1
00:00:00,330 --> 00:00:01,705
- [Worker] All set?

2
00:00:01,705 --> 00:00:02,538
- All set.
- The straighter you go.

3
00:00:02,538 --> 00:00:03,740
- [Crowd Member] But I had to
ask if we were going straight.

4
00:00:03,740 --> 00:00:06,630
- [Announcer] I'd like to
introduce, I know you all

5
00:00:06,630 --> 00:00:08,400
know her, Mary Cushman.

6
00:00:08,400 --> 00:00:10,470
So Mary's the university's
distinguished professor

7
00:00:10,470 --> 00:00:13,380
of medicine and pathology
and laboratory medicine

8
00:00:13,380 --> 00:00:14,310
here at LCOM.

9
00:00:15,600 --> 00:00:18,570
She is vice chair for
emerging researchers,

10
00:00:18,570 --> 00:00:20,610
the Department of Medicine,
and is also director

11
00:00:20,610 --> 00:00:23,400
of the Thrombosis and Hemostasis Program

12
00:00:23,400 --> 00:00:24,927
in the Vermont division.

13
00:00:24,927 --> 00:00:26,190
And of course, as you all know,

14
00:00:26,190 --> 00:00:29,883
she's co-director of
VCCBH with Mark Nelson.

15
00:00:30,960 --> 00:00:34,791
Mary completed her MD residency
and fellowship here at UVM

16
00:00:34,791 --> 00:00:37,950
and her Master of Science
and Epidemiology at Harvard,

17
00:00:37,950 --> 00:00:41,223
and she's been continuously
funded now for over 28 years.

18
00:00:42,450 --> 00:00:46,500
Mary's work focuses on positive
factors for heart disease,

19
00:00:46,500 --> 00:00:49,770
stroke, cognitive impairment,
venous thrombosis,

20
00:00:49,770 --> 00:00:52,743
and others using molecular
epidemiology approaches.

21
00:00:53,640 --> 00:00:55,230
She's very prolific.

22
00:00:55,230 --> 00:00:59,040
She's authored over 750
original research articles

23
00:00:59,040 --> 00:01:01,083
and she has an h-index 160.

24
00:01:02,610 --> 00:01:05,310
Mary has won many, many awards,

25
00:01:05,310 --> 00:01:07,920
including the American Heart Association

26
00:01:07,920 --> 00:01:10,770
2018 Population Research Prize

27
00:01:10,770 --> 00:01:14,580
recognizing research
achievement, the AHA 2020 award

28
00:01:14,580 --> 00:01:17,910
Meritorious Achievement for
impact in promoting careers

29
00:01:17,910 --> 00:01:21,990
of early professionals and
women in science and medicine.

30
00:01:21,990 --> 00:01:24,690
She is one of only
three Vermont physicians

31
00:01:24,690 --> 00:01:26,880
selected as a member of the Association

32
00:01:26,880 --> 00:01:30,000
of American Physicians and
today she'll be talking

33
00:01:30,000 --> 00:01:33,780
about her work in stroke
and cognitive impairment

34
00:01:33,780 --> 00:01:37,150
in an ongoing national biracial cohort

35
00:01:39,000 --> 00:01:42,720
funded by NIH that's a REGARDS
study, which is Reasons

36
00:01:42,720 --> 00:01:45,810
for Geographic and Racial
Differences in Stroke.

37
00:01:45,810 --> 00:01:46,760
Mary?
- Thank you.

38
00:01:52,560 --> 00:01:57,510
So I just put up some
objectives for the talk.

39
00:01:57,510 --> 00:02:00,450
We'll start with discussing
risk factors for stroke

40
00:02:00,450 --> 00:02:04,050
and cognitive impairment and
the commonalities in those

41
00:02:04,050 --> 00:02:08,130
and then discuss the role of biomarkers

42
00:02:08,130 --> 00:02:11,580
in determining risk and
studying pathogenesis

43
00:02:11,580 --> 00:02:14,650
of cognitive impairment
and stroke, and then

44
00:02:18,180 --> 00:02:21,480
utilize this framework to discuss research

45
00:02:21,480 --> 00:02:23,580
on biomarkers of brain
health and how this can help

46
00:02:23,580 --> 00:02:27,090
improve brain health in the population.

47
00:02:27,090 --> 00:02:31,080
So I'll be discussing racial
differences in health,

48
00:02:31,080 --> 00:02:34,230
comparing Black Americans
to white Americans,

49
00:02:34,230 --> 00:02:36,720
and reference to race is in the context

50
00:02:36,720 --> 00:02:39,300
of understanding social
determinants of health.

51
00:02:39,300 --> 00:02:42,063
We all know race is not
a biological construct.

52
00:02:43,050 --> 00:02:46,800
There are cases where we
might consider genetic markers

53
00:02:46,800 --> 00:02:49,980
that are more common in
people of African ancestry,

54
00:02:49,980 --> 00:02:51,660
and which in turn are more common

55
00:02:51,660 --> 00:02:54,000
in Black than white people
in the United States,

56
00:02:54,000 --> 00:02:56,910
but race is a different
thing than ancestry.

57
00:02:56,910 --> 00:02:58,910
Just making sure we're on the same page.

58
00:03:00,000 --> 00:03:04,302
So some basic facts to get
everybody also on the same page.

59
00:03:04,302 --> 00:03:05,135
- [Crowd Member] Hey, I was curious,

60
00:03:05,135 --> 00:03:06,930
is that NIH definitions
or does that come from...

61
00:03:06,930 --> 00:03:09,750
Where does that, the epidemiology, this-

62
00:03:09,750 --> 00:03:10,770
- Where does what come from?

63
00:03:10,770 --> 00:03:11,847
- [Crowd Member] Oh, the last slide.

64
00:03:11,847 --> 00:03:14,264
Is that-
- It's just the fact.

65
00:03:15,189 --> 00:03:16,022
It's just the fact.

66
00:03:16,022 --> 00:03:17,010
- Yeah.
- It's our current

67
00:03:17,010 --> 00:03:20,180
understanding of what race
in America, you know...

68
00:03:21,780 --> 00:03:26,013
Race is a social construct,
not a biological one.

69
00:03:28,080 --> 00:03:30,750
- [Crowd Member] So is that common across

70
00:03:30,750 --> 00:03:33,120
all publications, across journals?

71
00:03:33,120 --> 00:03:35,490
- [Crowd Member] It is
among everybody I know, yes.

72
00:03:35,490 --> 00:03:36,360
- Okay.
- And it was

73
00:03:36,360 --> 00:03:37,920
at my journal that I edited.

74
00:03:37,920 --> 00:03:39,938
- I don't publish in this
area so I don't use that

75
00:03:39,938 --> 00:03:42,950
when I describe basic science
with you, so I'm just-

76
00:03:42,950 --> 00:03:44,479
- If you wanna talk about-

77
00:03:44,479 --> 00:03:45,547
- You know, when you do both-

78
00:03:45,547 --> 00:03:47,220
- -the genetic architecture or whatever,

79
00:03:47,220 --> 00:03:52,220
you can talk about ancestry,
which is more complicated

80
00:03:52,440 --> 00:03:55,683
in a way because ancestry is,

81
00:03:58,290 --> 00:04:01,436
most people have multiple
(chuckles) ancestries.

82
00:04:01,436 --> 00:04:02,848
So yeah.

83
00:04:02,848 --> 00:04:03,838
- [Crowd Member] Okay.

84
00:04:03,838 --> 00:04:07,620
- Okay, so stroke is the
fifth leading cause of death

85
00:04:07,620 --> 00:04:10,620
and the leading cause of disability.

86
00:04:10,620 --> 00:04:13,320
Well, I was corrected
recently that it's among

87
00:04:13,320 --> 00:04:15,300
the leading causes of disability

88
00:04:15,300 --> 00:04:19,770
and every 40 seconds someone has a stroke.

89
00:04:19,770 --> 00:04:24,450
So it's a disease that I have to say

90
00:04:24,450 --> 00:04:27,720
working in stroke now for
more than 20 years, kinda hard

91
00:04:27,720 --> 00:04:29,880
to get stroke research
published in mainstream,

92
00:04:29,880 --> 00:04:31,680
like general medical journals.

93
00:04:31,680 --> 00:04:33,540
They're just not as interested
in it for some reason

94
00:04:33,540 --> 00:04:35,733
compared to like birth and cancer.

95
00:04:36,630 --> 00:04:40,980
But it's an important morbid condition,

96
00:04:40,980 --> 00:04:44,850
leads to cognitive problems
and dementia as well itself.

97
00:04:44,850 --> 00:04:49,290
And dementia is also
rising rapidly, mostly due

98
00:04:49,290 --> 00:04:52,800
to the aging of the
population, not an actual rise.

99
00:04:52,800 --> 00:04:57,630
And 10% of people over age 70 are affected

100
00:04:57,630 --> 00:05:00,380
and it's the seventh leading
cause of death in the U.S.

101
00:05:01,740 --> 00:05:04,530
And we have common risk factors
for these two conditions

102
00:05:04,530 --> 00:05:08,310
and they're shown here, the major ones.

103
00:05:08,310 --> 00:05:11,730
So hypertension, diabetes, heart disease,

104
00:05:11,730 --> 00:05:14,700
smoking, atrial fibrillation.

105
00:05:14,700 --> 00:05:17,550
A little weaker relationship
with cognitive impairment

106
00:05:17,550 --> 00:05:20,820
than stroke, but these
are the critical things,

107
00:05:20,820 --> 00:05:23,640
and if none of these would exist,

108
00:05:23,640 --> 00:05:26,373
we would have much lower
rates of these diseases.

109
00:05:30,840 --> 00:05:34,890
We have persistent strong differences,

110
00:05:34,890 --> 00:05:38,310
large differences by race in stroke death

111
00:05:38,310 --> 00:05:39,630
in the United States.

112
00:05:39,630 --> 00:05:42,840
So stroke death is
twice as common in Black

113
00:05:42,840 --> 00:05:47,840
than white Americans, and
you can see on the graph

114
00:05:48,048 --> 00:05:53,048
from 2000 to 2020, the
non-Hispanic Black group

115
00:05:53,160 --> 00:05:57,252
is at the top in blue,
and the overall line

116
00:05:57,252 --> 00:06:00,300
is the Black line, and then white people

117
00:06:00,300 --> 00:06:02,372
pretty much overlap with the overall line.

118
00:06:02,372 --> 00:06:05,906
And then the other race and ethnic groups

119
00:06:05,906 --> 00:06:10,800
in the U.S. have lower death
rates than white individuals.

120
00:06:10,800 --> 00:06:13,604
So this is a disparity.

121
00:06:13,604 --> 00:06:17,115
You look at the line,
that has not changed,

122
00:06:17,115 --> 00:06:19,953
and it's actually starting to worsen.

123
00:06:21,270 --> 00:06:24,240
So why would it be worsening?

124
00:06:24,240 --> 00:06:26,550
Why aren't Black individuals benefiting

125
00:06:26,550 --> 00:06:29,700
from this decline over 20 years

126
00:06:29,700 --> 00:06:32,400
that's largely due to
blood pressure control,

127
00:06:32,400 --> 00:06:34,050
risk factor control?

128
00:06:34,050 --> 00:06:35,583
Why is this happening?

129
00:06:36,780 --> 00:06:37,860
I'll get to that at the end of the talk.

130
00:06:37,860 --> 00:06:39,843
We don't really know.

131
00:06:41,010 --> 00:06:43,020
Dementia is also 40% more common

132
00:06:43,020 --> 00:06:44,733
in Black than white individuals.

133
00:06:45,780 --> 00:06:49,443
So back around 2000 I get
a call from my friend,

134
00:06:49,443 --> 00:06:51,960
well, soon to become my
friend, George Howard.

135
00:06:51,960 --> 00:06:53,490
We were acquainted.

136
00:06:53,490 --> 00:06:56,370
We met at a meeting and I
was like a middle author

137
00:06:56,370 --> 00:06:58,803
of like 15 people on
a paper that he wrote.

138
00:07:00,815 --> 00:07:03,840
And he said, "Hey, Mary, do you think

139
00:07:03,840 --> 00:07:06,090
if we could get into the
homes of 30,000 people

140
00:07:06,090 --> 00:07:09,000
all around the country,
could you get blood

141
00:07:09,000 --> 00:07:10,950
or do anything with it?"

142
00:07:10,950 --> 00:07:13,740
George is a biostatistician who functions

143
00:07:13,740 --> 00:07:15,870
like also an epidemiologist.

144
00:07:15,870 --> 00:07:18,259
And I said, "Yeah, why not?"

145
00:07:18,259 --> 00:07:22,920
So we proceeded to write a grant

146
00:07:22,920 --> 00:07:24,690
to try to study these racial differences

147
00:07:24,690 --> 00:07:27,312
and also geographic differences in stroke

148
00:07:27,312 --> 00:07:29,550
using a national sample, because George

149
00:07:29,550 --> 00:07:30,960
was really intensively interested

150
00:07:30,960 --> 00:07:33,660
in geographic differences in stroke,

151
00:07:33,660 --> 00:07:36,502
which you can't study if you
have just a few clinic sites

152
00:07:36,502 --> 00:07:38,940
enrolling patients or participants.

153
00:07:38,940 --> 00:07:42,720
You have to blanket the country.

154
00:07:42,720 --> 00:07:45,043
So we got a great score at that time.

155
00:07:45,043 --> 00:07:48,210
A 10 percentile was pretty sure, you know,

156
00:07:48,210 --> 00:07:50,730
for funding, not like now.

157
00:07:50,730 --> 00:07:55,323
But we got a great score
and the study was born.

158
00:07:56,190 --> 00:07:59,463
And now we have multi
PIs, who are shown here.

159
00:08:00,750 --> 00:08:05,750
And the funding actually started in 2002.

160
00:08:06,900 --> 00:08:10,563
Our funding comes from NINDS and NIA.

161
00:08:11,670 --> 00:08:15,270
And we have George Howard, Suzanne Judd,

162
00:08:15,270 --> 00:08:18,151
who is a biostatistician
nutritional epidemiologist,

163
00:08:18,151 --> 00:08:20,460
Jennifer Manley who's
one of the leading people

164
00:08:20,460 --> 00:08:23,823
in the country on cognitive
epidemiology, and myself.

165
00:08:24,840 --> 00:08:27,570
We also have now added to
the group Virginia Howard,

166
00:08:27,570 --> 00:08:29,400
who happens to be Georgia's spouse.

167
00:08:29,400 --> 00:08:34,400
She was here last spring
presenting the Nancy Jenny lecture.

168
00:08:34,860 --> 00:08:36,483
And Ginny is actually gonna receive

169
00:08:36,483 --> 00:08:39,873
the AHA Population
Research Prize on Monday

170
00:08:39,873 --> 00:08:42,843
at the AHA meeting, so
we're very proud of her.

171
00:08:43,805 --> 00:08:46,710
So REGARDS is a cohort study.

172
00:08:46,710 --> 00:08:48,210
What is a cohort study?

173
00:08:48,210 --> 00:08:50,936
It's a study where you
enroll lots of people,

174
00:08:50,936 --> 00:08:54,450
you collect as much information as you can

175
00:08:54,450 --> 00:08:56,910
about those people using validated

176
00:08:56,910 --> 00:08:58,997
instruments and techniques.

177
00:08:58,997 --> 00:09:02,220
You then follow the people
for clinical outcomes.

178
00:09:02,220 --> 00:09:05,250
So our outcomes were stroke
and cognitive function.

179
00:09:05,250 --> 00:09:07,190
The reason we did
cognitive function actually

180
00:09:07,190 --> 00:09:08,670
is not in the grant.

181
00:09:08,670 --> 00:09:10,078
It was because NINDS said

182
00:09:10,078 --> 00:09:11,960
you have to add cognitive testing.

183
00:09:11,960 --> 00:09:14,802
We're not gonna give you any extra money,

184
00:09:14,802 --> 00:09:16,830
but we have to do that too.

185
00:09:16,830 --> 00:09:19,530
And that was very wise on their part.

186
00:09:19,530 --> 00:09:22,050
And we're not cognitive researchers

187
00:09:22,050 --> 00:09:25,380
so we found a few people to
help us develop instruments

188
00:09:25,380 --> 00:09:28,122
we could administer by telephone.

189
00:09:28,122 --> 00:09:32,580
And so the dual kinda outcomes of REGARDS

190
00:09:32,580 --> 00:09:34,263
were born right at the beginning.

191
00:09:36,150 --> 00:09:38,189
Other diseases have been collected,

192
00:09:38,189 --> 00:09:41,640
disease outcomes, through what
we call ancillary studies.

193
00:09:41,640 --> 00:09:46,560
Somebody wants to study heart
failure or heart disease

194
00:09:46,560 --> 00:09:48,595
or venous thrombosis or whatever it is,

195
00:09:48,595 --> 00:09:51,960
they write their own grant
and we use our infrastructure

196
00:09:51,960 --> 00:09:56,670
to get those outcomes for them,
which they have to pay for

197
00:09:56,670 --> 00:09:58,080
with their grant money.

198
00:09:58,080 --> 00:10:00,660
And so ancillary studies
are a key part of REGARDS.

199
00:10:00,660 --> 00:10:03,210
It allows us to do so much more

200
00:10:03,210 --> 00:10:05,253
than what was originally envisioned.

201
00:10:07,500 --> 00:10:08,423
- [Crowd Member] Yeah.

202
00:10:09,719 --> 00:10:12,597
Since 2002 it's been going on
in such a high-impact study

203
00:10:12,597 --> 00:10:15,240
for so many years, like
the Framingham study.

