrontal cortex, and 37 angular gyrus, whereas T1DM-Aware subjects showed no caudate and insula changes, but 38 showed altered activation patterns in the prefrontal cortex and angular gyrus. Most strikingly, in 39 direct contrast to HC and T1DM-Aware subjects, T1DM-Unaware subjects failed to show any 40 hypoglycemia-induced changes in brain activity. These findings were also associated with 41 blunted hormonal counterregulatory responses and hypoglycemia symptoms scores during mild 42 hypoglycemia. 43 Conclusion: In T1DM, and in particular T1DM-Unaware patients, there is a progressive blunting 44 of brain responses in cortico-striatal and fronto-parietal neurocircuits in response to mild45 moderate hypoglycemia. These findings have implications for understanding why individuals 46 with impaired hypoglycemia awareness fail to respond appropriately to falling blood glucose 47 levels. 48 Funding: This study was supported in part by grants from the NIH R01DK020495 and P30 49 DK045735 (Sherwin), K23DK109284 (Hwang), K08AA023545 (Seo), the Yale Center for 50 Clinical Investigation supported by the Clinical Translational Science Award (UL1 RR024139). 51 52 53 INTRODUCTION 54 55 Patients with type 1 diabetes mellitus (T1DM) have long been constrained by the 56 adverse effects of insulin-induced hypoglycemia. The Diabetes Control and Complications Trial 57 (DCCT) established the benefits of restoring mean blood glucose to “near normal” levels in 58 patients with T1DM, and while this has produced clear benefits in terms of the microvascular 59 and macrovascular complications of T1DM, for many individuals, the widespread use of 60 intensified insulin therapy has resulted in a much higher rate of severe hypoglycemia (1). 61 Frequent episodes of hypoglycemia can lead to hypoglycemia unawareness, which prevents 62 patients from taking corrective action by eating. Thus, for many T1DM patients the immediate 63 fear of hypoglycemia exceeds the fear of long-term complications (2, 3). 64 In non-diabetic subjects, hypoglycemia is rare because, in response to falling blood 65 glucose levels, an integrated physiologic response is triggered which suppresses endogenous 66 insulin secretion, increases release of counterregulatory hormones, and provokes awareness of 67 hypoglycemia, which act together to rapidly restore euglycemia by stimulating glucose 68 production and food consumption. We have previously reported using the glucose clamp 69 technique together with functional magnetic resonance (fMRI) imaging, visual food cues, and 70 behavioral measures that brain regions involved in stimulating motivation to eat are exquisitely 71 sensitive to small reductions in glucose. In healthy humans, mild reductions in plasma glucose 72 (~68mg/dl) that were not sufficient to increase counterregulatory hormones were sufficient to 73 activate hypothalamic blood flow (4) as well as modulate brain motivation/reward and executive 74 control responses to food cues, which in turn resulted in a greater ‘wanting’ for high-calorie 75 foods (5). 76 In T1DM, this critical hypoglycemia defense system may be interrupted at every level. 77 Loss of endogenous insulin and reliance on peripheral exogenous hormone delivery make rapid 78 insulin reductions impossible. Beta cell destruction is also linked to loss of glucagon responses 79 to hypoglycemia, a defect that develops in nearly all T1DM patients (6, 7). As a result, T1DM 80 patients are particularly vulnerable to impairments in epinephrine release, which commonly 81 follows iatrogenic insulin-induced hypoglycemia (8-10). 82 Frequent episodes of hypoglycemia in T1DM individuals commonly lead to 83 hypoglycemia-associated autonomic failure (HAAF), whereby significantly lower blood glucose 84 levels are required to elicit a counterregulatory hormonal response as well as symptomatic 85 awareness of hypoglycemia (2, 3, 9). Whether loss of hypoglycemia awareness is also 86 accompanied by a failure to activate the drive to eat, which is clinically the most effective way to 87 reverse hypoglycemia, remains unknown. A study using fMRI reported that functional 88 connectivity in brain regions that have been implicated in the control of feeding behavior 89 including the basal ganglia, insula, and prefrontal cortex are altered in individuals with T1DM 90 (11). However, this study did not examine the specific effects of HAAF and hypoglycemia 91 unawareness on brain activity. Another study in a small number of individuals with T1DM who 92 were both aware or unaware of hypoglycemia using 18-F-Fluoro-2-deoxyglucose (FDG) PET 93 scanning suggested that acute hypoglycemia may increase ventral striatum FDG uptake and 94 that a small diminution of this response may have occurred in unaware patients (12). However, 95 FDG uptake may not accurately reflect glucose uptake during hypoglycemia, since acute 96 hypoglycemia (and likely antecedent hypoglycemia) alters the lumped constant used to 97 calculate glucose uptake (13). 