Outline Ea+ng Disorders and related ED and DSM-5 behaviours in - - PDF document

26/04/17 1 Ea+ng Disorders and related behaviours in adolescence: results from popula+on-based studies

Nadia Micali

• Dept. of Psychiatry

Icahn School of Medicine at Mount Sinai New York, NY US & Ins+tute of Child Health University College London London UK

Outline

• ED and DSM-5
• Epidemiology
• ED
• ED behaviours
• Three studies:
• prevalence of ED
• Correlates and adverse outcomes

ED: the old and the new

• Anorexia Nervosa

amenorrhea

• Bulimia Nervosa

Minimum frequency: once a week

• Ea+ng Disorders not Otherwise specified
• Binge Ea+ng Disorder

Minimum frequency: once a week

OSFED

The Incidence of ED amongst females aged 10-49 in the UK

Micali et al., 2013

26/04/17 2 The Incidence of ED amongst males aged 10-49 in the UK

Micali et al., 2013

Incidence rates of ED in females by age-bands in 2009

0" 20" 40" 60" 80" 100" 120" 140" 160" 180" 10(14" 15(19" 20(29" 30(39" 40(49" AN" BN" EDNOS" ALL"ED"

IR per 100,000 Age bands IR=1.7/1,000

Prevalence of adolescent AN

l US Na+onal Comorbidity Survey US adolescents aged 13-18 years: life+me prevalence: 0.3%, 12-mo prevalence 0.2% (M=F) l Portugal: 0.4% in adolescent females (aged 12-23) l Finland: 2.2% adolescent girls aged 15-19 (life+me)

l Overall about 0.4-2%

Swanson et al., 2011; Keski-Rakhonen et al., 2007, Machado et al., 2007

Prevalence of adolescent BN

l US: life+me 0.9%; M= 0.5%; F=1.3% l Portugal: 0.3% in adolescent girls (aged 12-23) l Finland: 1.7% in females aged 15-19

l Overall ~ 1-2%

Swanson et al., 2011; Keski-Rakhonen et al., 2009, Machado et al., 2007

26/04/17 3

Prevalence of adolescent EDNOS

l Portugal: 2.4% in adolescent females l US:

• BED: life+me 1.6%; M=0.8, F=2.3%

l EDNOS most common ED: BED

Swanson et al., 2011; Machado et al., 2007

Ea+ng disorders behaviours in US adolescents

Ackard et al., 2007

ED and ED behaviours in adolescence: prevalence, correlates and adverse outcomes ALSPAC and GUTS

Field et al, Pediatrics, 2012 Field et al. JAMA Pediatrics, 2013 Micali et al, JAACAP, 2015

26/04/17 4

Growing Up Today Study (GUTS)

• Established in 1996
• Par+cipants are the offspring of women in the

Nurses Health Study II (NHS II)

• 9039 girls and 7843 boys

– 9-15 years at baseline

Follow-up

• Ques+onnaires were mailed annually un+l 2001,

then biennially

• Non-responders are sent a reminder e-mail and/
• r postcard, followed by a another ques+onnaire
• Children who do not respond aher these

measures have been taken are sent an abbreviated survey

Weight & Weight Concerns

• Self-reported weight and height collected on all

surveys

• Weight concerns measured with the McKnight

Risk Factor Survey (MRFS)

Binge Ea+ng

• Binge ea+ng:

– at least monthly/weekly episodes of ea+ng a large amount of food AND feeling out of control during the episodes (LOC)

• Overea&ng:

– at least monthly/weekly episodes of ea&ng a large amount of food, but NOT feeling out of control during the episodes (no LOC)

26/04/17 5

Purging

• During the past year, did you make yourself

throw up to lose weight or keep from gaining weight?

• During the past year, did you take laxa+ves to

lose weight or keep from gaining weight?