204
00:10:15,240 --> 00:10:17,180
How do you keep the 30,000 participants?

205
00:10:17,180 --> 00:10:19,320
Do they dwindle with time or
do you continually enroll?

206
00:10:19,320 --> 00:10:21,978
Or how do you maintain
a cohort of that size?

207
00:10:21,978 --> 00:10:23,666
- I'll talk about the methods in a minute

208
00:10:23,666 --> 00:10:25,499
so you'll get a sense.

209
00:10:27,400 --> 00:10:28,233
That's a good question.

210
00:10:29,545 --> 00:10:32,550
You know, you have to have
people that have value

211
00:10:32,550 --> 00:10:34,353
in their participation.

212
00:10:36,420 --> 00:10:39,660
But I'll tell you just the
nuts and bolts of how it works.

213
00:10:39,660 --> 00:10:44,660
So in epidemiology, we
work in big teams usually.

214
00:10:47,550 --> 00:10:49,890
Think of us like herds of people

215
00:10:49,890 --> 00:10:52,440
collaborating together as one.

216
00:10:52,440 --> 00:10:54,240
We have a steering
committee for the study.

217
00:10:54,240 --> 00:10:57,540
This is over in Colchester
Research Facility

218
00:10:57,540 --> 00:11:00,186
with a steering committee
meeting that we had here,

219
00:11:00,186 --> 00:11:01,019
and the people sitting along the window

220
00:11:01,019 --> 00:11:02,970
are a bunch of our
trainees, 'cause they tend

221
00:11:02,970 --> 00:11:04,980
to not sit at the table.

222
00:11:04,980 --> 00:11:08,142
But REGARDS actually
has a really small team

223
00:11:08,142 --> 00:11:12,780
compared to other studies like MESA,

224
00:11:12,780 --> 00:11:16,350
or ERIC, or even Framingham.

225
00:11:16,350 --> 00:11:18,783
And so this is us in the early years.

226
00:11:19,665 --> 00:11:21,660
I won't like point myself out,

227
00:11:21,660 --> 00:11:24,183
but quite a different look back then.

228
00:11:25,170 --> 00:11:28,140
And this is down at UAB
on a nice sunny day.

229
00:11:28,140 --> 00:11:30,456
University of Alabama Birmingham

230
00:11:30,456 --> 00:11:33,660
is where the coordinating
center is, where George is.

231
00:11:33,660 --> 00:11:37,410
So the way we did the
recruitment really broke the mold

232
00:11:37,410 --> 00:11:39,090
of how you can do epidemiology research,

233
00:11:39,090 --> 00:11:41,400
'cause remember I said
we couldn't have clinics

234
00:11:41,400 --> 00:11:43,223
'cause we wanted to study geography,

235
00:11:43,223 --> 00:11:47,400
and if you have a clinic,
geography and city

236
00:11:47,400 --> 00:11:48,780
are totally confounded.

237
00:11:48,780 --> 00:11:51,026
You can't like separate them.

238
00:11:51,026 --> 00:11:53,310
You might have like, you're
interested in poor health

239
00:11:53,310 --> 00:11:55,860
in the south, then you can't have just

240
00:11:55,860 --> 00:11:58,230
two cities in the south and three cities

241
00:11:58,230 --> 00:11:59,220
outside the south.

242
00:11:59,220 --> 00:12:00,750
Doesn't work.

243
00:12:00,750 --> 00:12:04,650
So we recruited by telemarketing methods.

244
00:12:04,650 --> 00:12:07,350
We sent people our mailing and followed up

245
00:12:07,350 --> 00:12:11,250
with a phone call, and if they agreed,

246
00:12:11,250 --> 00:12:13,320
then we did a baseline interview.

247
00:12:13,320 --> 00:12:15,780
This was done as a
computer-assisted interview

248
00:12:15,780 --> 00:12:18,720
where just like
telemarketing, the interviewer

249
00:12:18,720 --> 00:12:21,570
is reading a script and
punching in answers.

250
00:12:21,570 --> 00:12:22,680
Some things were recorded.

251
00:12:22,680 --> 00:12:24,830
Some of the cognitive
testing was recorded.

252
00:12:26,400 --> 00:12:29,199
That was followed a few weeks
later by an in-home visit.

253
00:12:29,199 --> 00:12:32,250
So the key thing to get REGARDS to happen

254
00:12:32,250 --> 00:12:36,240
was that we've learned
from David Goff who's now

255
00:12:36,240 --> 00:12:38,703
the chief of heart section of NHLBI.

256
00:12:39,719 --> 00:12:41,752
At the time he was at Wake Forest

257
00:12:41,752 --> 00:12:44,283
and is a cardiovascular epidemiologist.

258
00:12:47,160 --> 00:12:50,359
David realized he needed
some more life insurance,

259
00:12:50,359 --> 00:12:53,910
so he applied for some life
insurance and they called him up

260
00:12:53,910 --> 00:12:55,770
and they said, "We're gonna
send a nurse to your house

261
00:12:55,770 --> 00:12:58,110
and do a physical." (laughs)

262
00:12:58,110 --> 00:13:00,270
And at the time George was at Wake Forest

263
00:13:00,270 --> 00:13:02,730
and David went in and said, "George,

264
00:13:02,730 --> 00:13:04,440
I know how you can do your study!"

265
00:13:04,440 --> 00:13:06,180
He said, "You're gonna hire this company.

266
00:13:06,180 --> 00:13:07,920
They came to my house, they drew my blood,

267
00:13:07,920 --> 00:13:10,890
they did an EKG, they
measured my height and weight,

268
00:13:10,890 --> 00:13:12,270
blah blah blah," right?

269
00:13:12,270 --> 00:13:16,950
So we hired EMSI, Examination
Management Services Inc.

270
00:13:16,950 --> 00:13:19,680
It was their first real research study.

271
00:13:19,680 --> 00:13:20,940
They have 8,000 employees.

272
00:13:20,940 --> 00:13:22,770
They trained 'em all in
human-subject research

273
00:13:22,770 --> 00:13:24,534
to work with us.

274
00:13:24,534 --> 00:13:25,950
It was really a nice partnership.

275
00:13:25,950 --> 00:13:28,410
So they went into the
home and did the visit

276
00:13:28,410 --> 00:13:30,510
to our specifications.

277
00:13:30,510 --> 00:13:32,880
- [Crowd Member] Did you get
baseline cognitive assessments?

278
00:13:32,880 --> 00:13:35,666
- We started shortly after baseline, yeah,

279
00:13:35,666 --> 00:13:38,610
'cause we were adding
it as we were starting.

280
00:13:38,610 --> 00:13:40,137
Yeah.
- So you could follow,

281
00:13:40,137 --> 00:13:41,670
but you got cognitive function early in.

282
00:13:41,670 --> 00:13:43,630
- Yes.
- Because it's not...

283
00:13:43,630 --> 00:13:45,147
Okay.
- Yes, yes.

284
00:13:45,147 --> 00:13:46,763
It was part of the baseline visit,

285
00:13:48,450 --> 00:13:49,680
or near to the baseline visit.

286
00:13:49,680 --> 00:13:50,880
It was all done by phone.

287
00:13:50,880 --> 00:13:52,320
- [Crowd Member] Yeah, it is
a tough thing is the outcome.

288
00:13:52,320 --> 00:13:54,480
We've done that for TBI it's
hard to do cognitive function

289
00:13:54,480 --> 00:13:55,313
as the outcome-
- Oh we know to-

290
00:13:55,313 --> 00:13:57,240
- [Crowd Member] -if you
haven't done it as baseline.

291
00:13:57,240 --> 00:13:58,680
You have to know who's gonna get a stroke

292
00:13:58,680 --> 00:14:00,210
or get injured, know the-

293
00:14:00,210 --> 00:14:01,440
- Yeah.

294
00:14:01,440 --> 00:14:02,660
- [Crowd Member] It's not just the outcome

295
00:14:02,660 --> 00:14:03,660
you get with the delta.
- Precisely.

296
00:14:03,660 --> 00:14:04,493
Yep.
- Gotcha.

297
00:14:04,493 --> 00:14:08,688
- So we are the central
lab, so I'll tell you

298
00:14:08,688 --> 00:14:09,570
a little bit about that.

299
00:14:09,570 --> 00:14:11,580
And then the follow up, this gets

300
00:14:11,580 --> 00:14:13,470
to your point about retention.

301
00:14:13,470 --> 00:14:16,530
Every six months the
participants are called

302
00:14:16,530 --> 00:14:19,620
to find out how they're
doing, to do certain things

303
00:14:19,620 --> 00:14:23,970
like cognitive testing, to
implement ancillary studies

304
00:14:23,970 --> 00:14:27,180
that might wanna know
about certain diseases

305
00:14:27,180 --> 00:14:28,979
or things that are going on with them.

306
00:14:28,979 --> 00:14:31,380
We've done things like
sending them devices

307
00:14:31,380 --> 00:14:32,910
like accelerometers so we can get

308
00:14:32,910 --> 00:14:34,080
physical activity measures.

309
00:14:34,080 --> 00:14:35,130
There's been all kinds of things.

310
00:14:35,130 --> 00:14:38,130
There's I think 170 ancillary
studies in the funding.

311
00:14:38,130 --> 00:14:41,463
Those are people providing
their own money to do it.

312
00:14:43,620 --> 00:14:47,550
And during those follow-up
visits, if a person indicates

313
00:14:47,550 --> 00:14:49,920
they were hospitalized,
we try to determine

314
00:14:49,920 --> 00:14:52,860
if they had a stroke or if
they think it was a stroke.

315
00:14:52,860 --> 00:14:55,468
The medical records are
obtained from all around

316
00:14:55,468 --> 00:14:57,690
the country, every hospital
that you can imagine.

317
00:14:57,690 --> 00:14:59,550
And those are validated by committee

318
00:14:59,550 --> 00:15:03,840
to determine if stroke occurred.

319
00:15:03,840 --> 00:15:06,180
So here's the team for the central lab.

320
00:15:06,180 --> 00:15:08,190
It was quite a process
to get this started.

321
00:15:08,190 --> 00:15:10,410
We had never done
anything quite like this.

322
00:15:10,410 --> 00:15:13,680
We usually work with clinical
sites that are sending us

323
00:15:13,680 --> 00:15:16,350
the samples like once a month
in batches and stuff like that

324
00:15:16,350 --> 00:15:19,230
from all the people they've
seen at their clinical site,

325
00:15:19,230 --> 00:15:22,587
but here we were gonna get
one box for each participant

326
00:15:22,587 --> 00:15:26,700
delivered to us overnight
after the in-home visit.

327
00:15:26,700 --> 00:15:29,820
So the first person I hired
was this guy, Andrew Betsy,

328
00:15:29,820 --> 00:15:33,630
who worked at a chocolate
factory as a systems engineer.

329
00:15:33,630 --> 00:15:37,680
And he helped us figure
out how we would like

330
00:15:37,680 --> 00:15:40,230
deal with this, how we would
set up like an assembly line

331
00:15:40,230 --> 00:15:43,320
and get all the data into data sets

332
00:15:43,320 --> 00:15:45,693
by just scanning the tubes and so forth.

333
00:15:46,680 --> 00:15:47,580
So here's how it works.

334
00:15:47,580 --> 00:15:49,980
You have your in-home
visit, the blood draw,

335
00:15:49,980 --> 00:15:52,320
which is relevant to
what we're interested in.

336
00:15:52,320 --> 00:15:54,333
DHL was the shipper that we used.

337
00:15:55,320 --> 00:15:59,100
EMSI was their biggest
client and EMSI was able

338
00:15:59,100 --> 00:16:00,510
to get us a discount on the shipping,

339
00:16:00,510 --> 00:16:03,810
which saved us some money
on our grant, (laughs)

340
00:16:03,810 --> 00:16:05,373
and the samples would come overnight.

341
00:16:05,373 --> 00:16:06,540
And if you haven't been there,

342
00:16:06,540 --> 00:16:08,310
that's Colchester Research Facility.

343
00:16:08,310 --> 00:16:11,100
If anybody ever wants to come
out, happy to show you around.

344
00:16:11,100 --> 00:16:14,220
It's quite a nice facility for us

345
00:16:14,220 --> 00:16:17,223
and we have a lot of space over there.

346
00:16:20,670 --> 00:16:22,983
We'd get as many as 100 boxes a day,

347
00:16:24,912 --> 00:16:28,200
Tuesday through Friday
only, 'cause we couldn't

348
00:16:28,200 --> 00:16:29,550
deal with weekend.

349
00:16:29,550 --> 00:16:32,250
But this is what they look
like when they come in.

350
00:16:32,250 --> 00:16:34,230
And we've limited that now.

351
00:16:34,230 --> 00:16:36,060
We have another study that's doing this

352
00:16:36,060 --> 00:16:39,510
and we've done REGARDS, another
home visit after baseline.

353
00:16:39,510 --> 00:16:42,710
We pretty much keep it to
no more than 40 in a day,

354
00:16:42,710 --> 00:16:45,270
'cause this takes a long time to handle.

355
00:16:45,270 --> 00:16:50,190
This is what the kit looks
like when we were to open it.

356
00:16:50,190 --> 00:16:52,110
There's paperwork, there's a consent form.

357
00:16:52,110 --> 00:16:54,810
There's some paperwork that of data

358
00:16:54,810 --> 00:16:56,040
that was collected during the visit.

359
00:16:56,040 --> 00:16:58,680
We take those out, we remove the samples,

360
00:16:58,680 --> 00:17:01,320
we re-centrifuge everything at high speed

361
00:17:01,320 --> 00:17:04,380
in refrigerated centrifuge
because the processing

362
00:17:04,380 --> 00:17:06,660
at the site into serum and plasma

363
00:17:06,660 --> 00:17:09,390
is done in kinda just tabletop centrifuges

364
00:17:09,390 --> 00:17:12,900
that the examiners have
often in their homes

365
00:17:12,900 --> 00:17:14,800
or the trunk of their car or whatever.

366
00:17:16,410 --> 00:17:19,650
And then they go into our freezers,

367
00:17:19,650 --> 00:17:22,827
and this is what our
freezer towers look like,

368
00:17:22,827 --> 00:17:25,110
and in each box there's a grid.

369
00:17:25,110 --> 00:17:29,850
And every aliquot that we
have stored is barcoded

370
00:17:29,850 --> 00:17:34,020
and we know exactly where
it is in which tower,

371
00:17:34,020 --> 00:17:37,170
which row of the tower,
which freezer, et cetera

372
00:17:37,170 --> 00:17:40,293
so we can retrieve what
we need when we need it.

373
00:17:41,430 --> 00:17:46,430
So the baseline visit filled
up 24 of these test reserves,

374
00:17:46,440 --> 00:17:50,863
750 liters of serum, plasma, and urine,

375
00:17:50,863 --> 00:17:55,835
whole blood, and half a million aliquots.

376
00:17:55,835 --> 00:17:58,770
We aliquot everything into small sizes

377
00:17:58,770 --> 00:18:00,820
so that we can minimize freestyle effects

378
00:18:02,666 --> 00:18:03,780
on the material.

379
00:18:03,780 --> 00:18:06,960
And that's our warehouse,
which is in Burlington

380
00:18:06,960 --> 00:18:11,640
and currently has upwards
of 200 freezers in it.

381
00:18:11,640 --> 00:18:13,170
Crazy.

382
00:18:13,170 --> 00:18:16,230
And this is where the
cohort lived at the end.

383
00:18:16,230 --> 00:18:19,380
So 30,239 participants.

384
00:18:19,380 --> 00:18:23,370
We targeted age range 45 to 100.

385
00:18:23,370 --> 00:18:25,890
We over-sampled the southeast
because we're interested

386
00:18:25,890 --> 00:18:30,390
in stroke disparities in the southeast

387
00:18:30,390 --> 00:18:32,610
where stroke rates are
higher and death rates

388
00:18:32,610 --> 00:18:34,980
from stroke are higher than
other parts of the country.

389
00:18:34,980 --> 00:18:36,240
There's other pockets of risk,

390
00:18:36,240 --> 00:18:39,780
but the southeast is a hotbed for stroke.

391
00:18:39,780 --> 00:18:41,100
- [Crowd Member] Can I just quickly ask,

392
00:18:41,100 --> 00:18:43,380
does that coincide with
the dementia as well?

393
00:18:43,380 --> 00:18:44,220
- Yes!

394
00:18:44,220 --> 00:18:47,223
We've discovered that,
with cognitive impairment.

395
00:18:49,110 --> 00:18:52,173
And you can see there are
like seven people in Vermont.

396
00:18:53,280 --> 00:18:57,480
Red is white, blue is African American,

397
00:18:57,480 --> 00:19:00,240
and of course the Vermont
participants are all white.

398
00:19:00,240 --> 00:19:01,350
Maria?

399
00:19:01,350 --> 00:19:06,350
- [Maria] Do they specifically
ask about Hispanic?

400
00:19:06,660 --> 00:19:08,679
So is this like white non-Hispanic?

401
00:19:08,679 --> 00:19:09,750
- It's not Hispanic.

402
00:19:09,750 --> 00:19:11,451
- [Maria] So it was
specifically put in here.

403
00:19:11,451 --> 00:19:13,110
- Yes. Yes.

404
00:19:13,110 --> 00:19:14,160
It's a good question.

405
00:19:16,255 --> 00:19:17,970
So that's where, and
you can see the people

406
00:19:17,970 --> 00:19:20,242
are pretty much where people live.

407
00:19:20,242 --> 00:19:21,963
But we have rural.

