1 00:00:08,400 --> 00:00:11,560 Welcome back to Crack in the Shell on psychological Science. 2 00:00:11,760 --> 00:00:14,920 I'm your host, Mike Doherty, and today I'm speaking with Doctor 3 00:00:14,920 --> 00:00:17,800 Desi Jones, a professor of psychology here at the 4 00:00:17,800 --> 00:00:20,840 University of Maryland and an expert on the topic of autism. 5 00:00:21,080 --> 00:00:23,720 Today, we're diving into an issue that's been dominating the 6 00:00:23,720 --> 00:00:25,520 headlines for the last couple of weeks. 7 00:00:25,640 --> 00:00:28,480 Does taking acetaminophen, better known as Tylenol, during 8 00:00:28,480 --> 00:00:31,560 pregnancy increase the likelihood of a child developing 9 00:00:31,600 --> 00:00:34,880 autism or ADHD? What do the data actually say, 10 00:00:35,120 --> 00:00:37,920 and are there reasons to be skeptical of claims suggesting a 11 00:00:37,920 --> 00:00:39,320 link? Let's find out. 12 00:00:49,540 --> 00:00:51,060 Good morning and welcome to the show, Desi. 13 00:00:51,620 --> 00:00:53,500 Hi, Mike. Yeah, I'm really excited to be 14 00:00:53,500 --> 00:00:55,020 here and to talk about this with you. 15 00:00:55,220 --> 00:00:57,540 Let's start with the research article in the journal 16 00:00:57,580 --> 00:01:00,580 Environmental Health by Didier Prada and colleagues at the 17 00:01:00,580 --> 00:01:03,980 Harvard School of Public Health. Now, what were the main findings 18 00:01:03,980 --> 00:01:05,720 of this article? Yeah. 19 00:01:05,720 --> 00:01:08,840 So this study, I want to clarify, it's not a meta 20 00:01:08,840 --> 00:01:10,400 analysis. They actually used something 21 00:01:10,400 --> 00:01:13,440 called qualitative synthesis. And what they did was they 22 00:01:13,440 --> 00:01:17,200 looked at a bunch of different studies that kind of examined 23 00:01:17,560 --> 00:01:21,600 Tylenol use and diagnosis for autism, ADHD and other 24 00:01:21,600 --> 00:01:24,920 developmental conditions. So they had 46 studies. 25 00:01:24,920 --> 00:01:27,640 They read through them and basically summarized them and 26 00:01:27,640 --> 00:01:30,680 summarized the findings. And they found that more of the 27 00:01:30,680 --> 00:01:34,200 studies, or most of the studies, tended to report a positive 28 00:01:34,200 --> 00:01:38,520 association between Tylenol use and later diagnosed with like 29 00:01:38,520 --> 00:01:41,240 autism and ADHD. Wasn't a meta analysis, so it 30 00:01:41,240 --> 00:01:43,160 was one of these systematic reviews. 31 00:01:43,520 --> 00:01:46,920 So the art making statistical claims on the basis of of this 32 00:01:46,920 --> 00:01:50,160 synthesis per SE, if they're just summarizing statistical 33 00:01:50,160 --> 00:01:52,440 claims of various other articles. 34 00:01:52,440 --> 00:01:53,880 Right. Yeah, exactly. 35 00:01:53,880 --> 00:01:57,520 So I think that their argument is presented in a very like 36 00:01:57,520 --> 00:01:59,880 causal type of way. But then when you look at their 37 00:01:59,880 --> 00:02:03,480 actual findings, they say like there may be an association. 38 00:02:03,800 --> 00:02:06,440 They make it clear it's not like A cause and effect type of 39 00:02:06,440 --> 00:02:07,640 thing. I think that's something that 40 00:02:07,640 --> 00:02:09,919 people need to be really careful about when they're reading 41 00:02:09,919 --> 00:02:12,400 research like this and interpreting things. 