98 It is noteworthy that prior studies using fMRI or PET scanning to assess the impact of 99 hypoglycemia on the brain amongst T1DM individuals have utilized very low glycemic targets 100 (typically 50 mg/dl or less). However, from a clinical perspective, by the time a T1DM patient’s 101 blood glucose falls into the low 50 mg/dl range they may already be at significantly higher risk 102 for hypoglycemia associated morbidity and mortality due to the failure to appropriately activate 103 multiple layers of protection against hypoglycemia. Therefore, in this study, we specifically 104 sought to determine how T1DM individuals both with or without hypoglycemia unawareness 105 respond to milder degrees of hypoglycemia in an effort to more effectively distinguish the CNS 106 defects at an earlier time point leading to unawareness in the course of developing moderate107 severe hypoglycemia. To do this, we used fMRI brain scanning in conjunction with a two-step 108 hyperinsulinemic euglycemic-hypoglycemic clamp technique to investigate how regional brain 109 activity, particularly the neurocircuits driving eating behavior, are altered amongst T1DM 110 individuals with hypoglycemia unawareness (T1DM-Unaware) as compared to those patients 111 with preserved awareness (T1DM-Aware) as well as healthy non-diabetic control subjects (HC) 112 during acute mild-moderate hypoglycemia (target ~60 mg/dl). 113 114 RESULTS 115 Participant characteristics 116 Thirteen healthy, non-diabetic individuals, 16 T1DM-Aware individuals as assessed by the 117 Clarke score (14), and 13 T1DM-Unaware individuals participated in this study. Demographic 118 and clinical characteristics are presented in Table 1. Compared to HC individuals, both T1DM119 aware individuals and T1DM-Unaware individuals were similar in age, gender, and education. 120 The T1DM-Unaware individuals were slightly older (P=0.01), had longer duration of disease 121 (P<0.001), and had slightly higher BMI (P=0.003) than the T1DM-Aware group. T1DM-Unaware 122 individuals also had significantly higher self-reported rates of severe hypoglycemic episodes in 123 the preceeding year (P=0.03). Both T1DM-Aware and T1DM-unaware groups were 124 indistinguishable in terms of HbA1C% and there were no differences across all three groups for 125 gender and education as well as measures of disordered eating and cognitive function (Table 126 1). 127 128 Two-step hyperinsulinemic euglycemic-hypoglycemic clamp 129 As seen in Figure 1B, both groups of individuals with T1DM had modestly higher blood 130 glucose levels at the beginning of the study compared to healthy control subjects. However, 131 using repeated measures linear regression analysis and adjusting for age, BMI, and gender, 132 there were no overall differences in plasma glucose levels during the course of the study 133 between T1DM-Aware and T1DM-Unaware subjects (LS mean 5.3 (-4.1, 14.7), P=0.27). 134 Furthermore, beginning at time 25 minutes, there was no significant statistical difference in pair135 wise comparisons of plasma glucose levels between T1DM-Unaware compared to T1DM-Aware 136 (P=0.11) as well as healthy control participants (P=0.14). Notably, during the times of fMRI 137 BOLD data acquisition (euglycemia at times 45-60 minutes and hypoglycemia at times 90-105 138 minutes), plasma glucose levels were virtually identical across all three groups and were at 139 target (mean plasma glucose at euglycemia 93.0±1.9 mg/dl and hypoglycemia 58.9±1.1 mg/dl) 140 (Figure 1B). There were also no differences in plasma insulin levels between the groups (P = 141 0.76) over time. In addition, there were no significant differences across groups in mean glucose 142 infusion rates (GIR) during euglycemia (GIR mg/kg/min, HC 9.5±1.1 vs. T1DM-Aware 8.2±1.4 143 vs. T1DM-Unaware 7.0±0.9, P = 0.35) as well as during hypoglycemia (HC 7.2±0.8 vs. T1DM144 Aware 6.6±0.7 vs. T1DM-Unaware 4.9±0.8, P = 0.12). 145 146 Hormonal and symptomatic responses to hypoglycemia 147 Mean plasma epinephrine, norepinephrine, glucagon, and cortisol levels at euglycemia 148 and hypoglycemia are shown in Figure 2. Baseline epinephrine levels were different across 149 groups (ANOVA P=0.04). This difference was primarily driven by the difference between T1DM150 Unaware compared to healthy control participants (P=0.024). T1DM-Unaware and T1DM-Aware 151 patients were not significantly different (P=0.07). Notably, plasma epinephrine levels rose 152 significantly in response to hypoglycemia in all three groups. Healthy controls and T1DM-Aware 153 subjects had a nearly 3-fold increase in epinephrine levels, whereas T1DM-Unaware individuals 154 had a much more modest response, i.e. only a 60-70% increase above euglycemic levels. In 155 contrast, only the healthy controls had a significant increase in plasma glucagon and cortisol 156 during the hypoglycemic phase of the study. No significant changes in plasma norepinephrine 157 were detected in the three groups during this relatively mild hypoglycemic stimulus. 158 While in the scanner and prior to the fMRI BOLD acquisitions (at time 30 min and time 159 75 min), participants were asked to rate their symptoms of hypoglycemia using the Edinburgh 160 hypoglycemia score (15). Both T1DM-Aware and control subjects exhibited a statistically 161 significant increase in symptom response during hypoglycemia, whereas there was no 162 significant change in symptoms in the T1DM-Unaware group (Figure 3). Interestingly, 163 hypoglycemia symptoms were different across groups during hypoglycemia (HC, 23.9 ± 7.0; 164 T1DM-Aware, 35.9 ± 14.2; T1DM-Unaware, 28.4 ± 12.4 (ANOVA P=0.03), with the highest 165 scores amongst T1DM-Aware individuals, which was due to differences between healthy control 166 and T1DM-Aware individuals (P=0.009). Furthermore, during hypoglycemia, the symptoms of 167 hypoglycemia correlated significantly and positively with plasma epinephrine levels for the 168 T1DM-Aware individuals (ρ = 0.58, P= 0.02), but not for T1DM-Unaware (P=0.54). Of note was 169 that one participant who was T1DM-Unaware based on Clarke score had a very pronounced 170 increase in plasma epinephrine levels during mild hypoglycemia (euglycemia 47 pg/ml, peak 171 hypoglycemia 329 pg/ml); however, this participant had minimal changes in hypoglycemia 172 symptom scores during acute hypoglycemia (Edinburgh score euglycemia 33, hypoglycemia 39) 173 despite achieving target glucose levels during both the euglycemic and hypoglycemia portions 174 of the clamp. As a result, all fMRI-based analyses were run with and without this participant. 