408
00:19:22,950 --> 00:19:26,000
We cover two thirds of
the counties in the U.S.

409
00:19:26,000 --> 00:19:28,503
We have thousands of rural people,

410
00:19:29,760 --> 00:19:30,780
so there's a lot of opportunity

411
00:19:30,780 --> 00:19:32,403
to study rural health as well.

412
00:19:33,360 --> 00:19:36,510
So all kinds of baseline
data that we have.

413
00:19:36,510 --> 00:19:38,613
I'm not gonna go through all of this.

414
00:19:39,660 --> 00:19:42,450
We actually, the genomics
data, I'm really excited

415
00:19:42,450 --> 00:19:44,670
'cause we never were able
to get our study section

416
00:19:44,670 --> 00:19:47,913
to help us do any genetics research,

417
00:19:48,833 --> 00:19:50,430
'cause a big barrier to genetics research

418
00:19:50,430 --> 00:19:52,950
is the cost of DNA extraction.

419
00:19:52,950 --> 00:19:55,380
It's about $35 a sample.

420
00:19:55,380 --> 00:19:58,200
Anything times 30,000 is a lot of money.

421
00:19:58,200 --> 00:20:00,180
And at the beginning on our first renewal,

422
00:20:00,180 --> 00:20:03,240
we wanted to get DNA
extraction covered (laughs)

423
00:20:03,240 --> 00:20:05,479
and the study section
said, "That's not science.

424
00:20:05,479 --> 00:20:07,200
No way, take it out."

425
00:20:07,200 --> 00:20:09,660
And so when we resubmitted
we had to take it out,

426
00:20:09,660 --> 00:20:13,420
but ultimately we were able to get

427
00:20:14,550 --> 00:20:17,820
DNA available on everyone
thanks to a recent collaboration

428
00:20:17,820 --> 00:20:20,550
with a NIA intramural group.

429
00:20:20,550 --> 00:20:24,540
They're completing GWAS on everyone

430
00:20:24,540 --> 00:20:27,573
that we already have it
on, which is really great.

431
00:20:28,920 --> 00:20:31,828
Really a multimillion dollar effort

432
00:20:31,828 --> 00:20:36,828
because they like us and it's a service

433
00:20:38,340 --> 00:20:40,593
to the research community and to America.

434
00:20:41,700 --> 00:20:46,500
So the follow up, as I
said, every six months.

435
00:20:46,500 --> 00:20:48,780
The cognitive battery, it was developed

436
00:20:48,780 --> 00:20:52,620
right after we started, so
there was a phase-in of it,

437
00:20:52,620 --> 00:20:55,170
but basically you have
close-to-baseline measures

438
00:20:55,170 --> 00:20:57,510
and then every two years there's a battery

439
00:20:57,510 --> 00:21:02,373
consisting of four tests,
and here's all the outcomes.

440
00:21:03,690 --> 00:21:05,850
Somewhat updated numbers, but you can see

441
00:21:05,850 --> 00:21:09,600
how many of everything that we've had,

442
00:21:09,600 --> 00:21:12,210
and these are outcomes among people

443
00:21:12,210 --> 00:21:15,060
who did not have that disease at baseline.

444
00:21:15,060 --> 00:21:20,060
So you see 2,000 strokes,
600 end-stage kidney disease,

445
00:21:20,190 --> 00:21:23,883
over 1,000 cognitive
impairment, dementia death.

446
00:21:25,650 --> 00:21:28,623
Lots of diabetes and hypertension
developing in new people.

447
00:21:31,590 --> 00:21:34,553
- [Crowd Member] So those
numbers drop in hypertension,

448
00:21:34,553 --> 00:21:39,120
is that is similar to the
other overall finding?

449
00:21:39,120 --> 00:21:41,793
- Pretty much. You mean to the country?

450
00:21:43,620 --> 00:21:44,940
- [Crowd Member] Yeah, so
that this is the number

451
00:21:44,940 --> 00:21:48,690
in regard positive parts and only

452
00:21:48,690 --> 00:21:53,310
the hypertension percentage,
they feel a little

453
00:21:53,310 --> 00:21:54,630
lower than that.
- Remember they're 45

454
00:21:54,630 --> 00:21:56,610
and older at baseline.

455
00:21:56,610 --> 00:21:59,400
- But that's-
- So about 60%

456
00:21:59,400 --> 00:22:01,710
of the Black people have
hypertension at baseline.

457
00:22:01,710 --> 00:22:02,550
- [Crowd Member] Oh, okay.

458
00:22:02,550 --> 00:22:04,410
- Right, they already have it.

459
00:22:04,410 --> 00:22:06,390
So you're catching new ones

460
00:22:06,390 --> 00:22:08,760
in largely the middle-aged people.

461
00:22:08,760 --> 00:22:11,010
- Thank you.
- Yeah, it's a good question.

462
00:22:11,010 --> 00:22:13,080
The characteristics of the
cohort are pretty similar

463
00:22:13,080 --> 00:22:15,480
to Black and white
Americans in general though.

464
00:22:15,480 --> 00:22:19,320
We were really proud of that,
'cause that's always an issue.

465
00:22:19,320 --> 00:22:21,270
Like is it generalizable?

466
00:22:21,270 --> 00:22:23,520
Do they represent real people, right?

467
00:22:23,520 --> 00:22:26,430
They're people who, honestly,
they answered the phone,

468
00:22:26,430 --> 00:22:28,200
they let us come into their house,

469
00:22:28,200 --> 00:22:30,883
they let us take their shirt
off to do an EKG (laughs)

470
00:22:30,883 --> 00:22:31,716
and give us their blood.

471
00:22:31,716 --> 00:22:33,270
I mean, would you even talk to a person

472
00:22:33,270 --> 00:22:35,160
who called you asking you to do that?

473
00:22:35,160 --> 00:22:35,993
Like think about it.

474
00:22:35,993 --> 00:22:39,489
They're really amazing
individuals who really did-

475
00:22:39,489 --> 00:22:43,170
- [Crowd Member] Were there
like incentives for them?

476
00:22:43,170 --> 00:22:47,250
- At the baseline visit, I
think we gave them like $25,

477
00:22:47,250 --> 00:22:50,700
and so the UAB got good at
processing these checks.

478
00:22:50,700 --> 00:22:52,590
And occasionally there was a participant

479
00:22:52,590 --> 00:22:55,770
who mailed it back (laughs)
and it was just like,

480
00:22:55,770 --> 00:22:58,951
that created a huge crisis at UAB,

481
00:22:58,951 --> 00:23:00,570
'cause managing the budget
and putting the money back

482
00:23:00,570 --> 00:23:02,610
or whatever was kind of a pain.

483
00:23:02,610 --> 00:23:06,630
But yeah, and I think at the
second visit we gave 'em $100.

484
00:23:06,630 --> 00:23:07,463
Inflation.

485
00:23:10,050 --> 00:23:12,240
The second visit was 10
years after the baseline,

486
00:23:12,240 --> 00:23:14,670
and I'll talk about
that a little bit more.

487
00:23:14,670 --> 00:23:15,903
So what have we learned?

488
00:23:17,430 --> 00:23:19,890
Fundamental information
about the stroke disparity.

489
00:23:19,890 --> 00:23:21,990
We didn't know when we
started if the disparity

490
00:23:21,990 --> 00:23:24,060
in stroke death in the
United States by race

491
00:23:24,060 --> 00:23:27,030
was due to higher incidences of stroke

492
00:23:27,030 --> 00:23:29,520
or higher mortality after stroke, right?

493
00:23:29,520 --> 00:23:32,040
Basic stuff that we didn't know.

494
00:23:32,040 --> 00:23:33,731
And who does it affect?

495
00:23:33,731 --> 00:23:36,180
And it turned out it's
affecting young people.

496
00:23:36,180 --> 00:23:41,180
So you can see here at age 45 to 55,

497
00:23:41,220 --> 00:23:43,140
the excess risk of stroke in Black

498
00:23:43,140 --> 00:23:46,710
compared to white is three times greater,

499
00:23:46,710 --> 00:23:51,710
and over age it gets weaker
such that the oldest group,

500
00:23:51,870 --> 00:23:54,993
75 plus, had no excess risk of stroke.

501
00:23:56,160 --> 00:23:58,503
And so this is really bad.

502
00:24:00,150 --> 00:24:02,400
This just shows the same thing on a graph.

503
00:24:02,400 --> 00:24:05,802
So the blue line is the relative risk

504
00:24:05,802 --> 00:24:09,780
and you can see the
three at the youngest age

505
00:24:09,780 --> 00:24:12,360
going down to no increased risk.

506
00:24:12,360 --> 00:24:13,650
There's one.

507
00:24:13,650 --> 00:24:15,600
The no increased risk at the older age.

508
00:24:15,600 --> 00:24:17,520
The dotted lines are
the confidence intervals

509
00:24:17,520 --> 00:24:19,293
for those estimates.

510
00:24:20,400 --> 00:24:24,420
And I told you that stroke is
a leading cause of disability.

511
00:24:24,420 --> 00:24:27,060
So this is terrible, right?

512
00:24:27,060 --> 00:24:31,083
So this is robbing these people
in the prime of their life

513
00:24:31,083 --> 00:24:35,280
of their ability to function
in the world, right?

514
00:24:35,280 --> 00:24:37,563
So it's really terrible.

515
00:24:38,663 --> 00:24:42,000
There's been estimates of the excess cost,

516
00:24:42,000 --> 00:24:46,200
like a cost analysis, something
like 10 or $20 billion

517
00:24:46,200 --> 00:24:49,143
a year in lost productivity
and economic impact.

518
00:24:50,940 --> 00:24:53,160
When you add stroke risk factors to this

519
00:24:53,160 --> 00:24:56,250
to try to figure out are the
risk factors explaining this?

520
00:24:56,250 --> 00:24:59,570
Does the higher prevalence
of hypertension or diabetes

521
00:24:59,570 --> 00:25:02,190
in Black people explain this relationship?

522
00:25:02,190 --> 00:25:05,520
What you see is some
reduction in that risk.

523
00:25:05,520 --> 00:25:09,450
So about half of the risk is explained

524
00:25:09,450 --> 00:25:12,510
by including risk factors in the analysis.

525
00:25:12,510 --> 00:25:14,670
If you then on top of that add income

526
00:25:14,670 --> 00:25:18,523
and education as proxies
for socioeconomic status,

527
00:25:20,880 --> 00:25:24,813
you see really not much
difference as the risk factors.

528
00:25:26,370 --> 00:25:28,860
So basically about half
of that racial excess

529
00:25:28,860 --> 00:25:31,410
is due to those risk
factors that I showed you

530
00:25:31,410 --> 00:25:32,283
at the beginning.

531
00:25:34,390 --> 00:25:37,200
What other things could there be?

532
00:25:37,200 --> 00:25:42,130
So physical activity, stress, racism,

533
00:25:43,110 --> 00:25:45,060
medication adherence, access to care.

534
00:25:45,060 --> 00:25:48,510
There's all kinds of things
that could be involved.

535
00:25:48,510 --> 00:25:53,510
Nutrition and pretty
much all of this stuff

536
00:25:54,210 --> 00:25:57,540
has been addressed over the
years in different manuscripts

537
00:25:57,540 --> 00:25:58,413
from the study.

538
00:26:01,260 --> 00:26:04,230
Blood pressure is a really key factor.

539
00:26:04,230 --> 00:26:09,230
So this shows the association
of systolic blood pressure,

540
00:26:09,510 --> 00:26:11,460
whether you have hypertension or not,

541
00:26:11,460 --> 00:26:13,200
and the risk of stroke.

542
00:26:13,200 --> 00:26:18,200
And so you see here everyone,
the relative risk of stroke

543
00:26:19,050 --> 00:26:23,610
is about 15% higher for
every 10 millimeters mercury

544
00:26:23,610 --> 00:26:26,313
greater for blood pressure, okay?

545
00:26:28,591 --> 00:26:31,380
In white people the relationship's

546
00:26:31,380 --> 00:26:34,503
a little weaker, 8% greater.

547
00:26:36,235 --> 00:26:38,820
In Black people it's 24% greater.

548
00:26:38,820 --> 00:26:41,880
So the relation of blood
pressure with stroke risk

549
00:26:41,880 --> 00:26:46,290
in Black individuals
is three times greater

550
00:26:46,290 --> 00:26:47,250
than a white individual.

551
00:26:47,250 --> 00:26:49,440
It's a more potent risk factor

552
00:26:49,440 --> 00:26:51,603
for the same incremental blood pressure.

553
00:26:54,720 --> 00:26:56,850
When you add in all the other risk factors

554
00:26:56,850 --> 00:26:58,920
to see if you can explain that away,

555
00:26:58,920 --> 00:27:01,473
doesn't change anything, right?

556
00:27:03,503 --> 00:27:05,280
So this is really revealing, and you know,

557
00:27:05,280 --> 00:27:07,590
so we determined that high blood pressure

558
00:27:07,590 --> 00:27:09,300
is more common in a Black individual.

559
00:27:09,300 --> 00:27:11,340
It's actually more difficult to treat

560
00:27:11,340 --> 00:27:13,413
and it's a more potent risk factor.

561
00:27:15,570 --> 00:27:19,020
Blood pressure was also
related to cognitive decline

562
00:27:19,020 --> 00:27:20,610
and the association was greater

563
00:27:20,610 --> 00:27:23,145
in Black than white individuals.

564
00:27:23,145 --> 00:27:25,560
And we know from clinical trials

565
00:27:25,560 --> 00:27:29,010
that intensive blood pressure control,

566
00:27:29,010 --> 00:27:32,760
systolic blood pressure
down to 120 as the target

567
00:27:32,760 --> 00:27:35,880
is associated with
reductions in risk of stroke

568
00:27:35,880 --> 00:27:38,130
and cognitive impairment, although I think

569
00:27:38,130 --> 00:27:40,560
that's not published
yet, but it was presented

570
00:27:40,560 --> 00:27:42,150
last year at AHA.

571
00:27:42,150 --> 00:27:43,320
- [Worker] Have you made a distinction

572
00:27:43,320 --> 00:27:45,360
between male and female?

573
00:27:45,360 --> 00:27:47,943
- Yeah, we often look for
sex differences as well,

574
00:27:48,990 --> 00:27:51,720
and whole nother topic and talk.

575
00:27:51,720 --> 00:27:53,520
I'm not gonna get into that much,

576
00:27:53,520 --> 00:27:58,410
but I don't think these
differences differ by sex.

577
00:27:58,410 --> 00:27:59,243
Yeah.

578
00:28:01,080 --> 00:28:03,060
It's hard when you cut
it into four groups,

579
00:28:03,060 --> 00:28:06,048
race and sex, between
Black men and Black women

580
00:28:06,048 --> 00:28:07,830
and white men and white
women, but we try to do that

581
00:28:07,830 --> 00:28:10,083
as much as we can 'cause it's important.

582
00:28:12,570 --> 00:28:16,710
Research that occurred
this summer by Nels Olson

583
00:28:16,710 --> 00:28:20,700
and Liz Henelbury looked
at whether this difference

584
00:28:20,700 --> 00:28:23,500
in the blood pressure relation
with cognitive impairment

585
00:28:24,630 --> 00:28:26,340
is the same as it is for stroke,

586
00:28:26,340 --> 00:28:28,080
and she really didn't
find much difference.

587
00:28:28,080 --> 00:28:29,010
There was a little difference,

588
00:28:29,010 --> 00:28:31,710
but it wasn't statistically significant.

589
00:28:31,710 --> 00:28:36,060
So perhaps the same pattern,
but not clear enough

590
00:28:36,060 --> 00:28:38,920
in the data that we have so far.

591
00:28:38,920 --> 00:28:40,779
- [Crowd Member] So sorry, Mary.

592
00:28:40,779 --> 00:28:42,683
So the last slide, you're
saying that that difference

593
00:28:42,683 --> 00:28:44,600
in the pressure is not-

594
00:28:45,633 --> 00:28:46,660
- That this racial difference-

595
00:28:46,660 --> 00:28:47,790
- [Crowd Member] Okay, okay.

596
00:28:47,790 --> 00:28:49,020
- It doesn't look like that.

597
00:28:49,020 --> 00:28:52,072
It looks more like that. (laughs)

598
00:28:52,072 --> 00:28:52,920
- [Crowd Member] Oh,
okay. OKay, okay, okay.

599
00:28:52,920 --> 00:28:55,980
- Yeah, and the confidence
intervals are wide

600
00:28:55,980 --> 00:28:58,890
and it wasn't as robust.

601
00:28:58,890 --> 00:29:00,420
- Okay.
- Yeah.

602
00:29:00,420 --> 00:29:01,470
So she's working on that paper.

603
00:29:01,470 --> 00:29:03,930
I'm not involved with that paper,

604
00:29:03,930 --> 00:29:05,780
but that's what I saw at lab meeting.

605
00:29:10,500 --> 00:29:14,815
So we did a second visit
10 years after baseline.

606
00:29:14,815 --> 00:29:19,800
That grant renewal, part
of the purpose was to study

607
00:29:19,800 --> 00:29:21,960
why there are racial
differences in development

608
00:29:21,960 --> 00:29:24,060
of risk factors to begin with.

609
00:29:24,060 --> 00:29:26,967
Why do Black people
have higher hypertension

610
00:29:26,967 --> 00:29:29,522
and diabetes in particular?

611
00:29:29,522 --> 00:29:32,970
And so we got funded every, you know,

612
00:29:32,970 --> 00:29:34,800
five years just like everybody else.