42 00:02:12,440 --> 00:02:14,920 Can you summarize maybe some of the specific claims they're 43 00:02:14,920 --> 00:02:16,880 trying to make? Specifically, they said that 44 00:02:16,880 --> 00:02:20,640 when women took Tylenol in the second and third trimesters of 45 00:02:20,640 --> 00:02:23,920 pregnancy, they were more likely to give birth to a child who 46 00:02:23,920 --> 00:02:27,040 then had a diagnosis of ADHD or autism. 47 00:02:27,040 --> 00:02:29,680 Because this was like a broad range of studies, they couldn't 48 00:02:29,680 --> 00:02:32,000 pull things together. So, you know, some studies had 49 00:02:32,000 --> 00:02:34,880 like actual clinician diagnosis that were reported. 50 00:02:34,880 --> 00:02:37,880 Some had parents doing a checklist about symptoms and 51 00:02:37,880 --> 00:02:40,080 things like that. Likewise, there wasn't actually 52 00:02:40,080 --> 00:02:42,880 a dose response relationship. This wasn't an experimental 53 00:02:42,880 --> 00:02:44,480 study. So they didn't say this woman 54 00:02:44,480 --> 00:02:47,880 took this much Tylenol and like when her kid was diagnosed. 55 00:02:48,080 --> 00:02:51,000 Instead they just had reports of prescriptions that people were 56 00:02:51,000 --> 00:02:55,040 given from registries or self report retrospectively from 57 00:02:55,040 --> 00:02:57,720 parents on how much Tylenol you know, they reported that they 58 00:02:57,720 --> 00:03:00,080 used when they were pregnant. You know, one of the things that 59 00:03:00,080 --> 00:03:02,720 as a methodologist we start to be concerned about are things 60 00:03:02,720 --> 00:03:04,240 like, OK, what was the control condition? 61 00:03:04,280 --> 00:03:06,480 Well, in this case, there is no control condition because 62 00:03:06,480 --> 00:03:08,840 they're just summarizing what are basically mostly 63 00:03:08,840 --> 00:03:11,240 correlational studies. Should we get into some of the 64 00:03:11,240 --> 00:03:14,520 limitations of this article? Then when I'm teaching students 65 00:03:14,520 --> 00:03:17,440 about research, I like to talk about confounder or third 66 00:03:17,440 --> 00:03:20,240 variables or things that could impact this relationship. 67 00:03:20,600 --> 00:03:23,240 A good fun example that's given is sometimes you see a 68 00:03:23,240 --> 00:03:26,560 correlation between like ice cream sales and shark attack. 69 00:03:26,640 --> 00:03:29,800 And it's not that selling more ice cream causes more sharks to 70 00:03:29,800 --> 00:03:32,680 attack people, right? Instead, if you think about it, 71 00:03:32,680 --> 00:03:34,760 in the summer, people are eating more ice cream. 72 00:03:34,760 --> 00:03:36,640 They're also going to be at the beach swimming more. 73 00:03:36,640 --> 00:03:39,320 So they're going to be more likely to eat and buy more ice 74 00:03:39,320 --> 00:03:42,240 cream and also more likely to get attacked by a shark. 75 00:03:42,240 --> 00:03:45,040 When we look at correlational studies with autism and 76 00:03:45,040 --> 00:03:47,880 especially this Tylenol thing, there are a lot of potential 77 00:03:47,880 --> 00:03:50,600 confounders that, you know, the authors of the study kind of 78 00:03:50,600 --> 00:03:52,720 mentioned, but they don't really get into them. 79 00:03:52,720 --> 00:03:55,360 They can't control for them in their analysis, things like 80 00:03:55,360 --> 00:03:57,640 that. So some potential confounders 81 00:03:57,640 --> 00:04:00,120 that could actually be contributing to autism. 82 00:04:00,320 --> 00:04:04,000 The biggest one is fevers. When women are pregnant and they 83 00:04:04,000 --> 00:04:07,080 have a fever, doctors typically prescribe Tylenol for them. 84 00:04:07,080 --> 00:04:09,640 That's like kind of seen as the safest, most effective 85 00:04:09,640 --> 00:04:11,960 treatment. However, fevers during pregnancy 86 00:04:11,960 --> 00:04:15,000 actually are linked to the development of many of these 87 00:04:15,040 --> 00:04:18,800 neurodevelopmental conditions, or they can basically impact the 88 00:04:18,800 --> 00:04:20,920 developing fetus when a mom has a fever. 