175 Given that there were no significant changes in the results, this participant was included in all 176 subsequent analysis. 177 178 Neural responses to mild hypoglycemia 179 Overall relationship between groups and glycemia (Group X condition effects): Across all 180 three groups, there was a significant group difference in brain response to hypoglycemia and 181 euglycemia in the right striatum (dorsal/ventral), particularly in the caudate (Figure 4) even after 182 adjusting for age and BMI and using the current standard threshold for significance of P<0.001 183 (whole brain, Family-wise error corrected) (16). To give a sense of directionality of change, a 184 region of interest was defined from the significant cluster in the right caudate and mean GLM β185 weights were extracted for each subject. In response to hypoglycemia, healthy control subjects 186 had relatively decreased activity in the caudate, whereas T1DM-Aware and T1DM-Unaware 187 individuals had minimal changes (Figure 4B). Across all three groups, we did not find any 188 significant (at P<0.001) interactions between Group X Glycemia X Task or Group X Task. Thus, 189 all analysis using all 3 groups were collapsed across tasks (visual food and non-food cues). 190 Furthermore, although all three groups had similar plasma glucose levels by 20 minutes prior to 191 the time of BOLD acquisitions, the T1DM-Aware group had higher plasma glucose levels at the 192 start of the clamps. To assess whether these differences in starting glucose levels affected brain 193 activity during euglycemia BOLD acquisitions (~45 minutes later), we assessed across groups 194 and between group interactions at euglycemia alone and found no significant differences. 195 We then examined each group’s brain activity response to hypoglycemia separately 196 (Figure 5 and Supplemental Table 1). The healthy control, T1DM-Aware and T1DM-Unaware 197 subjects had strikingly different patterns of brain responses to mild hypoglycemia, even after 198 adjusting for age and BMI. In particular, in response to hypoglycemia, healthy control subjects 199 had relatively decreased activity in the ventral striatum/caudate, insula, orbitofrontal cortex 200 (OFC), ventromedial prefrontal cortex (vmPFC), ventrolateral prefrontal cortex (vlPFC), 201 dorsolateral prefrontal cortex (dlPFC), and left angular gyrus. In contrast, while the T1DM202 Aware individuals also had relatively decreased activity in the vmPFC and OFC, they did not 203 have any significant differences in activity in the caudate, insula, or dlPFC. Interestingly, the 204 T1DM-Aware individuals had relatively increased activity in the inferior parietal lobe, particularly 205 the right angular gyrus as well as the right ventrolateral prefrontal cortex (vlPFC). In contrast, 206 T1DM-Unaware individuals showed no significant changes in brain activity in any of the regions 207 that were different amongst the other two groups. 208 Given that changes in plasma epinephrine levels are believed to be a particularly 209 sensitive marker for defective counterregulation amongst T1DM individuals, we assessed the 210 relationship between changes in plasma epinephrine levels and changes in brain responses in 211 the regions identified in Figure 5. A smaller change in plasma epinephrine levels was associated 212 with a smaller degree of deactivation in the striatum/caudate (ρ = -0.43, P = 0.005), vmPFC (ρ = 213 -0.46, P = 0.003), right insula (ρ = 0.37, P = 0.02), vlPFC (ρ = 0.34, P = 0.03), and angular 214 gyrus (ρ = -0.41, P = 0.007) consistent with the association of the blunted epinephrine response 215 with the blunted brain responses. There were no associations between brain activity in any of 216 the above regions and epinephrine levels at euglycemia or hypoglycemia alone. Amongst the 217 T1DM subjects, the Edinburgh hypoglycemia symptom score correlated inversely with activity 218 only in the vmPFC during hypoglycemia (ρ = -0.410, P=0.03). 219 220 Effect of hypoglycemia unawareness on brain responses to high calorie food cues 221 To address the question of whether hypoglycemia unawareness modulates the brain’s 222 response to hypoglycemia while viewing high calorie food stimuli, we performed an analysis 223 focused on only the T1DM-Aware and T1DM-Unaware groups. While viewing high calorie food 224 cues (~75% of the high calorie cues were also high carbohydrate), there was a significant 225 Group X Glycemia interaction (P<0.001), even after covarying for age, BMI, and duration of 226 diabetes. This interaction was not present under non-food visual stimuli conditions. Notably, 227 T1DM-Aware individuals had a significant decrease in brain activity during high calorie food in 228 the medial orbitofrontal cortex (Brodmann area 11), while T1DM-Unaware individuals showed 229 no statistically significant change in brain activity in this region (Figure 6). There were no 230 significant correlations between brain activity in this region and counterregulatory hormones. 231 DISCUSSION 232 In this study, we show that hypoglycemia unawareness in T1DM patients is associated 233 with a diminished brain response to mild hypoglycemia (plasma glucose ~60 mg/dl). Moreover, 234 the pattern of loss of brain responses appears to involve cortico-striatal and fronto-parietal 235 neurocircuits that are known to play important roles in regulating motivation and goal-directed 236 behavior as well as attention, and thus are likely to have implications for understanding why 237 individuals with hypoglycemia unawareness fail to respond appropriately to falling blood glucose 238 levels. 