613
00:29:34,800 --> 00:29:36,930
We have to get our grant renewed.

614
00:29:36,930 --> 00:29:41,100
And we tried to get it done sooner,

615
00:29:41,100 --> 00:29:43,860
but it took a couple tries, so it ended up

616
00:29:43,860 --> 00:29:46,140
being 10 years after
baseline, and it was just

617
00:29:46,140 --> 00:29:50,152
the same thing we did before,
and there were 16,000 people

618
00:29:50,152 --> 00:29:52,770
who participated in the phone interview.

619
00:29:52,770 --> 00:29:55,050
Not all of them did the home visit part.

620
00:29:55,050 --> 00:29:58,740
This gets to the point of, you know,

621
00:29:58,740 --> 00:30:00,960
they might have been losing
interest or what have you.

622
00:30:00,960 --> 00:30:02,307
We called them every six months.

623
00:30:02,307 --> 00:30:03,720
For the most part they like us,

624
00:30:03,720 --> 00:30:06,690
and we had higher retention
than we thought we would.

625
00:30:06,690 --> 00:30:11,100
a 97.5% per-year retention
rate, which was a problem,

626
00:30:11,100 --> 00:30:13,150
because the call center budget

627
00:30:15,914 --> 00:30:18,450
could not cover (laughs) all
the calls we needed to do.

628
00:30:18,450 --> 00:30:20,280
So it's a good problem to have,

629
00:30:20,280 --> 00:30:22,800
but we had to divert money
from other parts of the grant

630
00:30:22,800 --> 00:30:25,113
to make sure we can do those phone calls.

631
00:30:26,400 --> 00:30:30,120
So basically we observed
that Black individuals

632
00:30:30,120 --> 00:30:33,060
had twice the risk of
developing hypertension

633
00:30:33,060 --> 00:30:35,640
during follow-up as white individuals,

634
00:30:35,640 --> 00:30:40,380
and there's been a wide range of projects

635
00:30:40,380 --> 00:30:42,213
looking at reasons for this.

636
00:30:43,830 --> 00:30:48,390
And this paper was a labor
of love, I have to say.

637
00:30:48,390 --> 00:30:52,680
So George led this paper on determinants

638
00:30:52,680 --> 00:30:54,360
of the excess risk of hypertension

639
00:30:54,360 --> 00:30:55,320
in Black compared to white.

640
00:30:55,320 --> 00:30:58,140
Just like we did for
stroke, what are the factors

641
00:30:58,140 --> 00:31:00,540
that could explain this?

642
00:31:00,540 --> 00:31:05,540
And this went into JAMA after
three or four revisions.

643
00:31:06,000 --> 00:31:08,947
An 80-page response document.

644
00:31:08,947 --> 00:31:11,610
I told George, "Why are
you torturing yourself

645
00:31:11,610 --> 00:31:12,614
like this?" (laughs)

646
00:31:12,614 --> 00:31:14,897
And he said, "I wanna get it in JAMA."

647
00:31:14,897 --> 00:31:17,673
And he did succeed, but
it was really painful.

648
00:31:18,720 --> 00:31:22,380
And the differences in
the explanatory factors

649
00:31:22,380 --> 00:31:25,920
for developing hypertension
did differ by sex.

650
00:31:25,920 --> 00:31:29,970
And so for men it was
lower socioeconomic status,

651
00:31:29,970 --> 00:31:32,580
poor diet, high dietary sodium,

652
00:31:32,580 --> 00:31:35,103
and increased levels of inflammation,

653
00:31:36,030 --> 00:31:39,300
and for women there were added impacts

654
00:31:39,300 --> 00:31:43,203
of body size measures, obesity measures.

655
00:31:45,140 --> 00:31:46,790
So some differences by sex there.

656
00:31:49,680 --> 00:31:52,623
For diabetes, same sort of thing.

657
00:31:54,150 --> 00:31:57,960
So back in the data I
just showed you earlier

658
00:31:57,960 --> 00:32:00,480
about the factors that
explain stroke disparity,

659
00:32:00,480 --> 00:32:05,250
diabetes explained about
20 to 40% depending

660
00:32:05,250 --> 00:32:07,290
of the racial difference in stroke.

661
00:32:07,290 --> 00:32:10,110
And we observed that Black individuals

662
00:32:10,110 --> 00:32:12,120
had a twofold higher risk of diabetes,

663
00:32:12,120 --> 00:32:14,250
just like hypertension, developing over

664
00:32:14,250 --> 00:32:16,560
those 10 years of follow up.

665
00:32:16,560 --> 00:32:19,650
And the main mediators
which didn't differ by sex

666
00:32:19,650 --> 00:32:22,440
were adiposity measures, that poor diet

667
00:32:22,440 --> 00:32:24,903
and lower socioeconomic status.

668
00:32:25,800 --> 00:32:28,380
Brittany Palermo wrote this nice paper

669
00:32:28,380 --> 00:32:31,570
about interleukin six levels
in the risk of diabetes

670
00:32:32,910 --> 00:32:36,660
and showed that the
relationship was very large

671
00:32:36,660 --> 00:32:39,513
but similar in Black and white people.

672
00:32:40,650 --> 00:32:42,600
Brittany is a medical student here

673
00:32:42,600 --> 00:32:45,900
and just received the Research Week Award

674
00:32:45,900 --> 00:32:47,250
for best medical student paper.

675
00:32:47,250 --> 00:32:48,907
I was very proud of her.

676
00:32:51,090 --> 00:32:53,640
And Brittany's going into
the match for anesthesia,

677
00:32:54,660 --> 00:32:56,913
so she'll be graduating in the spring.

678
00:32:58,680 --> 00:33:00,870
So basically in terms of stroke,

679
00:33:00,870 --> 00:33:02,760
these traditional risk factors explain

680
00:33:02,760 --> 00:33:04,980
about half the disparity.

681
00:33:04,980 --> 00:33:07,470
Blood pressure, we call
it the triple curse

682
00:33:07,470 --> 00:33:10,080
because it's more common, it's
more difficult to control,

683
00:33:10,080 --> 00:33:12,319
it's more potent as a risk factor.

684
00:33:12,319 --> 00:33:14,220
This suggests that treatment goals

685
00:33:14,220 --> 00:33:16,320
for blood pressure maybe
needs to be different

686
00:33:16,320 --> 00:33:17,960
for Black individuals, but
boy wouldn't it be nice

687
00:33:17,960 --> 00:33:19,620
if we could prevent high blood pressure

688
00:33:19,620 --> 00:33:22,443
in Black individuals
starting in the first place.

689
00:33:23,310 --> 00:33:27,240
That's primordial prevention,
preventing the risk factors.

690
00:33:27,240 --> 00:33:29,613
That would be the holy grail really.

691
00:33:31,108 --> 00:33:33,510
And a better understanding of the origins

692
00:33:33,510 --> 00:33:36,600
of the race difference and
risk factors might allow

693
00:33:36,600 --> 00:33:38,550
the best way to reduce stroke risk

694
00:33:38,550 --> 00:33:40,290
rather than try to treat the risk factor.

695
00:33:40,290 --> 00:33:42,780
And I showed you just a little bit of that

696
00:33:42,780 --> 00:33:45,483
and there's a lot of work
going on in that area.

697
00:33:47,340 --> 00:33:52,340
So this is kind of the
glass half full argument,

698
00:33:53,130 --> 00:33:55,350
which was actually in the
title of the paper that showed

699
00:33:55,350 --> 00:33:57,810
the risk factors that
explain stroke disparities.

700
00:33:57,810 --> 00:33:59,310
I debated George about that.

701
00:33:59,310 --> 00:34:01,500
I wasn't, like kinda stupid to put that

702
00:34:01,500 --> 00:34:02,580
in the title (laughs) of the paper.

703
00:34:02,580 --> 00:34:04,800
He's like, "No, I won't have it."

704
00:34:04,800 --> 00:34:07,410
But basically you have
this unexplained risk

705
00:34:07,410 --> 00:34:11,520
that justifies us getting more grants

706
00:34:11,520 --> 00:34:15,570
and more funding to study,
but it's really important.

707
00:34:15,570 --> 00:34:17,520
So the question is whether lab markers

708
00:34:17,520 --> 00:34:22,520
could help explain this,
and those lab markers

709
00:34:22,650 --> 00:34:24,450
could work in a couple of different ways.

710
00:34:24,450 --> 00:34:28,740
They could be mediators like
the analyses I've shown you.

711
00:34:28,740 --> 00:34:31,680
They're in the pathway, they're involved.

712
00:34:31,680 --> 00:34:34,530
At least the biology they
represent is involved.

713
00:34:34,530 --> 00:34:36,090
Or they're moderators, which means

714
00:34:36,090 --> 00:34:37,620
they're race-specific risk factors

715
00:34:37,620 --> 00:34:41,010
meaning there's differences
in the relationship by race.

716
00:34:41,010 --> 00:34:43,210
Might be stronger in
one group than another.

717
00:34:44,408 --> 00:34:48,003
Have lots of people working
on this here over the years.

718
00:34:48,960 --> 00:34:53,850
Lot of med students,
postdocs, some fellows,

719
00:34:53,850 --> 00:34:56,760
faculty members, and that's really great

720
00:34:56,760 --> 00:34:58,774
to keep adding to that picture.

721
00:34:58,774 --> 00:35:00,090
So it takes a lot of memory.

722
00:35:00,090 --> 00:35:02,934
You have to snip it, make it smaller

723
00:35:02,934 --> 00:35:06,150
as you get to having more people
working with you. (laughs)

724
00:35:06,150 --> 00:35:08,100
All those photographs are a lot.

725
00:35:08,100 --> 00:35:12,140
So we embarked on something
called a case cohort study

726
00:35:12,140 --> 00:35:14,460
to study laboratory markers.

727
00:35:14,460 --> 00:35:17,460
And when you have 30,000
people in the study,

728
00:35:17,460 --> 00:35:19,980
you don't need to measure
lab markers on everyone.

729
00:35:19,980 --> 00:35:22,740
You measure them in these subsets,

730
00:35:22,740 --> 00:35:26,850
and this design will approximate

731
00:35:26,850 --> 00:35:28,920
what the results would
be if you measured it

732
00:35:28,920 --> 00:35:31,260
at everyone if you do it right.

733
00:35:31,260 --> 00:35:35,910
So a case cohort study is when
you have followed the people

734
00:35:35,910 --> 00:35:38,333
for some number of years
and then you go back

735
00:35:38,333 --> 00:35:41,490
and identify cases, in this case

736
00:35:41,490 --> 00:35:43,710
stroke and cognitive impairment,

737
00:35:43,710 --> 00:35:47,100
and then we take a random
sample from the cohort,

738
00:35:47,100 --> 00:35:49,170
and in this case it was 1,100,

739
00:35:49,170 --> 00:35:51,117
and we had those numbers of stroke

740
00:35:51,117 --> 00:35:54,420
and cognitive impairment
cases during follow up,

741
00:35:54,420 --> 00:35:57,776
and pull the samples out of the
freezer just on those people

742
00:35:57,776 --> 00:36:00,840
and measure a whole variety of things.

743
00:36:00,840 --> 00:36:04,530
And some of the measures we'll go through.

744
00:36:04,530 --> 00:36:07,710
There's a lot more, but
just to give you examples,

745
00:36:07,710 --> 00:36:10,323
you can have genetic-related markers,

746
00:36:12,420 --> 00:36:14,460
have circulating proteins.

747
00:36:14,460 --> 00:36:18,450
There's a guy at MGH
who's doing metabolomics

748
00:36:18,450 --> 00:36:19,860
who's written some really nice papers.

749
00:36:19,860 --> 00:36:22,203
I don't have time to go into his work.

750
00:36:23,850 --> 00:36:25,980
He did that in an ancillary study,

751
00:36:25,980 --> 00:36:29,760
and he's linking diet
to metabolomic markers

752
00:36:29,760 --> 00:36:33,360
and trying to make causal
pathway sorts of analyses

753
00:36:33,360 --> 00:36:35,520
and found some really
interesting findings.

754
00:36:35,520 --> 00:36:38,193
So sickle trait was a
high-priority thing for us.

755
00:36:39,511 --> 00:36:44,130
As time was going on, we
have learned over the years

756
00:36:44,130 --> 00:36:47,460
that sickle cell disease
and sickle cell trait

757
00:36:47,460 --> 00:36:49,470
are prothrombotic conditions.

758
00:36:49,470 --> 00:36:52,410
Sickle cell trait, the carrier
ship of sickle cell disease,

759
00:36:52,410 --> 00:36:57,240
hemoglobin AS is present
in 8% of African Americans.

760
00:36:57,240 --> 00:37:02,240
And it's been reported by us and others

761
00:37:02,670 --> 00:37:05,340
in different studies
that sickle cell disease

762
00:37:05,340 --> 00:37:07,470
and sickle cell trait are risk factors

763
00:37:07,470 --> 00:37:09,330
for venous thrombosis.

764
00:37:09,330 --> 00:37:13,140
And the idea is probably that
you have, venous thrombosis

765
00:37:13,140 --> 00:37:16,050
starts on the backside of these valves,

766
00:37:16,050 --> 00:37:20,110
the venous valves in the legs, and when

767
00:37:22,500 --> 00:37:26,786
these valves are hypoxic
actually, the PO2 I think it's 50

768
00:37:26,786 --> 00:37:28,959
in those valves.

769
00:37:28,959 --> 00:37:33,959
And so we think that even
heterozygous carriers

770
00:37:34,350 --> 00:37:37,050
of sickle trait, that the
red cells could sickle

771
00:37:37,050 --> 00:37:38,250
in that hypoxic environment.

772
00:37:38,250 --> 00:37:40,890
And interestingly, sickle
cell trait is also related

773
00:37:40,890 --> 00:37:43,263
to the risk of developing kidney disease

774
00:37:43,263 --> 00:37:47,550
where there are also in the
nephron areas of hypoxia

775
00:37:47,550 --> 00:37:50,077
like around an O2 750.

776
00:37:52,034 --> 00:37:54,390
So a really interesting
concept which led us

777
00:37:54,390 --> 00:37:56,850
to think about, well, maybe
it would relate to stroke

778
00:37:56,850 --> 00:37:58,290
and cognitive impairment as well

779
00:37:58,290 --> 00:38:00,600
because of hypercoagulability.

780
00:38:00,600 --> 00:38:04,410
And so this is a paper that merged data

781
00:38:04,410 --> 00:38:06,630
from multiple studies, including REGARDS,

782
00:38:06,630 --> 00:38:10,650
led by Hyacinth Hyacinth,
who was here last year

783
00:38:10,650 --> 00:38:14,250
giving this presentation,
and we showed no association

784
00:38:14,250 --> 00:38:16,650
of sickle cell trait with stroke risk.

785
00:38:16,650 --> 00:38:18,727
So here's your head ratio of one

786
00:38:18,727 --> 00:38:21,450
and the risk estimates for each study,

787
00:38:21,450 --> 00:38:23,880
and nothing is significantly different

788
00:38:23,880 --> 00:38:26,883
than a one, which means
there's no increased risk.

789
00:38:28,680 --> 00:38:29,793
So that was cool.

790
00:38:30,900 --> 00:38:35,520
Here Christina Cahill who
was an medical student here

791
00:38:35,520 --> 00:38:39,510
did that analysis for
cognitive impairment in REGARDS

792
00:38:39,510 --> 00:38:42,240
and she looked at it different ways,

793
00:38:42,240 --> 00:38:45,120
both as becoming impaired
when you were normal

794
00:38:45,120 --> 00:38:47,020
and there's a definition for that

795
00:38:47,020 --> 00:38:48,450
that I'm not gonna go
into, but also looking

796
00:38:48,450 --> 00:38:52,290
at the trajectories of
test scores over time

797
00:38:52,290 --> 00:38:54,990
in people with and without sickle trait

798
00:38:54,990 --> 00:38:59,103
and saw no association
with cognitive impairment.

799
00:39:01,860 --> 00:39:05,160
Lipoprotein(a) is another
genetic related factor.

800
00:39:05,160 --> 00:39:09,180
So Lp(a) is an atherogenic lipoprotein

801
00:39:09,180 --> 00:39:13,980
and it promotes pro-inflammatory milieu

802
00:39:13,980 --> 00:39:16,500
in atherosclerotic plaques, and it also

803
00:39:16,500 --> 00:39:18,243
enhances abnormal clotting.

804
00:39:19,440 --> 00:39:21,900
It's mostly genetically determined.

805
00:39:21,900 --> 00:39:23,340
Pretty much you only
have to measure it once

806
00:39:23,340 --> 00:39:25,227
in a person's life and you
know what their value is

807
00:39:25,227 --> 00:39:26,823
and it really doesn't change.

808
00:39:29,070 --> 00:39:30,871
One in four people have high levels,

809
00:39:30,871 --> 00:39:35,283
more common in Black than
white people to be high.

810
00:39:36,870 --> 00:39:39,000
What we observed is
this really interesting

811
00:39:39,000 --> 00:39:40,260
difference by race.

812
00:39:40,260 --> 00:39:43,860
So with stroke risk, this is Pankaj Arora

813
00:39:43,860 --> 00:39:46,165
who's a cardiologist down at UAB,

814
00:39:46,165 --> 00:39:50,160
and the relationship of
higher Lp(a) with stroke

815
00:39:50,160 --> 00:39:53,790
was present only in Black individuals.

816
00:39:53,790 --> 00:39:56,880
So they had a twofold
greater risk for high Lp(a)

817
00:39:56,880 --> 00:39:58,650
after accounting for other risk factors

818
00:39:58,650 --> 00:40:01,000
while for white people
there was no difference.