89 00:04:20,920 --> 00:04:24,000 So if a woman has more fevers during pregnancy, that could be 90 00:04:24,000 --> 00:04:26,160 the reason that her kid, you know, is born with a 91 00:04:26,160 --> 00:04:29,520 neurodevelopmental condition rather than the Tylenol usage. 92 00:04:29,680 --> 00:04:34,040 And so in some perverse way, right, pulling back on Tylenol 93 00:04:34,040 --> 00:04:37,800 could have an adverse effect on the prevalence rates. 94 00:04:37,880 --> 00:04:39,040 Yeah. I mean, if that ends up being, I 95 00:04:39,040 --> 00:04:40,360 mean, there are many confounders, right? 96 00:04:40,360 --> 00:04:41,560 So it's not like that's the only one. 97 00:04:41,680 --> 00:04:44,800 Yeah, there's many confounders. But yeah, I mean, a lot of 98 00:04:44,800 --> 00:04:48,280 research has gone into this found that fevers and maternal 99 00:04:48,280 --> 00:04:51,920 sickness during pregnancy can contribute to differences in the 100 00:04:51,960 --> 00:04:55,360 child when they're born. Another example is migraines. 101 00:04:55,360 --> 00:04:57,960 Women who have more migraines during pregnancy take more 102 00:04:57,960 --> 00:04:59,640 Tylenol. However, something could be 103 00:04:59,640 --> 00:05:02,160 contributing to the migraines. You know, maybe the mom has 104 00:05:02,160 --> 00:05:03,680 health conditions, things like that. 105 00:05:03,680 --> 00:05:04,680 I don't. Know if you're going to get into 106 00:05:04,680 --> 00:05:08,840 this, but the genetic risk. We find that autism is like 107 00:05:08,840 --> 00:05:11,520 pretty heritable, meaning there's a pretty strong genetic 108 00:05:11,520 --> 00:05:14,160 component to autism. People who are autistic often 109 00:05:14,160 --> 00:05:16,600 have family members who are autistic, things like that. 110 00:05:16,800 --> 00:05:19,600 And one really good way to look at heritability, that genetic 111 00:05:19,600 --> 00:05:22,160 component, is to do sibling studies. 112 00:05:22,160 --> 00:05:23,920 So. So what happens when we control 113 00:05:23,920 --> 00:05:25,440 for siblings? Yep. 114 00:05:25,440 --> 00:05:27,880 So there is a big Swedish study that actually did this. 115 00:05:27,880 --> 00:05:32,040 They had like 2 million families in there and like a million of 116 00:05:32,040 --> 00:05:35,720 them included siblings And when they controlled for this sibling 117 00:05:35,720 --> 00:05:39,120 relationship, this affected like relationship between Tylenol 118 00:05:39,120 --> 00:05:42,280 usage and autism diagnosis completely disappeared. 119 00:05:42,280 --> 00:05:44,640 It was no longer there, no longer significant. 120 00:05:44,640 --> 00:05:45,920 So and. So the key component of that, 121 00:05:45,920 --> 00:05:48,040 sorry to interrupt, but yeah, component of that one was right. 122 00:05:48,240 --> 00:05:52,480 If you just looked at the the initial analysis of those two 123 00:05:52,480 --> 00:05:55,720 million families, there appears to be an association, but the 124 00:05:55,760 --> 00:05:58,680 key analysis is one that controls for that genetic 125 00:05:58,680 --> 00:06:01,200 factor, right? And once that is controlled for, 126 00:06:01,520 --> 00:06:05,640 the results go to null, right? Yes, exactly. 