239 The basal ganglia, and in particular the caudate, has been consistently shown in studies 240 across species and imaging modalities to play an important role in the ability to respond 241 appropriately to environmental changes and to regulate goal directed behavioral inputs (17-21). 242 The caudate has direct physical and functional connections with executive control regions in the 243 frontal cortex including the medial, ventral, and dorsolateral prefrontal cortex (22, 23). Amongst 244 healthy, non-diabetic individuals, mild hypoglycemia was sufficient to elicit changes in the 245 caudate, cortical regions such as the vmPFC and vlPFC, and the insula, which is consistent with 246 previous studies that have shown that the caudate, PFC, and insula are responsive to changes 247 in circulating glucose levels (5, 12, 24, 25). In contrast, T1DM-Aware individuals had altered 248 patterns of cortico-striatal activity with no significant changes in the caudate or insula during 249 hypoglycemia. We also observed differences across groups in the patterns of 250 activation/deactivation in the dorsolateral prefrontal cortex (dlPFC) and angular gyrus. The 251 angular gyrus, located in the inferior parietal lobe, has direct projections to the dlPFC (26) and 252 together they are part of a larger, well-studied, fronto-parietal circuit (27-29). The angular gyrus, 253 in particular, has been shown to play a role in regulating how one’s attention shifts towards 254 higher salient stimuli (30-33). Interestingly, amongst healthy control subjects, mild hypoglycemia 255 induced changes in activity in the left dlPFC and left angular gyrus, which is consistent with a 256 previous study in healthy control subjects during hypoglycemia (plasma glucose 50 mg/dl) while 257 performing cognitive tasks (34). In contrast, T1DM-Aware individuals had no brain responses in 258 the left dlPFC or left angular gyrus, but instead showed markedly increased activity in the right 259 angular gyrus. Taken together with our findings that T1DM-Aware individuals had higher ratings 260 for symptoms at hypoglycemia, these observations suggest that increased activity in inferior 261 parietal lobe/angular gyrus may be a compensatory adaptation to the disruption in cortico262 striatal and fronto-parietal neurocircuits that are involved in sensing mild hypoglycemia. The 263 markedly increased angular gyrus activity seen in the T1DM-Aware group during mild 264 hypoglycemia may reflect differences in attention to or sensing of the stimulus (35). Thus, the 265 T1DM-Aware individuals may have heightened awareness to hypoglycemia sensory inputs 266 compared to healthy control subjects, which would be consistent with their higher reported 267 ratings of hypoglycemia symptoms both at euglycemia and at hypoglycemia. 268 Most strikingly, compared to T1DM-Aware and healthy control subjects, the T1DM269 Unaware participants showed virtually no changes in brain activity in response to mild 270 hypoglycemia. Very little is known about the impact of hypoglycemia unawareness on regional 271 brain responses; however, these findings would be consistent with the blunted symptom scores 272 as well as the blunted counterregulatory hormone responses to hypoglycemia observed in the 273 T1DM-Unaware group. The underlying mechanism mediating the lack of change amongst the 274 T1DM-Unaware individuals remains uncertain; however, it is likely due to brain adaptations to 275 frequent episodes of severe hypoglycemia in the preceeding year of the study. Recurrent 276 hypoglycemia alters brain glucose transport kinetics as well as promotes increased utilization of 277 alternate fuels such as monocarboxylic acids (lactate, ketones and acetate) in humans when the 278 availability of glucose diminishes (36, 37). Furthermore, T1DM individuals with hypoglycemia 279 unawareness may have alterations in cerebral blood flow during hypoglycemia (38, 39), which 280 may also affect BOLD signal. Interestingly, a recent study has reported that individuals with 281 T1DM and hypoglycemia unawareness have increased cerebral blood flow during acute 282 hypoglycemia compared to T1DM-Aware and healthy controls (39). The current findings would 283 be consistent with these observations that the brain adapts to ensure sufficient substrate 284 (glucose) delivery to the brain. In keeping with these human studies, data in rodents have also 285 demonstrated that prior exposure to hypoglycemia induces up-regulation of blood-brain barrier 286 glucose transport, leading to more efficient glucose utilization during hypoglycemia (40, 41). 287 Thus, the lack of change in brain activity amongst T1DM-Unaware individuals in response to 288 mild hypoglycemia may be the culmination of a variety of adaptive changes in cerebral blood 289 flow, glucose transport, cerebral glucose metabolism, or some combination of each of these 290 factors. 