819
00:40:02,723 --> 00:40:04,470
And we saw the same pattern

820
00:40:04,470 --> 00:40:06,480
for the risk of cognitive impairment.

821
00:40:06,480 --> 00:40:08,520
So cognitive impairment, this definition

822
00:40:08,520 --> 00:40:10,830
is kind of like a mild
cognitive impairment

823
00:40:10,830 --> 00:40:13,080
or worse phenotype.

824
00:40:13,080 --> 00:40:16,320
It's not dementia, if you
will, based on the test scores

825
00:40:16,320 --> 00:40:19,690
and how your scores relate to other people

826
00:40:20,730 --> 00:40:23,943
and what scores signify an impaired state.

827
00:40:25,252 --> 00:40:28,410
So here we saw the same
thing, weaker relationships.

828
00:40:28,410 --> 00:40:30,660
Often the relationships
of the stroke risk factors

829
00:40:30,660 --> 00:40:33,420
with cognition using
this outcome in REGARDS

830
00:40:33,420 --> 00:40:35,703
are weaker than they are for stroke.

831
00:40:37,380 --> 00:40:41,880
But the same sort of pattern
seen for cognitive impairment.

832
00:40:41,880 --> 00:40:42,960
- [Crowd Member] Mary,
can I ask a question?

833
00:40:42,960 --> 00:40:46,920
So you're studying these
risk factors by themselves.

834
00:40:46,920 --> 00:40:49,590
You're not cross correlating
a whole bunch of risk factors?

835
00:40:49,590 --> 00:40:51,390
- What we're doing in these analyses,

836
00:40:51,390 --> 00:40:53,267
we're adjusting for the
traditional factors,

837
00:40:53,267 --> 00:40:55,110
- Right.
- The ones I was describing

838
00:40:55,110 --> 00:40:56,520
in the first part of the talk.

839
00:40:56,520 --> 00:40:57,660
- But not all-
- We're not throwing

840
00:40:57,660 --> 00:40:59,250
all the biomarkers together.

841
00:40:59,250 --> 00:41:00,473
- [Crowd Member] Is that all the time?

842
00:41:01,410 --> 00:41:02,243
- You could do it.

843
00:41:02,243 --> 00:41:04,140
I mean we're trying to pick apart biology,

844
00:41:05,160 --> 00:41:09,000
so I'm not really as interested in...

845
00:41:09,000 --> 00:41:11,490
You know, like Lp(a) levels for example

846
00:41:11,490 --> 00:41:15,150
don't really relate much to sickle trait.

847
00:41:15,150 --> 00:41:17,094
They're not gonna be different, right?

848
00:41:17,094 --> 00:41:18,390
- So that analysis-
- So there's not really

849
00:41:18,390 --> 00:41:20,259
a point to putting them
up against each other

850
00:41:20,259 --> 00:41:22,050
to study the biology.

851
00:41:22,050 --> 00:41:24,690
If you wanted to know like
which ones could be used

852
00:41:24,690 --> 00:41:27,330
in the clinic, you could throw 'em all in

853
00:41:27,330 --> 00:41:30,333
to pick the best ones,
but to study biology,

854
00:41:32,070 --> 00:41:34,440
potentially if you wanted to
adjust for other biomarkers,

855
00:41:34,440 --> 00:41:36,150
you would adjust for ones that you think

856
00:41:36,150 --> 00:41:38,283
are part of that pathway.

857
00:41:39,150 --> 00:41:40,680
- [Crowd Member] But
with multi-correlation,

858
00:41:40,680 --> 00:41:42,360
maybe 100 biomarkers-

859
00:41:42,360 --> 00:41:43,800
- Becomes problematic.

860
00:41:43,800 --> 00:41:44,633
- Right.
- Yeah.

861
00:41:44,633 --> 00:41:46,260
- [Crowd Member] So you don't
know the interaction with-

862
00:41:46,260 --> 00:41:48,120
- You can test for that.

863
00:41:48,120 --> 00:41:49,710
We're gonna be doing some of that later.

864
00:41:49,710 --> 00:41:53,460
I'll introduce that at the end. Yeah.

865
00:41:53,460 --> 00:41:57,750
So Lp(a), there are drugs in development

866
00:41:57,750 --> 00:42:02,160
for lowering Lp(a) and they
are being tested in people

867
00:42:02,160 --> 00:42:04,160
with established cardiovascular disease.

868
00:42:05,250 --> 00:42:07,108
We're waiting for the trials to end.

869
00:42:07,108 --> 00:42:09,573
Should be maybe in the next year.

870
00:42:11,250 --> 00:42:13,680
And you can make a
hypothesis that these drugs,

871
00:42:13,680 --> 00:42:16,320
at least from the point of
view of stroke and cognition,

872
00:42:16,320 --> 00:42:19,410
might have greater effects
in Black than white people

873
00:42:19,410 --> 00:42:22,023
or be more useful in
Black than white people.

874
00:42:24,720 --> 00:42:27,180
So I wanna talk about
inflammatory cytokines

875
00:42:27,180 --> 00:42:32,180
and I want to point out this
work was done by Nancy Jenny.

876
00:42:33,680 --> 00:42:35,670
I'm not sure anyone here knew her.

877
00:42:35,670 --> 00:42:39,450
So Nancy was a faculty member
in our group in Colchester,

878
00:42:39,450 --> 00:42:44,450
and Maria knew her, and Nancy
really was a wonderful person,

879
00:42:45,477 --> 00:42:48,310
and Nancy died suddenly in an accident

880
00:42:49,440 --> 00:42:53,940
about five years ago?

881
00:42:53,940 --> 00:42:54,773
Six years ago.

882
00:42:55,830 --> 00:42:58,740
And Nancy's office was next to mine.

883
00:42:58,740 --> 00:43:01,673
We roomed together when we
traveled, we worked together.

884
00:43:01,673 --> 00:43:05,670
She was an amazing, giving,
and caring scientist.

885
00:43:05,670 --> 00:43:07,230
Very collaborative.

886
00:43:07,230 --> 00:43:10,380
And I wanna take a moment
just to mention her,

887
00:43:10,380 --> 00:43:14,843
because this was a tragic
thing for our group

888
00:43:14,843 --> 00:43:16,830
to lose a person like that.

889
00:43:16,830 --> 00:43:18,960
There one day, gone the next.

890
00:43:18,960 --> 00:43:21,840
No warning, no indication
anything was wrong.

891
00:43:21,840 --> 00:43:23,310
It was an accident.

892
00:43:23,310 --> 00:43:26,373
So just appreciate the people around you.

893
00:43:26,373 --> 00:43:31,373
This really taught me to just
take a little extra effort

894
00:43:31,584 --> 00:43:34,260
to appreciate the people
that are around you,

895
00:43:34,260 --> 00:43:36,243
because they might not stay.

896
00:43:37,260 --> 00:43:40,275
So anyway, after she passed away,

897
00:43:40,275 --> 00:43:43,230
this paper she had been
finishing up at the time,

898
00:43:43,230 --> 00:43:45,000
we were able to get it published,

899
00:43:45,000 --> 00:43:45,900
which we were very proud of.

900
00:43:45,900 --> 00:43:48,060
And then the journal "Neurology"
made this infographic

901
00:43:48,060 --> 00:43:50,910
for us, which is nice, 'cause
you can chop up pieces of it

902
00:43:50,910 --> 00:43:53,940
and put it in a talk like this.

903
00:43:53,940 --> 00:43:58,800
So she looked at interleukin 6,

904
00:43:58,800 --> 00:44:01,967
IL-8, and IL-10 as the cytokines
we had chosen to measure

905
00:44:01,967 --> 00:44:04,197
in that case cohort study of stroke

906
00:44:04,197 --> 00:44:05,760
and cognitive impairment risk.

907
00:44:05,760 --> 00:44:10,350
And higher IL-6 was the
only one that was related

908
00:44:10,350 --> 00:44:13,320
to stroke, and it was a
substantial relationship.

909
00:44:13,320 --> 00:44:16,560
So levels in the top
quarter of the distribution

910
00:44:16,560 --> 00:44:20,190
compared to the bottom
doubled stroke risk.

911
00:44:20,190 --> 00:44:24,843
So the inflammatory basis
of stroke, very important.

912
00:44:26,400 --> 00:44:29,937
And it was independent of adding

913
00:44:29,937 --> 00:44:32,310
all the typical stroke risk factors.

914
00:44:32,310 --> 00:44:34,500
- [Crowd Member] I had a quick question.

915
00:44:34,500 --> 00:44:37,140
Are all these results you're describing,

916
00:44:37,140 --> 00:44:40,350
are these coming from what you
say were ancillary studies?

917
00:44:40,350 --> 00:44:41,640
- No, this was the main study.

918
00:44:41,640 --> 00:44:45,052
That case cohort study was
an aim in one of our...

919
00:44:45,052 --> 00:44:47,979
But ancillary studies that
built on the framework

920
00:44:47,979 --> 00:44:50,310
of that case cohort study to add things,

921
00:44:50,310 --> 00:44:52,410
like the guy that did the metabolomics.

922
00:44:52,410 --> 00:44:55,230
He said, well, I'm gonna
use those same people,

923
00:44:55,230 --> 00:44:57,165
'cause you've showed us that it works,

924
00:44:57,165 --> 00:45:00,990
and do metabolomics
layered on top of this.

925
00:45:00,990 --> 00:45:04,950
So if we wanted to look at
pathways for metabolomic markers,

926
00:45:04,950 --> 00:45:08,910
we could kind of play around with the data

927
00:45:08,910 --> 00:45:10,320
using the other biomarkers we've...

928
00:45:10,320 --> 00:45:13,560
We've measured a whole variety
of things on these people.

929
00:45:13,560 --> 00:45:16,123
So yeah, but good question.

930
00:45:16,123 --> 00:45:19,320
So what we discovered in this paper,

931
00:45:19,320 --> 00:45:23,973
what Nancy discovered, is
that the IL-6 mediation

932
00:45:27,150 --> 00:45:29,890
of the race disparity
in stroke was actually

933
00:45:29,890 --> 00:45:32,880
through the other risk factors.

934
00:45:32,880 --> 00:45:34,290
I'm not gonna show you all the statistics

935
00:45:34,290 --> 00:45:36,360
around that, but just trust me.

936
00:45:36,360 --> 00:45:40,620
And the way we concluded
that part of the paper

937
00:45:40,620 --> 00:45:43,830
is that the stroke risk
factors are pro-inflammatory

938
00:45:43,830 --> 00:45:48,660
which leads to stroke in
a way that's more relevant

939
00:45:48,660 --> 00:45:52,293
for Black individuals than
white individuals, okay?

940
00:45:53,700 --> 00:45:58,680
So the question would be from that

941
00:45:58,680 --> 00:46:02,010
is maybe you could manipulate inflammation

942
00:46:02,010 --> 00:46:03,150
to prevent stroke.

943
00:46:03,150 --> 00:46:04,470
Now, we already do that.

944
00:46:04,470 --> 00:46:07,230
Statins lower inflammation, for example.

945
00:46:07,230 --> 00:46:09,300
Physical activity lowers inflammation.

946
00:46:09,300 --> 00:46:11,790
So a lot of things that are
healthy lifestyle things

947
00:46:11,790 --> 00:46:16,143
and primary prevention
treatments lower inflammation.

948
00:46:17,250 --> 00:46:21,870
And if you could figure out pathways

949
00:46:21,870 --> 00:46:24,865
that might be targetable in different ways

950
00:46:24,865 --> 00:46:27,720
than the things you already
do, maybe lower inflammation

951
00:46:27,720 --> 00:46:30,510
would reduce racial disparity.

952
00:46:30,510 --> 00:46:32,820
Daniela Zambrano in vascular neurology

953
00:46:32,820 --> 00:46:35,970
is working on a paper now
looking at these cytokines

954
00:46:35,970 --> 00:46:37,820
and the risk of cognitive impairment.

955
00:46:41,400 --> 00:46:43,980
Turns out IL-6 is probably
the most important one

956
00:46:43,980 --> 00:46:44,813
for that too.

957
00:46:46,170 --> 00:46:47,940
So here's just some other biomarkers.

958
00:46:47,940 --> 00:46:50,410
I'm not gonna spend much time on it

959
00:46:50,410 --> 00:46:51,870
'cause I wanna get through the rest,

960
00:46:51,870 --> 00:46:55,050
but just some sampling of some
of the things we've measured.

961
00:46:55,050 --> 00:46:58,470
And what you can see in
general, and this holds

962
00:46:58,470 --> 00:47:01,350
for other ones that we've
done, is the relationships

963
00:47:01,350 --> 00:47:05,730
are often concordant for stroke
and cognitive impairment.

964
00:47:05,730 --> 00:47:08,763
It reflects the underlying
common pathophysiology, right?

965
00:47:10,025 --> 00:47:11,550
And the relationships
with cognitive impairment

966
00:47:11,550 --> 00:47:15,227
often are a little weaker
than they are for stroke,

967
00:47:15,227 --> 00:47:17,580
except for blood group, which
is interesting. (laughs)

968
00:47:17,580 --> 00:47:21,780
You know, blood group
AB or non-O blood groups

969
00:47:21,780 --> 00:47:23,881
are related to hypercoagulability

970
00:47:23,881 --> 00:47:27,183
and impaired endothelial function.

971
00:47:28,290 --> 00:47:31,140
And so blood group AB
is an interesting proxy

972
00:47:31,140 --> 00:47:33,990
for really interesting biologies

973
00:47:33,990 --> 00:47:36,212
that relate to cardiovascular diseases,

974
00:47:36,212 --> 00:47:40,260
and so that was similar for
stroke and cognitive impairment.

975
00:47:40,260 --> 00:47:42,752
When my postdoc published
this paper (laughs)

976
00:47:42,752 --> 00:47:46,830
on cognitive impairment, it
was like a media firestorm.

977
00:47:46,830 --> 00:47:47,910
It was all over the world.

978
00:47:47,910 --> 00:47:50,548
It was on all the media outlets, you know,

979
00:47:50,548 --> 00:47:54,930
saying that your blood group
tells you about your cognition.

980
00:47:54,930 --> 00:47:57,150
It was a very like sexy topic.

981
00:47:57,150 --> 00:47:59,700
We had emails from people,
I had emails from this lady

982
00:47:59,700 --> 00:48:03,240
in England who said,
"Oh, I finally understand

983
00:48:03,240 --> 00:48:05,755
why my brain is the way it is!" (laughs)

984
00:48:05,755 --> 00:48:08,219
And she's like, "Thank you
for doing this research."

985
00:48:08,219 --> 00:48:09,235
It was so funny.

986
00:48:09,235 --> 00:48:11,910
But it never seemed to stop.

987
00:48:11,910 --> 00:48:13,710
It lasted for like three years.

988
00:48:13,710 --> 00:48:17,250
Every once in a while a
health writer would contact us

989
00:48:17,250 --> 00:48:19,310
asking to talk about that paper.

990
00:48:19,310 --> 00:48:20,553
It was so weird.

991
00:48:22,650 --> 00:48:26,100
So this part I wanna
talk about the importance

992
00:48:26,100 --> 00:48:30,300
of seeing who's around you in your travels

993
00:48:30,300 --> 00:48:33,810
and interacting with
people and why networking

994
00:48:33,810 --> 00:48:36,297
is so important,
especially for those of you

995
00:48:36,297 --> 00:48:37,710
who are just starting.

996
00:48:37,710 --> 00:48:41,310
So the International Society
on Thrombosis and Hemostasis,

997
00:48:41,310 --> 00:48:44,430
one of my groups, has
an early career lounge

998
00:48:44,430 --> 00:48:45,750
at their annual meeting, which I think

999
00:48:45,750 --> 00:48:48,218
most groups have these days.

1000
00:48:48,218 --> 00:48:52,620
And I was in the early career
lounge as the journal editor

1001
00:48:52,620 --> 00:48:56,070
for the "Society" journal
wanting to just talk to people,

1002
00:48:56,070 --> 00:48:58,350
meet people, answer their questions

1003
00:48:58,350 --> 00:49:00,390
about publishing and what have you,

1004
00:49:00,390 --> 00:49:04,740
and this guy, Robbie
Campbell, comes up to me

1005
00:49:04,740 --> 00:49:07,387
and he says, "Oh, can we
talk about this project?"

1006
00:49:07,387 --> 00:49:08,670
And I'm like okay.

1007
00:49:08,670 --> 00:49:12,600
So Robbie had a postdoc, Frederik Denorme,

1008
00:49:12,600 --> 00:49:16,650
working on this project in the lab

1009
00:49:16,650 --> 00:49:20,340
looking at a Power Four variant,

1010
00:49:20,340 --> 00:49:25,340
which is a platelet protein

1011
00:49:26,460 --> 00:49:31,460
that is a variant

1012
00:49:33,150 --> 00:49:34,980
more common in Black individuals,

1013
00:49:34,980 --> 00:49:37,530
in people of African
ancestry, I should say.

1014
00:49:37,530 --> 00:49:42,530
So they were studying a
mouse that had this mutation

1015
00:49:44,970 --> 00:49:48,570
that is more common in
people of African ancestry

1016
00:49:48,570 --> 00:49:50,190
and they were showing that these mice

1017
00:49:50,190 --> 00:49:52,410
compared to wild type had bigger strokes

1018
00:49:52,410 --> 00:49:54,090
in the stroke model.

1019
00:49:54,090 --> 00:49:59,090
And they were working out the
physiology of that in the lab.