127 00:06:05,920 --> 00:06:08,720 And you know, again, thinking about these third variables and 128 00:06:08,720 --> 00:06:12,440 stuff like with genetics, it could be that, you know, autism 129 00:06:12,440 --> 00:06:15,000 is associated with a lot of other Co occurring conditions, 130 00:06:15,000 --> 00:06:18,720 chronic pain, things like that, a moms autistic or maybe she had 131 00:06:18,720 --> 00:06:21,000 some health issues so she's taking more Tylenol when she's 132 00:06:21,000 --> 00:06:23,200 pregnant, things like that. Something else could be 133 00:06:23,200 --> 00:06:26,280 increasing the Tylenol usage that you know isn't actually 134 00:06:26,400 --> 00:06:29,080 related to this autism diagnosis, right? 135 00:06:29,080 --> 00:06:31,400 The key thing is like when we look at sort of gold standard 136 00:06:31,400 --> 00:06:34,440 science, right, the, the stuff that we really feel like we can 137 00:06:34,440 --> 00:06:36,840 trust. So we look at 4 studies, like 138 00:06:36,840 --> 00:06:41,160 the Swedish study that controls for a whole host of confounders. 139 00:06:41,400 --> 00:06:44,120 And in this case, like it really is when we're dealing with 140 00:06:44,120 --> 00:06:47,400 something that has a heritability component that has 141 00:06:47,400 --> 00:06:49,880 to be controlled for. And so when we control for that, 142 00:06:49,880 --> 00:06:53,560 we know what's it, those results go to zero. 143 00:06:53,640 --> 00:06:56,680 And we should probably add with 2 million or 1,000,000 families 144 00:06:56,680 --> 00:07:00,600 including the study, this is a very high statistically powered 145 00:07:00,600 --> 00:07:04,120 study that we would typically need to detect really small 146 00:07:04,120 --> 00:07:07,200 effects and yet no measurable relationship. 147 00:07:07,240 --> 00:07:10,240 How about we go into some of the other issues that you and I both 148 00:07:10,240 --> 00:07:13,040 have about this, this article, I don't know if you want to start. 149 00:07:13,560 --> 00:07:15,800 I have a big one I want to lay out there, but I'll I'll let you 150 00:07:15,800 --> 00:07:17,040 start. Yeah. 151 00:07:17,040 --> 00:07:19,880 So I mean, I think the the one that you're about to say is 152 00:07:20,600 --> 00:07:22,440 similar to what kind of what I'm thinking on. 153 00:07:23,320 --> 00:07:26,800 Yes, exactly. So it's like the same classic 154 00:07:26,800 --> 00:07:29,760 story that you see, you know, back when people were 155 00:07:29,760 --> 00:07:31,840 associating vaccines with autism, there was also a 156 00:07:31,840 --> 00:07:33,640 conflict of interest with that researcher. 157 00:07:33,800 --> 00:07:35,400 So you're talking about the Andrew Wakefield. 158 00:07:35,400 --> 00:07:39,400 So, OK, so this is the Seminole studies that purported, I mean, 159 00:07:39,400 --> 00:07:44,520 use the word purported to show a relationship between vaccines 160 00:07:44,520 --> 00:07:48,000 and the development of autism. It has now, to be clear, been 161 00:07:48,000 --> 00:07:49,600 retracted. There were many problems with 162 00:07:49,600 --> 00:07:51,960 that. There was allegations, I think 163 00:07:51,960 --> 00:07:55,080 substantiated allegations of research misconduct, but there 164 00:07:55,080 --> 00:07:56,400 were some other issues that I was unaware. 165 00:07:56,400 --> 00:08:00,240 Of Yep so this was specifically related to the MMR vaccine, 166 00:08:00,440 --> 00:08:03,440 measles, mumps and rubella. And he was saying that that 167 00:08:03,440 --> 00:08:06,280 combined vaccine was, you know, dangerous and kids shouldn't 168 00:08:06,280 --> 00:08:08,680 take it. However, two big conflicts of 169 00:08:08,680 --> 00:08:11,240 interest weren't disclosed. The first was the that he was 170 00:08:11,240 --> 00:08:13,720 actually trying to patent a rival vaccine. 