291 It is important to note that induction of hypoglycemia results in a series of dynamic 292 changes in brain activation and deactivation, and thus time intervals when the scans are 293 acquired over the course of hypoglycemia may directly impact the directionality and regional 294 changes observed (24). This, as well as other factors such as hypoglycemia target, timing of 295 image acquisition, imaging modality, may all contribute to the heterogeneity of brain responses 296 to hypoglycemia previously reported in the literature. For example, we did not observe 297 hypoglycemia-induced changes in the hypothalamus, which has been reported by some groups 298 (25), but not others (42) to be altered during hypoglycemia in T1DM individuals. Thus, our 299 findings must be interpreted cautiously given that we are only observing a snapshot of the 300 dynamic brain changes produced over the course of falling blood glucose levels, a critical time 301 for prevention of hypoglycemia-induced brain injury. 302 Importantly, it remains uncertain whether lower glycemic thresholds will be able to elicit 303 changes in brain activation responses amongst T1DM-Unaware individuals and whether the 304 brain responses will be in a similar pattern to that observed amongst T1DM-Aware individuals. 305 Studies of glucose transport kinetics in hypoglycemia unaware T1DM individuals have found 306 that glucose transport is preserved even at glucose levels as low as 50 mg/dl (43). However, it 307 remains uncertain whether lower glucose thresholds are the only difference between T1DM308 Aware and Unaware individuals. Thus, this current study highlights the need for further studies 309 designed to assess the contribution of additional factors such as age of onset of diabetes, 310 duration of diabetes, severity of diabetes during childhood/adolescence when the brain is still 311 developing in determining the propensity for developing hypoglycemia unawareness. 312 Furthermore, whether these changes are reversible and whether strict avoidance of 313 hypoglycemia can restore brain responses remains to be assessed. Of note, prior studies using 314 strict avoidance of hypoglycemia have also resulted in worsening of glycemic control (44-46), 315 which could also have an impact on glucose transport capacity into the brain. 316 Because one of the earliest and best defenses against falling blood glucose levels is to 317 eat, we also examined the brain responses to high calorie/high carbohydrate food cues. 318 Amongst non-diabetic individuals, high calorie food cues have been shown to elicit robust 319 changes in brain activity in reward, motivation, and executive control regions during both 320 euglycemia (47) and mild hypoglycemia (5). Consistent with these findings reported in non321 diabetic individuals, the current data demonstrate that T1DM Hypoglycemia-Aware individuals 322 also had a pronounced change in the medial orbitofrontal cortex (OFC) when viewing high 323 calorie food cues that was not present when looking at pictures of non-food objects. Notably, the 324 medial OFC plays an important role in reward-guided decision-making (48, 49). Furthermore, 325 because it has dense direct connections with the hypothalamus (50, 51), it has been shown to 326 play a particularly important role in regulating feeding behavior (52-54). Thus, it is particularly 327 noteworthy that in contrast to T1DM-Aware individuals, high calorie food cues had no effect on 328 medial OFC brain activity during mild hypoglycemia in T1DM-Unaware individuals, suggesting a 329 diminished drive to eat, which may be a critical early defect in the defense against 330 hypoglycemia. Interestingly, we found no relationship between changes in brain activity to high 331 calorie foods and the counterregulatory hormone response. Whether the lack of brain response 332 is due to intrinsic central nervous system differences or secondary to the blunted rise in 333 circulating counterregulatory hormone levels remains unclear and further studies will be needed 334 to address this question and prove causality. However, given that in non-diabetic subjects 335 changes in brain activity induce and occur prior to changes in counterregulatory hormones (4), it 336 is likely that changes in brain activity are not primarily driven by the counterregulatory response, 337 but rather play the key role in protecting the brain by initiating appropriate defenses against 338 falling glucose levels. Prior studies have also noted a dissociation between cournterregulatory 339 hormone responses and awareness of hypoglycemia (45). 340 It is noteworthy that there are some considerations and limitations to the current study. 341 While we defined our groups using widely accepted and validated questionnaires for 342 hypoglycemia unawareness, the Clarke and Gold scores, these are subjective reports and we 343 did not collect data on glycemic variability and objective rates of hypoglycemia in the months 344 preceeding our studies. In addition, our T1DM-Unaware participants were approximately 10 345 years older and had diabetes for a longer duration than the T1DM Aware group. Although we 346 co-varied for age, BMI, and duration of diabetes, our findings amongst the T1DM-Unaware 347 individuals should still be interpreted cautiously with recognition that it may be very difficult 348 experimentally to separate the effects of age and longer duration of T1DM from the effects of 349 hypoglycemia unawareness itself. Of note in this regard, increasing age has been associated 350 with increases in baseline epinephrine levels (55) and our T1DM-Unaware cohort was slightly 351 older and had higher baseline epinephrine levels; however, we did not observe any 352 relationships between epinephrine levels at euglycemia or hypoglycemia and brain responses. 353 Furthermore, prior studies have examined the effects of age on counterregulatory responses to 354 hypoglycemia (amongst non-diabetic individuals). In these studies, where the mean age of the 355 older groups was markedly older than our cohort (age 60-70’s), they found modest (55) or no 356 (56) differences in counterregulatory responses to hypoglycemia. 