1020
00:49:59,610 --> 00:50:03,360
And he said, "Can we look
at that variant in REGARDS?"

1021
00:50:03,360 --> 00:50:06,600
And I said, "Sure, we have
GWAS on all Black individuals,

1022
00:50:06,600 --> 00:50:09,930
so we can look at the
association of this variant

1023
00:50:09,930 --> 00:50:12,747
with stroke risk in Black individuals."

1024
00:50:14,220 --> 00:50:19,220
And so with this variant,

1025
00:50:19,770 --> 00:50:21,990
anti-platelet drugs seem less effective

1026
00:50:21,990 --> 00:50:22,980
in the animal model.

1027
00:50:22,980 --> 00:50:24,703
That's kind of interesting.

1028
00:50:25,800 --> 00:50:29,790
And what we found is that the variant

1029
00:50:29,790 --> 00:50:32,910
was more common in African ancestry,

1030
00:50:32,910 --> 00:50:36,303
or by proxy Black individuals in REGARDS.

1031
00:50:37,860 --> 00:50:41,545
62% of them had the
variant compared to 21%

1032
00:50:41,545 --> 00:50:43,683
of the white participants.

1033
00:50:45,720 --> 00:50:50,640
And in the mice, Frederik had observed,

1034
00:50:50,640 --> 00:50:53,070
as I said, that they have worse strokes.

1035
00:50:53,070 --> 00:50:55,023
They had greater net formation.

1036
00:50:56,430 --> 00:50:59,013
Part of the possible pathway for that.

1037
00:51:00,150 --> 00:51:04,380
And we showed an increased risk of stroke

1038
00:51:04,380 --> 00:51:07,140
with that variant, 1.264 increase.

1039
00:51:07,140 --> 00:51:10,110
Modest relationship for a gene variant

1040
00:51:10,110 --> 00:51:13,623
that's a pretty good, pretty solid number.

1041
00:51:15,150 --> 00:51:18,180
Most gene variants are, you
know, mild risk factors.

1042
00:51:18,180 --> 00:51:21,510
So 25% increased risk if
you have this variant,

1043
00:51:21,510 --> 00:51:24,360
which is much more common
in Black individuals,

1044
00:51:24,360 --> 00:51:26,310
so hey, there's another one, right?

1045
00:51:26,310 --> 00:51:29,180
Maybe it is partly explaining
this pattern we see

1046
00:51:29,180 --> 00:51:31,230
in the population.

1047
00:51:31,230 --> 00:51:36,230
And then Robbie is like, "Well,
the mice get worse strokes.

1048
00:51:38,160 --> 00:51:39,767
Can we look at the size of the stroke

1049
00:51:39,767 --> 00:51:41,661
or the severity of the stroke?"

1050
00:51:41,661 --> 00:51:43,290
And I'm like, we've
never done that before.

1051
00:51:43,290 --> 00:51:45,570
This is why it's so good
to collaborate with people

1052
00:51:45,570 --> 00:51:47,190
who think differently than you.

1053
00:51:47,190 --> 00:51:50,610
And it turned out that
there was this project

1054
00:51:50,610 --> 00:51:53,550
that had extracted from the charts

1055
00:51:53,550 --> 00:51:58,550
of several hundred of
the stroke cases we had

1056
00:51:58,860 --> 00:51:59,910
stroke severity measures.

1057
00:51:59,910 --> 00:52:02,460
There are validated measures
that physicians use,

1058
00:52:02,460 --> 00:52:04,665
kind of required for them to calculate

1059
00:52:04,665 --> 00:52:07,950
in the records in every stroke patient.

1060
00:52:07,950 --> 00:52:11,880
And indeed, we showed that
if they had this mutation,

1061
00:52:11,880 --> 00:52:15,990
they had higher odds, a higher risk

1062
00:52:15,990 --> 00:52:18,210
of more unfavorable stroke outcome.

1063
00:52:18,210 --> 00:52:21,000
So 70% higher risk of unfavorable.

1064
00:52:21,000 --> 00:52:23,970
So maybe they were having
also bigger strokes

1065
00:52:23,970 --> 00:52:24,870
if they have this.

1066
00:52:27,660 --> 00:52:29,250
So that was cool!

1067
00:52:29,250 --> 00:52:32,010
And I would say any of you in the room,

1068
00:52:32,010 --> 00:52:33,945
you know, working in vascular function,

1069
00:52:33,945 --> 00:52:37,380
and Osama, we've tried to (laughs) do

1070
00:52:37,380 --> 00:52:39,953
some translational work around a PAO1

1071
00:52:41,091 --> 00:52:42,060
and we had a lot of challenges with it

1072
00:52:42,060 --> 00:52:43,833
that someday we'll maybe overcome,

1073
00:52:45,450 --> 00:52:47,940
but this kind of work is not hard to do,

1074
00:52:47,940 --> 00:52:50,147
especially if you have
an identified mutation

1075
00:52:50,147 --> 00:52:54,390
that you're studying
that is prevalent enough

1076
00:52:54,390 --> 00:52:56,613
in the population that we can study it.

1077
00:52:58,050 --> 00:52:59,460
What about clinical applications?

1078
00:52:59,460 --> 00:53:01,410
This is really not related
to race differences,

1079
00:53:01,410 --> 00:53:05,970
but Sam Short, who graduated
from medical school

1080
00:53:05,970 --> 00:53:09,400
here the year before last, he
also won the Best Paper Award

1081
00:53:11,160 --> 00:53:14,910
last year after he left
when he was an intern

1082
00:53:14,910 --> 00:53:16,284
down at Chapel Hill.

1083
00:53:16,284 --> 00:53:20,640
He looked at, we had some
money in our carry forward,

1084
00:53:20,640 --> 00:53:23,100
like a lot of money during the pandemic,

1085
00:53:23,100 --> 00:53:24,840
and we decided we're gonna
measure a whole bunch

1086
00:53:24,840 --> 00:53:27,227
of biomarkers in everybody in REGARDS

1087
00:53:27,227 --> 00:53:29,820
to baseline atrial fibrillation,
because atrial fibrillation

1088
00:53:29,820 --> 00:53:32,040
is an important risk factor for stroke.

1089
00:53:32,040 --> 00:53:35,490
And we use anticoagulant therapy

1090
00:53:35,490 --> 00:53:38,220
to prevent stroke in people with AFib,

1091
00:53:38,220 --> 00:53:41,070
and so we thought, well, can
we make that risk prediction

1092
00:53:41,070 --> 00:53:43,530
for who should get anticoagulation better

1093
00:53:43,530 --> 00:53:44,700
by using biomarkers?

1094
00:53:44,700 --> 00:53:47,000
There's risk scores that physicians use

1095
00:53:47,000 --> 00:53:50,520
in primary care to decide who
should get anticoagulation

1096
00:53:50,520 --> 00:53:52,140
or who needs it and who doesn't,

1097
00:53:52,140 --> 00:53:55,062
because these scores predict stroke risk.

1098
00:53:55,062 --> 00:54:00,062
And so Sam is a great analyst
and he's a wonderful creator

1099
00:54:01,260 --> 00:54:05,638
of imagery, (laughs)
and so this was a figure

1100
00:54:05,638 --> 00:54:08,700
he created looking at these biomarkers

1101
00:54:08,700 --> 00:54:10,833
that cut across different domains.

1102
00:54:11,700 --> 00:54:13,530
Clotting activity, cardiac function,

1103
00:54:13,530 --> 00:54:16,953
kidney function, vascular
dysfunction, et cetera.

1104
00:54:17,910 --> 00:54:19,380
Liver dysfunction.

1105
00:54:19,380 --> 00:54:22,710
And you can see several
of these biomarkers

1106
00:54:22,710 --> 00:54:26,940
from here over had greater
risk of stroke in AFib.

1107
00:54:26,940 --> 00:54:30,690
And these are all in
people who were not taking

1108
00:54:30,690 --> 00:54:34,170
an anticoagulant at baseline,
so they're the kind of people

1109
00:54:34,170 --> 00:54:36,480
who would be risk assessed, for example,

1110
00:54:36,480 --> 00:54:38,780
for whether they should
take an anticoagulant.

1111
00:54:39,720 --> 00:54:43,560
And accounting for all the
risk factors for stroke,

1112
00:54:43,560 --> 00:54:46,740
you see these relationships,
and like we saw

1113
00:54:46,740 --> 00:54:49,020
in the general population,
the largest association

1114
00:54:49,020 --> 00:54:53,700
was for NT-proBNP, measure of atriopathy,

1115
00:54:53,700 --> 00:54:57,363
left atrial dysfunction, which makes sense

1116
00:54:57,363 --> 00:55:01,050
for its relationship with
AFib with stroke risk,

1117
00:55:01,050 --> 00:55:03,420
'cause it's the atrium that's sick in AFib

1118
00:55:03,420 --> 00:55:07,347
and creating clots that
embolize to the brain.

1119
00:55:07,347 --> 00:55:11,910
GDF-15 was another one
which is turning out

1120
00:55:11,910 --> 00:55:13,590
to be a really robust marker.

1121
00:55:13,590 --> 00:55:15,283
We haven't worked with it much before,

1122
00:55:15,283 --> 00:55:19,713
but it's a marker of kinda
inflammation fibrosis probably,

1123
00:55:20,790 --> 00:55:23,223
but it predicts a lot
of vascular outcomes.

1124
00:55:24,458 --> 00:55:26,850
There's Lp(a) you see and cystatin C,

1125
00:55:26,850 --> 00:55:29,610
which is kidney function, and IL-6.

1126
00:55:29,610 --> 00:55:32,460
So you can see the theme maybe emerging

1127
00:55:32,460 --> 00:55:35,010
from some of the other
findings for stroke overall.

1128
00:55:37,050 --> 00:55:41,640
What Sam did was essentially
developed a new model

1129
00:55:41,640 --> 00:55:42,570
for risk prediction.

1130
00:55:42,570 --> 00:55:45,630
So CHADSVASc is the name
of the score that we use

1131
00:55:45,630 --> 00:55:49,535
and he made a CHADSVASc biomarkers model

1132
00:55:49,535 --> 00:55:52,937
and showed that it was superior
to the CHADSVASc model.

1133
00:55:52,937 --> 00:55:56,292
So you can add improved risk
classification of people

1134
00:55:56,292 --> 00:55:59,940
by adding actually just
a couple of biomarkers.

1135
00:55:59,940 --> 00:56:02,724
NT-proBNP and GF-15 is all you need to add

1136
00:56:02,724 --> 00:56:05,190
and you get a better risk prediction.

1137
00:56:05,190 --> 00:56:09,150
So perhaps that is something
that could potentially lead

1138
00:56:09,150 --> 00:56:12,260
to practice change and how stroke risk

1139
00:56:12,260 --> 00:56:14,790
is assessed in AFib if it catches on.

1140
00:56:14,790 --> 00:56:17,550
You never know if these
things are gonna catch on.

1141
00:56:17,550 --> 00:56:20,340
So essentially we showed some biomarkers

1142
00:56:20,340 --> 00:56:22,620
underlie the effects of
risk factors on stroke.

1143
00:56:22,620 --> 00:56:25,320
I didn't show in CRP, but interluekin 6

1144
00:56:25,320 --> 00:56:28,170
and C-reactive protein are
other inflammatory marker.

1145
00:56:28,170 --> 00:56:31,500
Some appear to be race-specific
risk factors for stroke

1146
00:56:31,500 --> 00:56:33,123
or cognitive impairment.

1147
00:56:34,200 --> 00:56:37,500
Gluconic acid is the one
that the metabolomics work

1148
00:56:37,500 --> 00:56:39,355
came up with in that regard.

1149
00:56:39,355 --> 00:56:42,600
Some did not relate, which is, you know,

1150
00:56:42,600 --> 00:56:43,890
what you're gonna find.

1151
00:56:43,890 --> 00:56:45,753
Not all your hypotheses are correct.

1152
00:56:47,130 --> 00:56:50,220
And so the hypercoagulability
of sickle trait

1153
00:56:50,220 --> 00:56:52,683
is not enough to raise stroke risk.

1154
00:56:53,520 --> 00:56:56,370
And then some could suggest
possible new treatments

1155
00:56:56,370 --> 00:57:00,150
to improve brain health, and
the example with the AFib

1156
00:57:00,150 --> 00:57:02,450
that I showed you and
others that I mentioned.

1157
00:57:03,300 --> 00:57:05,670
So I'm giving a talk to this group.

1158
00:57:05,670 --> 00:57:07,380
So how can I talk not about

1159
00:57:07,380 --> 00:57:09,573
cerebral small vessel disease, right?

1160
00:57:11,700 --> 00:57:14,190
So Hyacinth and I and
Nels Olson just submitted

1161
00:57:14,190 --> 00:57:17,430
an R1, an ancillary study through REGARDS

1162
00:57:17,430 --> 00:57:21,030
to try to untangle the pathways

1163
00:57:21,030 --> 00:57:23,160
of cerebral small vessel
disease with stroke

1164
00:57:23,160 --> 00:57:25,680
and cognitive impairment in REGARDS.

1165
00:57:25,680 --> 00:57:30,363
So Hyacinth has a grant
now. This is so cool.

1166
00:57:31,590 --> 00:57:34,500
And he didn't talk about this
in his talk when he came here.

1167
00:57:34,500 --> 00:57:38,520
He's collecting MRIs from
all of our participants

1168
00:57:38,520 --> 00:57:40,890
from the hospitals where
they were treated for stroke

1169
00:57:40,890 --> 00:57:45,570
or evaluated for TIA and
he's rereading those MRIs

1170
00:57:45,570 --> 00:57:48,030
for cerebral small
vessel disease phenotype.

1171
00:57:48,030 --> 00:57:49,230
It's so cool.

1172
00:57:49,230 --> 00:57:51,240
He's got about 1,200 and some in hand

1173
00:57:51,240 --> 00:57:53,123
and he's expecting to have 1,500.

1174
00:57:54,110 --> 00:57:56,280
And there's a lot of
interesting things about this.

1175
00:57:56,280 --> 00:57:59,310
The MRIs it turns out
from the smaller hospitals

1176
00:57:59,310 --> 00:58:01,800
tend to be better quality,
(laughs) which I found

1177
00:58:01,800 --> 00:58:04,320
like so weird.

1178
00:58:04,320 --> 00:58:06,120
But anyway, he's characterizing that

1179
00:58:06,120 --> 00:58:09,870
so we can start really
getting at the epidemiology

1180
00:58:09,870 --> 00:58:11,850
of small vessel disease.

1181
00:58:11,850 --> 00:58:14,580
It's not like we have MRIs
on everybody at baseline,

1182
00:58:14,580 --> 00:58:15,990
but it's better than nothing.

1183
00:58:15,990 --> 00:58:18,810
We can't afford to do MRIs
on everybody in REGARDS,

1184
00:58:18,810 --> 00:58:20,400
and it would be a logistical nightmare,

1185
00:58:20,400 --> 00:58:21,930
'cause we have no clinics.

1186
00:58:21,930 --> 00:58:26,520
But at any rate, what we're
gonna do in the proposed work,

1187
00:58:26,520 --> 00:58:30,870
which we just submitted, is
to look at the proteomics

1188
00:58:30,870 --> 00:58:34,920
of cerebral small vessel
disease using the latest,

1189
00:58:34,920 --> 00:58:39,920
newest Olink platform which
measures over 5,000 protein.

1190
00:58:40,500 --> 00:58:42,250
This will be done down at the Broad

1191
00:58:43,470 --> 00:58:45,090
and we'll relate those proteins

1192
00:58:45,090 --> 00:58:47,760
to cerebral small vessel
disease phenotypes,

1193
00:58:47,760 --> 00:58:52,760
all different kinds, and
then to ADRD and also stroke.

1194
00:58:54,720 --> 00:58:57,660
But basically post-stroke
cognitive impairment,

1195
00:58:57,660 --> 00:58:59,610
because even after people have the stroke,

1196
00:58:59,610 --> 00:59:00,963
we're still following them.

1197
00:59:01,920 --> 00:59:06,920
And use the imaging and proteomic
data to try to understand

1198
00:59:08,790 --> 00:59:10,950
or prove our understanding
of the racial disparity,

1199
00:59:10,950 --> 00:59:12,840
which gets to your point that you asked

1200
00:59:12,840 --> 00:59:15,240
about would you throw all these things

1201
00:59:15,240 --> 00:59:17,160
in the same model or whatever.

1202
00:59:17,160 --> 00:59:18,810
And we can use machine learning

1203
00:59:18,810 --> 00:59:20,220
and all kinds of different approaches

1204
00:59:20,220 --> 00:59:23,504
once we have this data
to try to understand

1205
00:59:23,504 --> 00:59:28,504
like pathways and stuff
with 5,000 protein.

1206
00:59:29,730 --> 00:59:33,960
So here's some of the images
from actual participants.

1207
00:59:33,960 --> 00:59:36,360
So this is one participant.
This is another.

1208
00:59:36,360 --> 00:59:38,100
They're not really easy to see,

1209
00:59:38,100 --> 00:59:42,180
but at the top you see the yellow

1210
00:59:42,180 --> 00:59:45,240
is white matter hyperintensities,

1211
00:59:45,240 --> 00:59:47,373
and this participant didn't have any.

1212
00:59:49,260 --> 00:59:53,200
We have here

1213
00:59:55,770 --> 00:59:58,530
three examples of three types of cerebral

1214
00:59:58,530 --> 01:00:01,290
small vessel disease that
are being characterized.

1215
01:00:01,290 --> 01:00:03,870
So here we have deep and periventricular

1216
01:00:03,870 --> 01:00:05,940
white matter hyperintensities.