171 00:08:13,720 --> 00:08:17,600 So he wanted a kind of solo, I think it was measles vaccine. 172 00:08:17,640 --> 00:08:20,240 So it was kind of in his best interest to discredit this 173 00:08:20,240 --> 00:08:22,360 vaccine that was already out there so that people would use 174 00:08:22,360 --> 00:08:24,480 his vaccine. And then the second was he was 175 00:08:24,480 --> 00:08:27,320 being paid by lawyers of families who were suing this 176 00:08:27,320 --> 00:08:30,000 vaccine makers. You know, again, he had a lot to 177 00:08:30,000 --> 00:08:33,320 benefit from, you know, if this vaccine was discredited. 178 00:08:33,559 --> 00:08:35,520 And that brings us right back to this article. 179 00:08:35,960 --> 00:08:41,600 Because of if you look closely at the byline, the 4th author, 180 00:08:42,080 --> 00:08:46,640 arguably the senior author is a consultant for a law firm that 181 00:08:46,640 --> 00:08:50,920 is suing the makers of Tylenol. So again, having a lot to 182 00:08:50,920 --> 00:08:53,960 benefit from potentially, you know, they said like this could 183 00:08:53,960 --> 00:08:55,640 be seen as a conflict of interest. 184 00:08:55,640 --> 00:08:57,440 But you know, I definitely think it's. 185 00:08:57,840 --> 00:08:59,600 Right. And, and let's just be honest 186 00:08:59,600 --> 00:09:02,120 that disclosing a conflict of interest does not make it go 187 00:09:02,120 --> 00:09:05,320 away, especially when there's a substantial financial benefit to 188 00:09:05,320 --> 00:09:06,840 be had. And in this case, at least for 189 00:09:07,000 --> 00:09:09,400 some of the other reporting I saw, he had made already 190 00:09:09,400 --> 00:09:12,640 $150,000 on consulting fees for this law firm. 191 00:09:12,960 --> 00:09:14,840 But there's another issue that we wanted to talk. 192 00:09:14,840 --> 00:09:16,800 About yes. So that's the issue of 193 00:09:16,800 --> 00:09:20,680 publication bias, right? So there is another study that 194 00:09:20,680 --> 00:09:23,680 came out recently and researchers actually did a meta 195 00:09:23,680 --> 00:09:28,040 analysis on all of these studies of Tylenol usage, ADHD and 196 00:09:28,040 --> 00:09:29,560 autism. And when they did the meta 197 00:09:29,560 --> 00:09:31,840 analysis, first of all, they didn't find the significant 198 00:09:31,840 --> 00:09:34,920 relationship between Tylenol usage and autism. 199 00:09:35,040 --> 00:09:39,120 They did find a very small one between ADHD and Tylenol usage. 200 00:09:39,120 --> 00:09:42,040 However, they really emphasized that this effect was so small it 201 00:09:42,040 --> 00:09:44,200 could be explained by publication bias. 202 00:09:44,200 --> 00:09:46,480 I know this is something that's interesting to you and people 203 00:09:46,480 --> 00:09:49,520 who are really interested in open science like us basically 204 00:09:49,800 --> 00:09:52,160 there for a long time in the field, right, there's been this 205 00:09:52,160 --> 00:09:55,320 emphasis where you could only publish significant findings, 206 00:09:55,320 --> 00:09:58,160 meaning that most of the research we have out here is 207 00:09:58,160 --> 00:10:00,800 ones that show a relationship between things. 