357 It is also noteworthy that increased age and duration of diabetes may be associated with 358 cerebrovascular dysfunction. Increased presence of cerebral small vessel disease such as 359 white matter hyperintensities and lacunes have been reported amongst individuals with T1DM 360 (mean age 50 years) (57, 58); however, other studies amongst older T1DM patients (mean age 361 ~60 years and with known microvascular complications) (59) have reported no significant 362 differences in white matter lesions or microinfarcts compared to control subjects. While we 363 cannot exclude the possibility that occult cerebrovascular disease may also contribute to the 364 differences observed in the T1DM-Unaware individuals, this appears less likely given our 365 participants had well-controlled diabetes, had no history of cerebrovascular disease or 366 cardiovascular disease, and were significantly younger (mean age 30 and 40 years for T1DM367 Aware and Unaware, respectively) than the groups reported in the literature. 368 Finally, even though our study includes larger numbers of T1DM-Aware and T1DM369 Unaware participants than prior fMRI-based studies investigating hypoglycemia unawareness, 370 our sample sizes remain a limitation. To minimize the risk of false positives, we used a p-value 371 threshold of P<0.001 (16). Currently, best practice guidelines for conducting fMRI based studies 372 typically recommend at least 20 subjects per group to minimize false positives (60); however, 373 these recommendations may not be directly applicable to studies amongst relatively rare 374 disease groups such as individuals with T1DM and hypoglycemia unawareness or in study 375 designs using highly controlled physiologic manipulations such as in a two step euglycemic376 hypoglycemic clamp where individuals are compared to themselves at two well-defined, but 377 different states. 378 In conclusion, the current study highlights the differential CNS responses to mild 379 hypoglycemia amongst individuals with T1DM and preserved or diminished hypoglycemia 380 awareness. Our findings suggest that although T1DM-Aware individuals no longer exhibit 381 hypoglycemia-induced changes in reward and motivation brain regions (striatum), they have 382 developed compensatory increases in activity in regions associated with attention (i.e. angular 383 gyrus), which may be a protective adaptive mechanism to help maintain an appropriate 384 response to falling glucose levels. However, T1DM patients with hypoglycemia unawareness fail 385 to respond acutely to mild hypoglycemia in cortico-striatal and fronto-parietal brain regions. 386 Taken together with the blunted counterregulatory hormone and subjective hypoglycemia 387 symptom responses seen amongst these individuals, these CNS changes most likely play an 388 important role in causing the inability of T1DM patients with hypoglycemia unawareness to 389 detect and respond appropriately to falling plasma glucose levels. These findings underscore 390 the importance of future interventional studies to determine whether reduction of hypoglycemia 391 frequency can restore these changes in regional brain responses. 392 393 394 METHODS 395 Participants 396 Participants were recruited from the greater New Haven area as well as the Yale 397 Diabetes Center. Inclusion criteria for all subjects included: ages 18-60 years, BMI < 30 kg/m, 398 ability to read English. Inclusion criteria for individuals with T1DM included HbA1C < 9% and 399 duration of diabetes > 5 years. Exclusion criteria included inability to enter the MRI, smoking, 400 illicit drug or recent steroid use; known psychiatric or neurological disorders, active infection, 401 malignancy, abnormal thyroid function, cerebrovascular disease, cardiovascular disease, 402 hepatobiliary disease, weight change in the last 3 months, pregnancy or breastfeeding. 403 Sixty-seven potential subjects were screened at the Yale New Haven Hospital research 404 unit from November 2013 through July 2016 with a screening history, electrocardiogram, 405 physical examination, and laboratory testing at the Yale New Haven Hospital Research Unit 406 (HRU). Of the 67 subjects screened, 42 participants completed the study and were included in 407 the final analysis (see CONSORT diagram showing the flow of participants in the study, 408 Supplement Figure 1). They were divided into the following three groups: 13 healthy controls 409 (HC) (6 males/7 females, age 33±11 years, BMI 24.1±2.9 kg/m2, HbA1c 5.0±0.3%), 16 410 participants with type 1 diabetes and hypoglycemia awareness (T1DM-Aware) (5 males/11 411 females, age 30±8 years, BMI 23.2±3.4 kg/m2, HbA1c 7.0±0.8), and 13 participants with type 1 412 diabetes and hypoglycemia unawareness (T1DM-Unaware) (6 males/7 females, age 40±12 413 years, BMI 26.8±2.9 kg/m2, HbA1c 6.9±0.6). The Clarke score (14) was used to differentiate 414 participants with hypoglycemia awareness versus unawareness. If the Clarke score was not 415 classifiable (i.e. when individuals reported a score of 3 R) then the Gold (61) method was used 416 to determine whether they had impaired hypoglycemia awareness. 