Ea+ng Disorder Classifica+on

• AN

– Underweight

• < 18 years

Age-specific cut-off predic+ng BMI < 18.5 at age 18

• > 18 years

BMI < 18.5

– High concerns with weight and shape

Ea+ng Disorder Classifica+on

• BN

– Binge eat with loss of control > weekly (DSM-5) >monthly – Purging > weekly (DSM-5) >monthly

Ea+ng Disorder Classifica+on

• AN
• BN
• BED

– Binge eat with LOC > weekly > weekly (DSM-5) > monthly – No or infrequent purging

26/04/17 6

Ea+ng Disorder Classifica+on

• AN
• BN
• BED
• Purging Disorder (PD)

– Purge > weekly > weekly (DSM-5) > monthly – No or infrequent binge ea+ng

Ea+ng Disorder Classifica+on

• AN
• BN
• BED
• Purging Disorder (PD)
• EDNOS

– Overeat, but no loss of control – Binge weekly (DSM-IV)/monthly (DSM-5) – Purge weekly (DSM-IV)/monthly (DSM-5)

Age-specific prevalence of ea+ng disorders among girls in GUTS

0% 1% 2% 3% 4% 5% 9-12 yrs 13-15 yrs 16-18 yrs 19-22 yrs 23-26 yrs AN BN

Age-specific prevalence of ea+ng disorders among girls in GUTS

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 9-12 yrs 13-15 yrs 16-18 yrs 19-22 yrs 23-26 yrs AN BN BED

26/04/17 7

Age-specific prevalence of ea+ng disorders among girls in GUTS

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 9-12 yrs 13-15 yrs 16-18 yrs 19-22 yrs 23-26 yrs AN BN BED Purging Disorder

Age-specific prevalence of ea+ng disorders among girls in GUTS

0% 5% 10% 15% 20% 25%

9-12 yrs 13-15 yrs 16-18 yrs 19-22 yrs 23-26 yrs

AN BN BED Purging Disorder EDNOS

Methods

• All analyses restricted to 8594 females with

follow-up data

• Lagged analysis
• Generalized es+ma+ng equa+ons with an

independence working covariance

Methods

• Becoming overweight

– Age, BMI, die+ng

• Star+ng to use drugs other than marijuana
• Star+ng to binge drink frequently
• Developing high depressive symptoms

26/04/17 8

Methods

• Becoming overweight
• Star+ng to use drugs

– Age, sib who used drugs, sib who was drinking < 21, friends who use drugs, adult who drinks, region

• Star+ng to binge drink
• Developing high depressive symptoms

Methods

• Becoming overweight
• Star+ng to use drugs
• Star+ng to binge drink

– Age, sibling who used was drinking < 21, friends who use drugs, adult who drinks, region

• Developing high depressive symptoms

Methods

• Becoming overweight
• Star+ng to use drugs
• Star+ng to binge drink
• Developing high depressive symptoms

– Age, BMI, depressive symptoms

Ea+ng disorder subtypes predic+ng star+ng to use drugs

OR (95% CI) Non-disordered 1.0 (referent) BN 3.9 (1.8-8.4)1 BED 0.5 (0.2-1.5) PD 1.7 (1.0-3.1)1 EDNOS 1.5 (1.2-1.9)

1: no change when using monthly cut-off

26/04/17 9

Ea+ng disorder subtypes predic+ng star+ng to binge drinking frequently OR (95% CI) Non-disordered 1.0 (referent) BN

(monthly cut-off)

1.7 (1.0-3.1)

2.59 (1.76–3.81)

BED

(monthly cut-off)

1.1 (0.7-1.7)

1.42 (1.13–1.79)

PD

(monthly cut-off)

1.8 (1.3-2.7)

1.75 (1.34–2.27)

EDNOS

(monthly cut-off)

1.6 (1.4-1.9)

1.40 (1.09–1.79)

Ea+ng disorder subtypes predic+ng becoming overweight or obese

OR (95% CI) Non-disordered 1.0 (referent) BN 0.8 (0.2-2.8) BED 1.8 (1.1-2.9) PD

(monthly cut-off)