208 00:10:01,480 --> 00:10:04,160 If people had a paper that showed no significant 209 00:10:04,160 --> 00:10:07,920 relationship between Tylenol and ADHD or autism, maybe it was 210 00:10:07,920 --> 00:10:09,920 less likely to get published because it didn't have that 211 00:10:09,920 --> 00:10:12,840 flashy significant finding. Right, Not only that, but you 212 00:10:12,840 --> 00:10:15,520 know, you know, there is there is a systematic preference or 213 00:10:15,520 --> 00:10:18,400 historically in journals for publishing positive findings, 214 00:10:18,520 --> 00:10:20,480 right? But it's also the case that I 215 00:10:20,480 --> 00:10:23,480 think a lot of authors often times won't even bother to 216 00:10:23,480 --> 00:10:26,120 submit that word or publication. So what that means then 217 00:10:26,120 --> 00:10:29,000 obviously is, is that all of those findings out there, all 218 00:10:29,000 --> 00:10:31,800 the studies out there that have looked for an association 219 00:10:31,840 --> 00:10:35,920 between in our, in this case Tylenol and autism and failed to 220 00:10:35,920 --> 00:10:38,360 find that association went unpublished. 221 00:10:38,480 --> 00:10:40,040 Now we don't know what that looks like. 222 00:10:40,040 --> 00:10:43,640 They are systematically missing from the literature, but it does 223 00:10:43,640 --> 00:10:46,960 raise the question about the veracity of the literature. 224 00:10:46,960 --> 00:10:50,080 This, by the way, applies uniformly across science. 225 00:10:50,080 --> 00:10:51,840 It's not specific to this one domain. 226 00:10:51,960 --> 00:10:55,280 It's just in this case the absence of an effect is quite 227 00:10:55,320 --> 00:10:57,120 important. Yeah, exactly. 228 00:10:57,120 --> 00:10:59,760 And I think that's what the authors of this meta analysis 229 00:10:59,760 --> 00:11:02,240 were really kind of getting at. They said, you know, if even 230 00:11:02,240 --> 00:11:05,680 like one or two studies of a null effect weren't published 231 00:11:05,680 --> 00:11:08,400 because of this publication bias, it could have swung the 232 00:11:08,400 --> 00:11:10,400 results, you know, the other way for this. 233 00:11:10,400 --> 00:11:12,760 That's how small this effect was for ADHD. 234 00:11:12,760 --> 00:11:15,520 So that's a big problem. And I think we really need to be 235 00:11:15,520 --> 00:11:17,800 considerate of it when we're, you know, evaluating the 236 00:11:17,800 --> 00:11:21,000 literature out here, especially when you know, you're looking at 237 00:11:21,000 --> 00:11:23,440 trying to make causal inferences, which you really 238 00:11:23,440 --> 00:11:26,760 shouldn't be doing with without an experimental study anyway. 239 00:11:26,880 --> 00:11:29,040 I want to jump to another topic. We don't have a whole lot more 240 00:11:29,040 --> 00:11:32,320 time here left, but I know you've written a lot on stigma 241 00:11:32,320 --> 00:11:35,800 towards autistic people. How does this study and the 242 00:11:35,800 --> 00:11:39,080 framing of it perpetuate stigma? Yeah, So I think right now 243 00:11:39,080 --> 00:11:42,880 especially where in a time where a lot of this discourse around 244 00:11:42,880 --> 00:11:46,440 autism is driven by this fear narrative almost, right. 245 00:11:46,440 --> 00:11:49,440 So you constantly see these headlines like rising rate of 246 00:11:49,440 --> 00:11:53,240 autism diagnosis, one in 31 children diagnosed, things like 247 00:11:53,240 --> 00:11:55,080 that. And people are really focused on 248 00:11:55,080 --> 00:11:58,000 finding A cause of autism. But the way that they do this is 249 00:11:58,000 --> 00:12:00,600 very pathologizing. It treats autism as something 250 00:12:00,720 --> 00:12:03,360 risky, something very negative and harmful. 