417 Study protocol 418 All participants with T1DM were asked to wear a continuous glucose monitor (CGM) 419 (Dexcom G4) one week prior to their scheduled MRI visit in order to monitor for antecedent 420 hypoglycemia. If participants had any episodes of hypoglycemia (glucose < 70 mg/dl or a 421 symptomatic episode requiring assistance) in the 5 days prior to MRI scanning, then the scans 422 were postponed to a later date. On the day of the MRI, participants arrived to the HRU at 9AM. 423 All participants were instructed to eat breakfast as usual prior to arrival and those with diabetes 424 were further instructed to bolus insulin as usual for breakfast. At 10AM, all participants were 425 provided with a standardized snack consisting of 41 grams of carbohydrate (turkey sandwich, 426 apple, diet ginger ale) in order to neutralize feeling of hunger as previously described (5). 427 Participants with diabetes were instructed to inject a bolus of insulin as per their home insulin to 428 carbohydrate ratio. Intravenous catheters were placed in antecubital veins bilaterally: one for 429 blood sampling and the other for insulin and glucose infusion. Participants were informed that 430 their glucose levels would be reduced below normal using an insulin and glucose infusion, 431 which could lead to symptoms of hypoglycemia. Participants were blinded to the timing of 432 changes in glucose levels. Scanning began in the MRI center at 12PM simultaneously with 433 initiation of an insulin infusion at 2 milliunits/kg/hr. Euglycemia (~90 mg/dl) that was maintained 434 for the first phase of the study, after which glucose was decreased into the mild hypoglycemia 435 range (~60 mg/dl) (Figure 1A). Blood oxygen level dependent (BOLD) images were acquired 436 during euglycemia (between time 45-60 minutes) and hypoglycemia (between time 90-105 437 minutes) sessions. Participants completed a visual food task while BOLD images were 438 collected, as described below. Throughout the MRI scan, blood was sampled for glucose every 439 5 minutes. Counter-regulatory hormones (epinephrine, norepinephrine, glucagon and cortisol) 440 were sampled at times 0, 30, 45, 60, 75, 90, and 105. 441 442 Biochemical Analysis: Plasma glucose was measured enzymatically using glucose oxidase 443 (YSI, Yellow Springs, OH). Plasma free insulin, leptin, ghrelin, and glucagon were measured by 444 double antibody radioimmunoassay (Millipore). Double-antibody radioimmunoassay was used to 445 measure plasma cortisol (MP Biomedicals, Solon, OH). Plasma epinephrine and norepinephrine 446 were measured by high-performance liquid chromatography (ESA). 447 448 Visual Food Cue Task: The visual food cue task we used has been previously validated for 449 fMRI (5, 47). During each euglycemia and hypoglycemia session, we presented a total of 42 450 images (three runs of 14 pictures (7 high-calorie food images (HCF), 7 non food images (NF)) 451 each). HCF pictures included items such as hamburgers, pizza, ice cream, and chocolate as 452 previously described (5). 75% of the HCF foods were also high carbohydrate foods. NF pictures 453 consisted of objects such as buildings, books, and doors. Using an event-related design, 454 images were shown for 6 seconds. Each picture was displayed only once and the order of 455 pictures was counterbalanced and randomized within condition across participants. At the end 456 of each trial, a fixation cross appeared with a jittered inter-trial interval (mean, 6 seconds: range 457 3-9 seconds), during which participants relaxed until the onset of the next trial, as previously 458 described (5). This process was repeated for each of the three runs that were presented at both 459 euglycemia and hypoglycemia. 460 461 Hypoglycemia symptom assessments: Participants were asked to verbally rate their 462 sensation of hypoglycemic symptoms (unable to concentrate, blurry vision, anxiety, confusion, 463 difficulty speaking, double vision, drowsiness, tiredness, hunger, weakness, sweating, 464 trembling, warmness, heart racing) on a 7-point Likert scale (1 indicating “not at all” and 7 465 indicating “a lot”) based upon the Edinburgh hypoglycemia symptom score (15) at three 466 separate time points during the study: baseline (prior to entering the scanner) and then once 467 they had reached target glucose levels for euglycemia (at time 30 min) and hypoglycemia (at 468 time 90 min). 469 470 Statistical Analysis: One-way analysis of variance (ANOVA) was used to determine whether 471 there were statistical differences among the three groups for all demographic variables followed 472 by Fisher’s least significant difference (LSD) test for pairwise comparisons if the overall test was 473 statistically significant. Analyses of repeatedly measured variables such as plasma glucose was 474 performed using the mixed-effects regression model method, taking into account both between475 subject and within-subject correlations of repeated measures using a combination of pre476 specified compound symmetry covariance matrix and an autoregressive covariance matrix. Age, 477 gender, and BMI were adjusted as covariates (i.e., as fixed effects). Subsequently, pair-wise 478 comparisons at each time point were performed. Least square mean difference and its 95% 479 confidence interval were reported as a measure of effect size. To assess changes in 480 counterregulatory hormones, plasma hormone levels at euglycemia (times 45 and 60 min) and 481 hormone levels at hypoglycemia (times 90 and 105 min) were averaged together and compared 482 using paired T-tests. All analyses were performed using SAS, version 9.4 (Cary, NC) and 483 SPSS, version 22 (Armonk, NY). A two-tailed p-value of less than 0.05 was considered to be 484 statistically significant. Unless otherwise stated, data are presented as mean +/standard error 485 of mean (SEM). 486 487 Study Approval: The protocol was approved by the Yale University School of Medicine Human 488 Investigation Committee (New Haven, CT). All subjects provided informed, written consent 489 before participation. 