1.0 (0.5-1.9)

1.49 (1.00–2.21)

EDNOS 1.1 (0.9-1.5)

Ea+ng disorder subtypes predic+ng developing high depressive symptoms

OR (95% CI) Non-disordered 1.0 (referent) BN 0.4 (0.1-3.4) BED

(monthly cut-off)

2.3 (1.0-5.0)

1.77 (1.23–2.46)

PD 1.2 (0.6-2.2) EDNOS 1.3 (1.0-1.7)

Conclusions

• Fewer EDNOS cases with DSM-5, but s+ll the

largest group

• AN and BN are the least common disorders
• BED looks different than other ea+ng disorders:

prospec+ve associa+on with depressive sx,

• verweight/obesity
• PD looks like BN, should it be combined?

26/04/17 10

Sample-Avon Longitudinal Study of Parents and Children (ALSPAC)

• General popula+on sample of ~14,000 women

recruited in pregnancy and their children

• Cohort members have been followed up

regularly Qs sent

• 6,140 (58%) adolescents responded at age 14

years

• 5,069 (52%) at age 16 years

Sample

• Qs have also been regularly mailed to parents

ED symptoms Age 14/16 yrs

• Weight & shape concern:

3 Qs from the McKnight Risk Factor Survey

• ED behaviours:

Ques+ons about purging, binge ea+ng, fas+ng, excessive exercise derived from the Youth Risk Behavior Surveillance System ques+onnaire enquiring about the previous year

• BMI (age- and gender-adjusted)

Objec+ve weight and height

26/04/17 11 ED symptoms Age 14/16 yrs-parental report

• Parents completed the Developmental and

Well-being Assessment (DAWBA)-ED sec+on

• Sensi+vity analyses showed binge ea+ng and

purging not overlapping with youth report

• Parental report used for AN diagnosis

ED diagnoses

AN: Objec+ve underweight and food restric+on/ fas+ng/excessive exercise and shape and weight concern BN: Binge ea+ng and purging ≥once/week BED: Binge ea+ng ≥once/week and at least 3 cogni+ve symptoms (ea+ng fast or faster than normal; ea+ng un+l

stomach hurt or they felt sick, ea+ng large amounts when not hungry, ea+ng alone, feeling guilty about amount eaten)

ED diagnoses-OSFED

PD: Purging ≥once/week Subthreshold BN (S-BN): Binge ea+ng and purging ≥once/month Subthreshold BED (S-BED): Binge ea+ng ≥once/month and cogni+ve symptoms OSFED-other1: Monthly binge ea+ng/ purging/ excessive exercising or fas+ng OSFED-other2: < Monthly binge ea+ng/ purging/ excessive exercise or fas+ng

ED behaviours

Any frequency:

• Excessive exercise: exercising for weight loss or to avoid

weight gain AND exercising even when injured or sick

• r impact on school work due to the amount of +me

spent exercising;

• Purging: self-induced vomi+ng, laxa+ve use or other

medicines for weight loss or to avoid weight gain

• Fas+ng for weight loss or to avoid weight-gain

26/04/17 12

Outcomes

Obtained via Qs or structured interviews at age 15.5/16 AND 17.5/18

• Depression:
• Short Moods & Feelings ques+onnaire (sMFQ)
• Drug use (any):
• Specific Qs (ecstasy, amphetamines, crack/cocaine, opiates,

seda+ves, hallucinogens, solvents, other)

• Alcohol (binge drinking):
• Alcohol Use Disorders Iden+fica+on Test (AUDIT)
• Anxiety disorder (past 6 months):
• Developmental and wellbeing assessment (DAWBA)
• Deliberate Self-Harm
• Past month (DAWBA) at 15.5; -past year (CIS-R) at 17.5