251 00:12:03,360 --> 00:12:06,640 The same way you would treat like a disease that people have. 252 00:12:06,640 --> 00:12:08,960 They're trying to prevent it, look at ways to cure it, things 253 00:12:08,960 --> 00:12:11,360 like that. Another issue is it's like very 254 00:12:11,360 --> 00:12:13,400 blame focused. If you're narrowing it down to 255 00:12:13,400 --> 00:12:16,520 something like, well, a mom took too much Tylenol when she was 256 00:12:16,520 --> 00:12:19,600 pregnant, it's kind of making it so that it's the mom's fault, 257 00:12:19,600 --> 00:12:22,960 almost like blaming her for her child being neurodivergent. 258 00:12:23,040 --> 00:12:24,320 And that's really harmful, right? 259 00:12:24,520 --> 00:12:26,840 I think we need to focus on the people at the other end of it, 260 00:12:26,840 --> 00:12:29,320 these families, these actual autistic people who are living 261 00:12:29,320 --> 00:12:31,760 their lives. It doesn't do us any, you know, 262 00:12:31,760 --> 00:12:34,040 good to focus on things that happened in the past. 263 00:12:34,240 --> 00:12:37,000 Instead, I think we should really focus on the autistic 264 00:12:37,000 --> 00:12:39,440 people who are here with us today there and they have a lot 265 00:12:39,440 --> 00:12:42,360 of needs and priorities that aren't being met in research. 266 00:12:42,400 --> 00:12:44,440 It's really hard for people to get a diagnosis. 267 00:12:44,440 --> 00:12:47,800 Even up here in Maryland. There can be like a month long 268 00:12:47,800 --> 00:12:50,760 wait list just to get that diagnosis so that you can get 269 00:12:50,760 --> 00:12:53,800 the care and services you need, whether that's in school or at 270 00:12:53,800 --> 00:12:56,600 work, things like that. I guess my broader message here 271 00:12:56,600 --> 00:13:00,400 is that I think instead of focusing on, you know, these 272 00:13:00,400 --> 00:13:04,160 kind of tenuous causes of autism that aren't really fully 273 00:13:04,160 --> 00:13:06,200 supported by the research, we should really shift the 274 00:13:06,200 --> 00:13:09,200 narrative, talk to autistic people and their families, see 275 00:13:09,200 --> 00:13:11,760 what their priorities are. You know, these community 276 00:13:11,800 --> 00:13:14,040 priorities. I think in all health research 277 00:13:14,040 --> 00:13:16,760 is really important and finding ways to make the lived 278 00:13:16,760 --> 00:13:20,280 experiences of autistic people and their families better and as 279 00:13:20,280 --> 00:13:22,480 good as they can be, right? I feel like we could spend 280 00:13:22,480 --> 00:13:25,160 another two or three hours talking about this this, but 281 00:13:25,480 --> 00:13:27,320 Desi's been wonderful having you on the show. 282 00:13:27,320 --> 00:13:29,320 I'm glad you could come and hopefully we can have you back 283 00:13:29,320 --> 00:13:32,200 for either an update or some other work that you're doing on 284 00:13:32,200 --> 00:13:33,120 this topic. Yeah. 285 00:13:33,120 --> 00:13:34,720 Thank you so much. This has been really great to 286 00:13:34,720 --> 00:13:36,400 talk about. My pleasure, cheers. 287 00:13:36,440 --> 00:13:48,840 Bye bye bye. If you're interested in reading 288 00:13:48,840 --> 00:13:51,520 more about this topic, check out the links in the episode 289 00:13:51,520 --> 00:13:53,440 description. And if you're pregnant and 290 00:13:53,440 --> 00:13:56,160 wondering whether to take Tylenol, the best advice is to 291 00:13:56,160 --> 00:13:58,200 consult your doctor. They'll be able to tailor their 292 00:13:58,200 --> 00:14:00,320 recommendations to your specific situation. 293 00:14:00,520 --> 00:14:02,760 That's it for now, take care and be well.