490 491 fMRI analysis: The digital data (Digital Imaging and Communication in Medicine (DICOM)) was 492 converted to NIFTI using dcm2nii (62) and then the first 3 images were discarded from each 493 functional run to enable the signal to achieve steady-state equilibrium between radio-frequency 494 pulsing and relaxation leaving 271 images per slice per run for analysis. The data were motion 495 corrected using SPM8 (www.fil.ion.ucl.ac.uk/spm/software/spm8), and they were discarded if 496 linear motion was greater than 1·5 mm or rotation was greater than 2 degrees. Images were 497 iteratively smoothed until the smoothness for any image had a full width half maximum of 498 approximately 6mm (63). For individual subject data analysis, General Linear Model (GLM) was 499 used to determine the regions with changes in signal in response to the visual task (HCF or NF 500 image) in each session. To consider potential variability in baseline fMRI signal, drift correction 501 was included in the GLM with drift regressors used to remove the mean time course, linear, 502 quadratic, and cubic trends for each run. To adjust for anatomical differences in each 503 individual, the Yale Bio-Image Suite software package (http://www.bioimagesuite.org/) was used 504 to calculate two linear and one non-linear registration. These three registrations were 505 concatenated and applied as one registration to bring the data into a common reference brain 506 space. The Colin27 Brain in the Montreal Neurological Institute (MNI) space was used as the 507 reference brain. For group level data analysis, linear effects modeling using AFNI 3dLME 508 (http://afni.nimh.nih.gov) was implemented with a 3 (group: HC, T1DM-Aware, T1DM-Unaware) 509 X 2 (session: Euglycemia, Hypoglycemia) X 2 (Task: HCF and NF) design, while co-varying for 510 age, duration of diabetes, and BMI using the LME modeling program 3dLME from AFNI 511 (http://afni.nimh.nih.gov/sscc/gangc/lme.html). In this design, task and session were treated as 512 the within-subjects fixed-effect factors and group as the between-subjects factor and subject as 513 the random-effect factor. To correct for multiple comparisons, we used family-wise errors (FWE) 514 correction determined by Monte Carlo simulation using the AFNI 3dClustSim version (16.3.05, 515 October 2016) program. Results are shown at p<0.05 whole brain FWE corrected with an initial 516 p threshold of p<0.001, as described previously (16). 517 ACKNOWLEDGEMENTS 518 519 Author contributions: 520 Drs. Hwang and Sherwin had full access to all of the data in the study and take responsibility for 521 the integrity of the data and the accuracy of the data analysis. 522 Study concept and design: Hwang, Seo, Constable, Sinha, Sherwin 523 Acquisition of data: Hwang, Parikh, Seo, Schmidt, Hamza, Lam, Belfort DeAguiar 524 Analysis and interpretation of data: all authors 525 Writing of manuscript: all authors 526 Statistical analysis: Lacadie, Hwang, Dai, Lam, Hamza 527 528 529 Funding and Support: This study was supported in part by grants from the NIH R01DK020495 530 and P30 DK045735 (Sherwin), K23DK109284 (Hwang), K08AA023545 (Seo), the Yale Center 531 for Clinical Investigation supported by the Clinical Translational Science Award (UL1 532 RR024139). 533 534 Role of the sponsors: The funding agencies had no role in the design and conduct of the 535 study; collection, management, analysis, and interpretation of the data; or the preparation, 536 review, or approval of the manuscript. 537 538 We gratefully acknowledge the help of the Yale Core lab staff: Mikhail Smolgovsky, Irene 539 Chernyak, Ralph Jacob, Doreen Nemeth, Maria Batsu, Codruta Todeasa as well as the Yale 540 HRU nurses and staff: Joanne Caprio-Adams, Gina Solomon, Anne O’Connor, Catherine 541 Parmelee, Mary Scanlon, Lynda Knaggs, Carmen Galarza, Elizabeth O’Neal, Joyce Russell, 542 Gayle Pietrogallo, Cynthia Smith. 543 544 545 546 547 1. Group DP. Incidence and trends of childhood Type 1 diabetes worldwide 1990-1999. 548 Diabet Med. 2006;23(8):857-66. 549 2. Cryer PE. Banting Lecture. Hypoglycemia: the limiting factor in the management of 550 IDDM. Diabetes. 1994;43(11):1378-89. 551 3. Sherwin RS. Bringing light to the dark side of insulin: a journey across the blood-brain 552 barrier. Diabetes. 2008;57(9):2259-68. 553 4. Page KA, et al. Small decrements in systemic glucose provoke increases in 554 hypothalamic blood flow prior to the release of counterregulatory hormones. Diabetes. 555 2009;58(2):448-52. 556 5. Page KA, et al. Circulating glucose levels modulate neural control of desire for high557 calorie foods in humans. J Clin Invest. 2011;121(10):4161-9. 558 6. Gerich JE, et al. Lack of glucagon response to hypoglycemia in diabetes: evidence for 559 an intrinsic pancreatic alpha cell defect. Science. 1973;182(4108):171-3. 560 7. Bolli G, et al. Abnormal glucose counterregulation in insulin-dependent diabetes mellitus. 561 Interaction of anti-insulin antibodies and impaired glucagon and epinephrine secretion. 562 Diabetes. 1983;32(2):134-41. 563 8. Hirsch BR, and Shamoon H. Defective epinephrine and growth hormone responses in 564 type I diabetes are stimulus specific. Diabetes. 1987;36(1):20-6. 565 9. Cryer PE, et al. Hypoglycemia. Diabetes Care. 1994;17(7):734-55. 566 10. White NH, et al. Identification of type I diabetic patients at increased risk for 567 hypoglycemia during intensive therapy. N Engl J Med. 1983;308(9):485-91. 568 11. Bolo NR, et al. Functional Connectivity of Insula, Basal Ganglia, and Prefrontal 569 Executive Control Networks during Hypoglycemia in Type 1 Diabetes. J Neurosci. 57