1217
01:00:05,940 --> 01:00:10,230
Here we have basal ganglia,
enlarged perivascular spaces.

1218
01:00:10,230 --> 01:00:14,700
And here we have multiple
cerebral microbleeds,

1219
01:00:14,700 --> 01:00:17,697
which can be due to
cerebral amyloid angiopathy

1220
01:00:17,697 --> 01:00:20,430
or hypertensive encephalopathy.

1221
01:00:20,430 --> 01:00:24,390
So really important phenotyping.

1222
01:00:24,390 --> 01:00:25,650
- [Crowd Member] Is this also based...

1223
01:00:25,650 --> 01:00:27,600
Okay, sorry.

1224
01:00:27,600 --> 01:00:29,700
- Yeah. (laughs) I gotcha.

1225
01:00:29,700 --> 01:00:32,790
So we generated a bunch of
preliminary data on the grant,

1226
01:00:32,790 --> 01:00:37,470
and this is based on the first
1,000 or so MRIs that he got.

1227
01:00:37,470 --> 01:00:42,470
And what you see is that
every region of the brain

1228
01:00:44,400 --> 01:00:49,400
for this, this is for white
matter hyperintensity volume,

1229
01:00:50,700 --> 01:00:53,603
the amount of white matter
hyperintensities that you have,

1230
01:00:53,603 --> 01:00:57,300
you can see there were
all these differences.

1231
01:00:57,300 --> 01:00:59,460
They're all unfavorable in Black

1232
01:00:59,460 --> 01:01:01,140
compared to white individuals.

1233
01:01:01,140 --> 01:01:02,940
There were two regions, I took 'em off

1234
01:01:02,940 --> 01:01:05,760
'cause the slide was so busy, cerebellum

1235
01:01:05,760 --> 01:01:08,340
and one other region where
there wasn't a difference.

1236
01:01:08,340 --> 01:01:12,330
But all of these areas had differences,

1237
01:01:12,330 --> 01:01:14,820
sometimes very large differences

1238
01:01:14,820 --> 01:01:16,670
in terms of statistical significance.

1239
01:01:18,960 --> 01:01:22,440
And then we said okay, well,
we can show associations

1240
01:01:22,440 --> 01:01:24,300
of biomarkers with these things.

1241
01:01:24,300 --> 01:01:26,880
So this is just in 168 participants

1242
01:01:26,880 --> 01:01:30,540
who we had some lab data
in and risk factor data in

1243
01:01:30,540 --> 01:01:31,950
from that case cohort studies.

1244
01:01:31,950 --> 01:01:34,337
Remember, we had 650 strokes.

1245
01:01:34,337 --> 01:01:37,980
And so so far with the
numbers that we have

1246
01:01:37,980 --> 01:01:41,850
the full characterization
of small vessel disease,

1247
01:01:41,850 --> 01:01:45,630
we had 168 participants with lab data.

1248
01:01:45,630 --> 01:01:48,780
And this is shown here, and you can see

1249
01:01:48,780 --> 01:01:53,040
some of the biomarkers
relate to different types

1250
01:01:53,040 --> 01:01:54,240
of cerebral small vessel disease.

1251
01:01:54,240 --> 01:01:57,161
So basal ganglia in large
cardiovascular spaces.

1252
01:01:57,161 --> 01:01:58,290
We got D-dimer.

1253
01:01:58,290 --> 01:02:01,440
For some reason factor XI
is inversely associated.

1254
01:02:01,440 --> 01:02:04,770
It's probably just advanced
minding. (chuckles)

1255
01:02:04,770 --> 01:02:09,210
But HGF has a growth
factor strongly related

1256
01:02:09,210 --> 01:02:13,807
to cerebral microbleeds.

1257
01:02:13,807 --> 01:02:17,760
- [Crowd Member] So all of
these individuals are Black or-

1258
01:02:17,760 --> 01:02:18,593
- No.

1259
01:02:18,593 --> 01:02:20,790
- [Crowd Member] The 168
still includes both parties?

1260
01:02:20,790 --> 01:02:21,623
- Yeah, yeah.

1261
01:02:21,623 --> 01:02:23,700
I mean, there's not enough
ability to look at it,

1262
01:02:23,700 --> 01:02:27,733
but once we have everybody, 1,500 MRIs,

1263
01:02:27,733 --> 01:02:32,640
1,500 assays for 5,000 proteins,

1264
01:02:32,640 --> 01:02:36,767
then, you know, we'll be able
to pick apart more detail.

1265
01:02:36,767 --> 01:02:40,080
But look, you know, hypertension,
we all know this, right?

1266
01:02:40,080 --> 01:02:45,080
Basal ganglia, 25 or greater
risk of basal ganglia.

1267
01:02:45,090 --> 01:02:47,850
That hypertension is the
cause of that. We know that.

1268
01:02:47,850 --> 01:02:50,220
So it's really just good proof principle

1269
01:02:50,220 --> 01:02:52,020
that we can see these relationships.

1270
01:02:55,110 --> 01:02:59,400
So hopefully we get the score.

1271
01:02:59,400 --> 01:03:02,520
I like the grant, but
just 'cause I like it

1272
01:03:02,520 --> 01:03:04,760
doesn't mean the study's
actually gonna like it,

1273
01:03:04,760 --> 01:03:06,600
as we were talking about earlier.

1274
01:03:06,600 --> 01:03:09,990
So you know, I think we've
identified some actionable steps

1275
01:03:09,990 --> 01:03:13,683
to try to improve the health
of Black people in the U.S.

1276
01:03:14,970 --> 01:03:16,830
Always we hope that clinical trials

1277
01:03:16,830 --> 01:03:20,040
will represent those people appropriately.

1278
01:03:20,040 --> 01:03:22,470
We learned in COVID we could do it.

1279
01:03:22,470 --> 01:03:24,240
If you're interested in
that, we just published

1280
01:03:24,240 --> 01:03:28,079
in "NEJM Evidence" our
experience in the COVID trials,

1281
01:03:28,079 --> 01:03:29,910
'cause over half of our participants

1282
01:03:29,910 --> 01:03:31,803
were underrepresented minority.

1283
01:03:31,803 --> 01:03:33,660
But we know we can do that.

1284
01:03:33,660 --> 01:03:37,170
We have a poor history
of that in this country.

1285
01:03:37,170 --> 01:03:38,940
I bet you the people
doing those Lp(a) trials

1286
01:03:38,940 --> 01:03:41,610
have no idea of our data, right?

1287
01:03:41,610 --> 01:03:43,710
Or probably aren't even thinking about it.

1288
01:03:45,240 --> 01:03:47,280
And then there's so much ongoing research.

1289
01:03:47,280 --> 01:03:48,630
You know, Minaz is here.

1290
01:03:48,630 --> 01:03:52,950
She's studying the intersectionality

1291
01:03:52,950 --> 01:03:57,070
of social factors, race, sex, and class

1292
01:03:57,990 --> 01:03:59,193
with risk factors.

1293
01:04:01,800 --> 01:04:04,830
We have Nels Olson,
just got an R01 to study

1294
01:04:04,830 --> 01:04:07,976
the proteomics of cognitive impairment

1295
01:04:07,976 --> 01:04:09,270
where he probably is
gonna use dementia death

1296
01:04:09,270 --> 01:04:12,333
as the outcome, 'cause it's
the hardest outcome we have.

1297
01:04:13,440 --> 01:04:16,590
It's more strongly related
to all the risk factors

1298
01:04:16,590 --> 01:04:19,053
than our cognitive
impairment type measures.

1299
01:04:20,640 --> 01:04:22,980
There's more going on
in AFib and the effects

1300
01:04:22,980 --> 01:04:25,170
on other aspects of health.

1301
01:04:25,170 --> 01:04:27,930
Andres Cordova, who's a cardiology fellow,

1302
01:04:27,930 --> 01:04:31,287
is writing about
biomarkers that predict MI

1303
01:04:31,287 --> 01:04:33,720
in people with AFib, 'cause we know AFib

1304
01:04:33,720 --> 01:04:36,600
is a risk factor for MI and
we don't understand why.

1305
01:04:36,600 --> 01:04:39,090
Vinh Le, our postdoc, is looking at AFib

1306
01:04:39,090 --> 01:04:41,460
and the risk of cognitive impairment

1307
01:04:41,460 --> 01:04:44,880
and biomarkers that might
relate to cognitive impairment

1308
01:04:44,880 --> 01:04:46,470
just in people with AFib.

1309
01:04:46,470 --> 01:04:48,780
He's making great progress
during his one year

1310
01:04:48,780 --> 01:04:49,893
that he has with us.

1311
01:04:51,150 --> 01:04:55,230
And we submitted our renewal for REGARDS

1312
01:04:55,230 --> 01:04:58,200
the second time in 2023.

1313
01:04:58,200 --> 01:05:00,870
We got a bad study
section and a bad review.

1314
01:05:00,870 --> 01:05:02,280
Basically submitted the same grant

1315
01:05:02,280 --> 01:05:05,080
to a different study section
and we got a perfect score.

1316
01:05:06,120 --> 01:05:09,720
Just goes to show where you
send it is really important,

1317
01:05:09,720 --> 01:05:10,590
where it gets assigned.

1318
01:05:10,590 --> 01:05:12,120
Pay attention. (laughs)

1319
01:05:12,120 --> 01:05:14,250
I mean, it was in a SEP first time

1320
01:05:14,250 --> 01:05:15,930
'cause of conflict
actually, 'cause Hyacinth

1321
01:05:15,930 --> 01:05:19,080
was on the study section we usually go to.

1322
01:05:19,080 --> 01:05:21,840
And so it was in a SEP,
and there was not even

1323
01:05:21,840 --> 01:05:24,783
a vascular neurologist in the panel.

1324
01:05:26,460 --> 01:05:28,200
It was just ridiculous.

1325
01:05:28,200 --> 01:05:31,440
So it was torture and it
really caused a lot of delays

1326
01:05:31,440 --> 01:05:36,090
and really like picking
ourselves up by the bootstraps

1327
01:05:36,090 --> 01:05:39,090
and tightening up our budgets
to cover no-cost extension

1328
01:05:39,090 --> 01:05:41,790
and keep our people employed.

1329
01:05:41,790 --> 01:05:44,073
But what we're doing
here, this is the title.

1330
01:05:45,789 --> 01:05:50,253
This is what's going on since
the 60s in the United States.

1331
01:05:51,360 --> 01:05:55,680
This is the Black/white
stroke mortality disparity

1332
01:05:55,680 --> 01:06:00,680
getting larger, and since about 2013,

1333
01:06:03,270 --> 01:06:07,357
everyone is having an increasing rate

1334
01:06:08,580 --> 01:06:10,740
of stroke mortality.

1335
01:06:10,740 --> 01:06:12,570
It's rising.

1336
01:06:12,570 --> 01:06:13,560
You know, we've had 20 years

1337
01:06:13,560 --> 01:06:15,030
of declining in cardiovascular disease.

1338
01:06:15,030 --> 01:06:16,110
Now stroke is rising.

1339
01:06:16,110 --> 01:06:19,320
Not CHD. We don't know why.

1340
01:06:19,320 --> 01:06:24,320
So we want to understand
risk factor differences

1341
01:06:25,740 --> 01:06:28,980
in birth cohorts over
time, so we're gonna enroll

1342
01:06:28,980 --> 01:06:31,530
12,000 new people in REGARDS.

1343
01:06:31,530 --> 01:06:33,900
Gets to one of Cal's questions.

1344
01:06:33,900 --> 01:06:36,750
And they're gonna be middle
aged people, Black and white.

1345
01:06:36,750 --> 01:06:38,760
We're gonna compare
their respective profiles

1346
01:06:38,760 --> 01:06:42,018
to the original cohort of the same age

1347
01:06:42,018 --> 01:06:44,190
to try to understand what's happening.

1348
01:06:44,190 --> 01:06:47,340
We know the risk factors are
probably gonna be worse now

1349
01:06:47,340 --> 01:06:49,980
with the obesity epidemic
affecting these people

1350
01:06:49,980 --> 01:06:54,980
when they young adults, and so
we'll try to figure that out

1351
01:06:56,040 --> 01:06:59,150
and then understand how those differences

1352
01:06:59,150 --> 01:07:03,510
in risk factors might explain
this rising stroke risk.

1353
01:07:03,510 --> 01:07:07,140
We're also gonna do a lot on
Alzheimer's disease biomarkers

1354
01:07:07,140 --> 01:07:09,300
and circulation stuff to kind of really

1355
01:07:09,300 --> 01:07:10,773
get at their epidemiology.

1356
01:07:11,760 --> 01:07:14,340
So it's a really exciting time for us,

1357
01:07:14,340 --> 01:07:15,993
and we're gonna start in April.

1358
01:07:17,430 --> 01:07:20,610
So tons of people to acknowledge.

1359
01:07:20,610 --> 01:07:22,804
You can't fit them all
on one slide really.

1360
01:07:22,804 --> 01:07:26,433
But if you have other
questions, I know I'm over time,

1361
01:07:27,450 --> 01:07:29,790
but I'm happy to answer other
questions you might have.

1362
01:07:29,790 --> 01:07:33,300
And I wanna point out, you
know, Debora was postdoc,

1363
01:07:33,300 --> 01:07:36,600
is now faculty and funded by our center

1364
01:07:36,600 --> 01:07:41,250
to do her work on
intersectionality of social factors

1365
01:07:41,250 --> 01:07:43,620
and how that relates to
cardiovascular health

1366
01:07:43,620 --> 01:07:47,520
and what the proteomic profiles
are a biological embodiment

1367
01:07:47,520 --> 01:07:51,333
of those social injustices
in the population.

1368
01:07:52,300 --> 01:07:54,660
Minaz is her postdoc if
you haven't met her yet.

1369
01:07:54,660 --> 01:07:59,430
And Debora's up at AHA, so not here today.

1370
01:07:59,430 --> 01:08:01,530
Anyway, free to take questions.

1371
01:08:01,530 --> 01:08:03,263
- [Crowd Member] I'll
slip in a quick question.

1372
01:08:06,000 --> 01:08:08,550
So you talked about incentives
to these participants.

1373
01:08:08,550 --> 01:08:09,383
- Yes.

1374
01:08:12,390 --> 01:08:14,063
- [Crowd Member] One of
the incentives would be

1375
01:08:14,063 --> 01:08:16,920
to give a forward
prediction based on their

1376
01:08:16,920 --> 01:08:18,589
risk factors and biomarkers,
how long they're gonna live or-

1377
01:08:18,589 --> 01:08:19,422
- Oh.

1378
01:08:20,487 --> 01:08:25,320
Yeah, I mean, at the baseline,
they get their results back.

1379
01:08:25,320 --> 01:08:27,930
So they get their EKG,
their lipid profile,

1380
01:08:27,930 --> 01:08:31,980
their glucose, creatinine,
hemoglobin, A1C, blood pressure.

1381
01:08:31,980 --> 01:08:32,820
They get all that back.

1382
01:08:32,820 --> 01:08:34,710
- Right.
- So they get our results.

1383
01:08:34,710 --> 01:08:37,770
So when the examiner is in the home,

1384
01:08:37,770 --> 01:08:38,880
there's a little card they fill out

1385
01:08:38,880 --> 01:08:40,560
with their blood pressure and stuff.

1386
01:08:40,560 --> 01:08:43,470
And then we send a results
letter with the lab data

1387
01:08:43,470 --> 01:08:47,580
and the ECG data, and
so they can act on that,

1388
01:08:47,580 --> 01:08:49,453
if they should choose.

1389
01:08:49,453 --> 01:08:50,957
We have to be careful though.

1390
01:08:50,957 --> 01:08:53,670
- [Crowd Member] I know it's
probably politically incorrect,

1391
01:08:53,670 --> 01:08:55,920
but all your risk factors, if
you assemble them together,

1392
01:08:55,920 --> 01:08:58,170
you would get a forward prediction model

1393
01:08:58,170 --> 01:09:00,221
of their likelihood of getting a stroke

1394
01:09:00,221 --> 01:09:01,500
in the next 10 years.

1395
01:09:01,500 --> 01:09:02,940
That would be useful information.

1396
01:09:02,940 --> 01:09:05,700
If they're like, oh, I'm gonna die at 60,

1397
01:09:05,700 --> 01:09:07,290
give or take three years, it might change

1398
01:09:07,290 --> 01:09:09,060
the way they live their life.

1399
01:09:09,060 --> 01:09:12,090
- Yeah, so there are standard
risk prediction scores

1400
01:09:12,090 --> 01:09:15,210
that are used in practice,
like the whole cohort equation,

1401
01:09:15,210 --> 01:09:18,990
the ASCD risk score, which is used

1402
01:09:18,990 --> 01:09:21,450
for statin prescription.

1403
01:09:21,450 --> 01:09:24,000
There's Life's Simple 7
and now Life's Simple 8

1404
01:09:24,000 --> 01:09:26,340
which are health assessments
that can help people

1405
01:09:26,340 --> 01:09:28,713
identify areas they need to change.

1406
01:09:29,700 --> 01:09:31,714
We have to be a little
careful about doing care

1407
01:09:31,714 --> 01:09:35,010
for the people, because
they can take that report

1408
01:09:35,010 --> 01:09:36,734
to their physician.

1409
01:09:36,734 --> 01:09:38,340
Everything on that report is stuff

1410
01:09:38,340 --> 01:09:40,740
they should be having anyway pretty much

1411
01:09:40,740 --> 01:09:42,420
in the course of their primary care.

1412
01:09:42,420 --> 01:09:45,120
We have to be a little careful not to like

1413
01:09:45,120 --> 01:09:48,493
change their life too much, right?

1414
01:09:48,493 --> 01:09:52,056
'Cause if you do too
much of an intervention,

1415
01:09:52,056 --> 01:09:54,690
then you're affecting
potentially the outcomes

1416
01:09:54,690 --> 01:09:55,620
that we're interested in.

1417
01:09:55,620 --> 01:09:57,780
This is the observational research.

1418
01:09:57,780 --> 01:10:00,000
A great example, when we
were starting this study,

1419
01:10:00,000 --> 01:10:02,400
the Multi-Ethnic Study
of Atherosclerosis, MESA,

1420
01:10:02,400 --> 01:10:04,140
started a couple years before REGARDS,

1421
01:10:04,140 --> 01:10:07,230
and I was really deeply
engaged in the design of MESA.

1422
01:10:07,230 --> 01:10:09,690
Russ Tracy was my mentor at the time

1423
01:10:09,690 --> 01:10:11,640
and he was on the steering committee.

1424
01:10:11,640 --> 01:10:15,690
And MESA was a study of racial differences

1425
01:10:15,690 --> 01:10:19,170
across four groups in subclinical
atherosclerosis measure

1426
01:10:19,170 --> 01:10:21,120
and how they relate to clinical outcomes.

1427
01:10:21,120 --> 01:10:24,300
They were all people who never
had cardiovascular disease

1428
01:10:24,300 --> 01:10:26,580
and they had high-tech imaging
for the different types

1429
01:10:26,580 --> 01:10:28,650
of subclinical disease.

1430
01:10:28,650 --> 01:10:32,210
So coronary calcium was the hot
new thing at the time by CT.

1431
01:10:32,210 --> 01:10:35,160
If you detect coronary artery calcium,

1432
01:10:35,160 --> 01:10:37,110
that's atherosclerosis, right?

1433
01:10:37,110 --> 01:10:40,500
And so we wanted to understand
the epidemiology of that

1434
01:10:40,500 --> 01:10:42,720
and the prevalence of it
and the way it relates

1435
01:10:42,720 --> 01:10:46,440
to heart disease, stroke,
and other outcomes.

1436
01:10:46,440 --> 01:10:48,810
So there was a debate
during the design phase,

1437
01:10:48,810 --> 01:10:51,870
'cause these are NHLBI contracts.

1438
01:10:51,870 --> 01:10:55,290
We all apply to be the
sites and they pick us

1439
01:10:55,290 --> 01:10:56,760
and then we design the study.

1440
01:10:56,760 --> 01:11:00,000
So we're following loose
parameters of what they want,

1441
01:11:00,000 --> 01:11:03,545
but we're designing the study
how we see it should be.

1442
01:11:03,545 --> 01:11:05,660
And the steering committee
had a lot of debate

1443
01:11:05,660 --> 01:11:10,660
on the calcium, whether to
give people their scores,

1444
01:11:10,800 --> 01:11:13,260
should we randomize people
to getting their scores,

1445
01:11:13,260 --> 01:11:16,290
you know, what we were
gonna do with that data.

1446
01:11:16,290 --> 01:11:19,590
And I love the idea of
randomizing people to getting

1447
01:11:19,590 --> 01:11:22,620
the score or not, because
at the follow-up visits,

1448
01:11:22,620 --> 01:11:24,360
we could then say if they did anything

1449
01:11:24,360 --> 01:11:25,710
that improved their health.

1450
01:11:26,670 --> 01:11:29,580
There was a study done
in military recruits

1451
01:11:29,580 --> 01:11:32,970
that did that actually and showed them

1452
01:11:32,970 --> 01:11:37,710
the pictures of their calcium
and they had no effect, right?

1453
01:11:37,710 --> 01:11:40,920
So we debated doing this
as an embedded thing

1454
01:11:40,920 --> 01:11:43,189
and there was a lot of
debate and discussion.

1455
01:11:43,189 --> 01:11:44,027
In the end it was decided
just give 'em the score

1456
01:11:44,027 --> 01:11:47,497
and you know, tell 'em,
"See your doctor with this."

1457
01:11:47,497 --> 01:11:50,010
Of course at the time, primary care docs,

1458
01:11:50,010 --> 01:11:51,510
they didn't know what to
do with that information.

1459
01:11:51,510 --> 01:11:54,720
Now we know if we have
calcium and burden of calcium

1460
01:11:54,720 --> 01:11:58,047
is really strongly related
to heart-attack risk.

1461
01:11:58,047 --> 01:12:01,350
Obviously it's the
disease, but at the time

1462
01:12:01,350 --> 01:12:03,780
we didn't really know how
useful that test would be

1463
01:12:03,780 --> 01:12:04,743
as a clinical test.

1464
01:12:06,750 --> 01:12:09,600
So you can do things, but you
have to be a little cautious

1465
01:12:09,600 --> 01:12:10,700
about how much you do.

1466
01:12:12,990 --> 01:12:14,970
But you know, getting
their results letters,

1467
01:12:14,970 --> 01:12:16,570
most people are happy with that.

1468
01:12:17,910 --> 01:12:22,620
And our participants, they
just love to talk to us.

1469
01:12:22,620 --> 01:12:23,550
They send notes.

1470
01:12:23,550 --> 01:12:28,200
We occasionally get notes
in the boxes. (laughs)

1471
01:12:28,200 --> 01:12:31,529
We did a dried blood spot
collection during the pandemic

1472
01:12:31,529 --> 01:12:35,662
and we sent kits out to
people to collect blood,

1473
01:12:35,662 --> 01:12:36,840
put it on filter paper,
and mail it back to us

1474
01:12:36,840 --> 01:12:38,790
so we could do COVID serology.

1475
01:12:38,790 --> 01:12:41,160
We did it in 14 of these big studies.

1476
01:12:41,160 --> 01:12:45,660
We have 20,000-plus
blood spots from people.

1477
01:12:45,660 --> 01:12:47,947
And some people would send notes.

1478
01:12:47,947 --> 01:12:49,860
"Thank you so much for
the research you're doing,

1479
01:12:49,860 --> 01:12:51,003
blah, blah, blah," or whatever.

1480
01:12:51,003 --> 01:12:52,113
It's so funny.

1481
01:12:54,351 --> 01:12:55,620
But these are people who join.

1482
01:12:55,620 --> 01:12:59,080
So does that make sense?

1483
01:12:59,080 --> 01:13:01,140
- [Crowd Member] I was
actually thinking too,

1484
01:13:01,140 --> 01:13:03,290
could you cross reference
your geographical

1485
01:13:04,290 --> 01:13:07,950
risk associations with
cost of life insurance

1486
01:13:07,950 --> 01:13:09,780
and whether the health insurance companies

1487
01:13:09,780 --> 01:13:11,610
are tracking the same risk factors?

1488
01:13:11,610 --> 01:13:13,470
- The longevity and stuff?

1489
01:13:13,470 --> 01:13:15,360
Yeah, I don't think we've done that.

1490
01:13:15,360 --> 01:13:19,410
We've looked at, you know,
accessibility to care

1491
01:13:19,410 --> 01:13:21,540
and insurance status and things like that.

1492
01:13:21,540 --> 01:13:24,510
Of course, our population,
half of them are over 65,

1493
01:13:24,510 --> 01:13:27,753
so they're on Medicare, have
basic insurance at least.

1494
01:13:28,920 --> 01:13:31,860
But we ask questions
about all kinds of things.

1495
01:13:31,860 --> 01:13:34,118
There's a questionnaire
called You and Your Doctor

1496
01:13:34,118 --> 01:13:36,610
and it's getting at do the people trust

1497
01:13:37,464 --> 01:13:39,090
their healthcare teams?

1498
01:13:39,090 --> 01:13:40,410
How do they feel about what

1499
01:13:40,410 --> 01:13:41,550
their healthcare teams tell them?

1500
01:13:41,550 --> 01:13:43,535
What's their relationship like?

1501
01:13:43,535 --> 01:13:45,532
There's so many different things.

1502
01:13:45,532 --> 01:13:50,532
We wanted to track tobacco
taxation with smoking

1503
01:13:50,730 --> 01:13:55,170
and health outcomes to
see if we could tease out

1504
01:13:55,170 --> 01:13:59,910
the impact of tobacco taxation
on health as an example,

1505
01:13:59,910 --> 01:14:02,550
because in the south, there's
like no tobacco taxes.

1506
01:14:02,550 --> 01:14:03,660
They're minimal.

1507
01:14:03,660 --> 01:14:07,980
But when you are in basically
the blue states, you know,

1508
01:14:07,980 --> 01:14:11,972
your pack of cigarettes is
three, four, five times greater

1509
01:14:11,972 --> 01:14:16,350
than it is in the south
or on Indian reservations

1510
01:14:16,350 --> 01:14:18,390
where they don't have tobacco tax.

1511
01:14:18,390 --> 01:14:21,030
So we tried to look at that.
It didn't really pan out.

1512
01:14:21,030 --> 01:14:24,480
We tried to look at sugary
beverage taxes as well,

1513
01:14:24,480 --> 01:14:26,103
or sugary beverage intake.

1514
01:14:28,380 --> 01:14:30,480
It's hard to do some of
these policy questions

1515
01:14:30,480 --> 01:14:35,480
in such a small sample, because
it is only 30,000 people

1516
01:14:35,490 --> 01:14:36,940
scattered all over the place.

1517
01:14:38,520 --> 01:14:41,043
But yeah, so it's endless
the things you can do.

1518
01:14:43,650 --> 01:14:45,090
And we can link to other data.

1519
01:14:45,090 --> 01:14:48,810
So there's linkage to Medicare claims data

1520
01:14:48,810 --> 01:14:50,520
for the people on Medicare, for example,

1521
01:14:50,520 --> 01:14:54,690
and that's used by certain
groups to do specific things

1522
01:14:54,690 --> 01:14:58,877
because it gives richer outcomes
data for certain things.

1523
01:15:10,840 --> 01:15:12,910
- [Crowd Member] It's
like a mix, you know,

1524
01:15:12,910 --> 01:15:13,897
so I reckon that'd influence-

1525
01:15:13,897 --> 01:15:15,056
- [Crowd Member] Yeah, no, I figured.

1526
01:15:15,056 --> 01:15:16,503
- Oh, so Sully's asking...

1527
01:15:16,503 --> 01:15:17,400
- [Crowd Member] It's stuck in my head.

1528
01:15:17,400 --> 01:15:19,660
- Can you hear me, Sully?
Are you still there?

1529
01:15:19,660 --> 01:15:20,743
- [Crowd Member] That changed.

1530
01:15:20,743 --> 01:15:21,576
- [Crowd Member] How is that?

1531
01:15:21,576 --> 01:15:23,220
- [Sully] Yes, I can hear you, Mary.

1532
01:15:24,270 --> 01:15:29,270
- So fatty liver disease,
we use something called

1533
01:15:29,352 --> 01:15:31,688
the fatty liver index.

1534
01:15:31,688 --> 01:15:34,355
It's calculated based on lab tests.

1535
01:15:34,355 --> 01:15:38,010
So we don't have imaging deliver.

1536
01:15:38,010 --> 01:15:39,053
- [Crowd Member] Okay, yeah.

1537
01:15:41,288 --> 01:15:44,490
- So we use the fatty liver
index, which is not a bad proxy.

1538
01:15:44,490 --> 01:15:45,510
Not great.

1539
01:15:45,510 --> 01:15:47,880
There was some resistance
in peer review to it,

1540
01:15:47,880 --> 01:15:49,960
but that's what we use, yeah.

1541
01:15:49,960 --> 01:15:52,140
- [Sully] Cool. Thanks, Mary.

1542
01:15:52,140 --> 01:15:54,601
- So we have that in that case folder.

1543
01:15:54,601 --> 01:15:56,907
- [Crowd Member] Yeah. Thanks.

1544
01:15:56,907 --> 01:15:57,777
- [Crowd Member] Yeah.

1545
01:15:57,777 --> 01:15:58,859
- [Crowd Member] How is it going?

1546
01:15:58,859 --> 01:16:01,544
- [Crowd Member] That's the question.

1547
01:16:01,544 --> 01:16:02,377
- [Crowd Member] It's going well.

1548
01:16:02,377 --> 01:16:03,210
- [Crowd Member] We'll see, yeah.

1549
01:16:03,210 --> 01:16:04,770
- [Crowd Member] Yeah, it's fine.

1550
01:16:04,770 --> 01:16:05,633
- [Crowd Member] Yeah.

1551
01:16:06,930 --> 01:16:09,990
- Yeah, Sully, I don't know
if you were trying to speak,

1552
01:16:09,990 --> 01:16:11,070
but I think what's happening here

1553
01:16:11,070 --> 01:16:12,933
was not turned on correctly.

1554
01:16:13,770 --> 01:16:16,320
Were you trying to say something?

1555
01:16:16,320 --> 01:16:18,210
- [Sully] Oh, I was just saying thank you.

1556
01:16:18,210 --> 01:16:20,060
That's good to know, Mary. Thank you.

1557
01:16:21,450 --> 01:16:23,010
- Yeah, I can barely hear you.

1558
01:16:23,010 --> 01:16:23,843
The sound-

1559
01:16:23,843 --> 01:16:25,403
- [Crowd Member] Yeah, yeah, exactly.

1560
01:16:27,150 --> 01:16:28,100
- Can we try again?

1561
01:16:29,130 --> 01:16:30,903
To see, just so we can test it.

1562
01:16:35,370 --> 01:16:36,330
Sully, can you hear?

1563
01:16:36,330 --> 01:16:37,569
- [Crowd Member] And so we have to-

1564
01:16:37,569 --> 01:16:39,537
- [Sully] Yep, I can hear
you good. Is that better?

1565
01:16:39,537 --> 01:16:41,190
- The sound wasn't on them.

1566
01:16:41,190 --> 01:16:43,100
The mic, you should...

1567
01:16:44,100 --> 01:16:46,050
I'm just gonna show Sierra the problem.

1568
01:16:47,220 --> 01:16:48,630
This said headphones.

1569
01:16:48,630 --> 01:16:50,262
- [Crowd Member] Ugh.

1570
01:16:50,262 --> 01:16:51,095
- [Crowd Member] That's crazy.

1571
01:16:51,095 --> 01:16:53,840
- So I just changed it to the same.

1572
01:16:53,840 --> 01:16:55,465
Turned the sound up.

1573
01:16:55,465 --> 01:16:56,298
Not that they can...

1574
01:16:56,298 --> 01:16:57,420
I think nobody was trying to speak there.

1575
01:16:57,420 --> 01:16:58,440
- [Crowd Member] Right.

1576
01:16:58,440 --> 01:17:00,737
I'm just gonna complete that
one now. Thank you, though.

1577
01:17:00,737 --> 01:17:04,710
- Sully, I'm sorry I didn't
go into the chat right away.

1578
01:17:04,710 --> 01:17:06,060
- [Sully] Oh, no, it's cool.

1579
01:17:06,060 --> 01:17:08,310
But thanks, Mary, thanks
for explaining that.

1580
01:17:08,310 --> 01:17:09,630
- All right, awesome.

1581
01:17:09,630 --> 01:17:11,603
I'm happy to talk to you about
it if you wanna look at it.

1582
01:17:11,603 --> 01:17:13,770
- [Sully] That'd be
great. Yeah, no, for sure.

1583
01:17:13,770 --> 01:17:15,057
That'd be great. Thank you.

1584
01:17:15,057 --> 01:17:16,920
- Yeah, FLI is kind of cool.

1585
01:17:16,920 --> 01:17:19,833
My postdoc tackled that one,

1586
01:17:21,030 --> 01:17:23,100
a postdoc that I had in the past.

1587
01:17:23,100 --> 01:17:24,900
- [Sully] The fatty liver disease?

1588
01:17:24,900 --> 01:17:26,220
- [Mary] Yeah, yeah.

1589
01:17:26,220 --> 01:17:27,053
- [Sully] Oh, cool.

1590
01:17:27,053 --> 01:17:27,886
- She worked on it.

1591
01:17:27,886 --> 01:17:30,510
The papers were a little
difficult to publish.

1592
01:17:30,510 --> 01:17:32,100
Actually, you know, there's a guy...

1593
01:17:32,100 --> 01:17:34,507
Oh, there's a paper.

1594
01:17:34,507 --> 01:17:39,120
There's a guy at Columbia
who I think did Thrive?

1595
01:17:39,120 --> 01:17:40,950
Maybe it was the fifth floor,

1596
01:17:40,950 --> 01:17:43,963
and he never finished the paper. (laughs)

1597
01:17:44,899 --> 01:17:46,623
But we need an author.

1598
01:17:47,820 --> 01:17:49,200
I think it was on stroke.

1599
01:17:49,200 --> 01:17:52,653
So if you know someone who
might be interested in that,

1600
01:17:53,731 --> 01:17:56,310
it's an unfinished project.

1601
01:17:56,310 --> 01:17:57,420
- [Sully] Yeah, that's cool.

1602
01:17:57,420 --> 01:17:59,520
We should chat for sure.

1603
01:17:59,520 --> 01:18:01,170
- Yeah, okay. Awesome.

1604
01:18:01,170 --> 01:18:02,070
Take care.

1605
01:18:02,070 --> 01:18:02,903
- [Sully] Thanks, Mary.

1606
01:18:02,903 --> 01:18:04,264
- Bye.
- Yeah, something with-

1607
01:18:04,264 --> 01:18:05,514
- [Crowd Member] Exactly.