Weight

• Objec+vely measured at 15.5 and 17.5 years
• f age
• WHO age and gender adjusted BMI cut offs for

underweight (18.5), overweight (25) and

• besity (30) based on UK reference charts as

per Cole et al. 2000, 2007 Underweight Overweight Obesity

Methods

• GEE models to inves+gate prospec+ve

associa+ons of ED with lagged outcomes- crude and adjusted

• MI by chained equa+ons used to impute

missing covariate data

• All analyses adjusted for wave of assessment,

gender, maternal educa+on, maternal parity

• Addi+onal adjustment for presence of
• utcome disorder at previous wave

ED prevalence (14 years) (n=6,140)

All (n=6,140) Girls (n=3,416) Boys (n=2,742) AN 153 (2.48%) 109 (3.19%) 44 (1.60%) BN 16 (0.26%) 14 (0.41%) 2 (0.07%) BED 30 (0.50%) 21 (0.61%) 9 (0.33%) OSFED PD 26 (0.42%) 21 (0.61%) 5 (0.18%) Subthreshold BN 82 (1.33%) 58 (1.70%) 24 (0.88%) Subthreshold BED 2 (0.03%) 1 (0.03%) 1 (0.04%) OSFED-other 1 463 (7.52%) 380 (11.12%) 83 (3.03%) OSFED-other 2 394 (6.40%) 240 (7.03%) 154 (5.62%) ALL ED 1,166 (18.95%) 844 (24.70%) 322 (11.74%) ALL ED (excluding OSFED-other) 160 (5.03%) 224 (6.55%) 85 (3.09%)

26/04/17 13 ED prevalence (16 years) (n=5,069)

All (n=5,069) Girls (n=3,059) Boys (n=2,154) AN 91 (1.75%) 72 (2.35%) 19 (0.88%) BN 42 (0.81%) 41 (1.34%) 1 (0.05%) BED 60 (1.15%) 47 (1.54%) 13 (0.60%) OSFED PD 80 (1.53%) 75 (2.45%) 5 (0.23%) Subthreshold BN 168 (3.22%) 137 (4.48%) 31 (1.44%) Subthreshold BED 22 (0.42%) 22 (0.72%) OSFED-other 1 629 (12.07%) 465 (15.20%) 164 (7.61%) OSFED-other 2 830 (15.92%) 656 (21.44%) 174 (8.08%) ALL ED 1,922(36.87%) 1,515 (49.52%) 407 (18.89%) ALL ED (excluding OSFED-other) 463 (9.13%) 394 (12.88%) 69 (3.20%)

Adjusted odds ra+os (95% Confidence intervals) for Depression

1 2 3 4 5 6 7 8 9 10 AN BN BED PD S-T BN S-T BED OSFED-Other1 OSFED-Other2 * * * ** ** *** ***

*: p≤0.05, **:p≤0.001, ***: p≤0.0001

Adjusted odds ra+os (95% Confidence intervals) for Anxiety

5 10 15 20 25 30 AN BN BED PD S-T BN S-T BED OSFED-Other1 OSFED-Other2

*: p≤0.05, **:p≤0.001, ***: p≤0.0001

* ** ** *** *** *** *** ***

Adjusted odds ra+os (95% Confidence intervals) for Drug use

5 10 15 20 25 30 AN BN BED PD S-T BN S-T BED OSFED-Other1 OSFED-Other2 ** ** * **

*: p≤0.05, **:p≤0.001, ***: p≤0.0001

** * *

26/04/17 14 Adjusted odds ra+os (95% Confidence intervals) for DSH

2 4 6 8 10 12 14 16 AN BN BED PD S-T BN S-T BED OSFED-Other1 OSFED-Other2 ** **

*: p≤0.05, **:p≤0.001, ***: p≤0.0001

*** *** *** ***

Weight outcomes

• AN predicted underweight at the following wave: adjusted

OR=2.43 (1.62-3.66), p≤0.0001

• BED predicted obesity OR=3.58 (1.06-12.14), p=0.04
• BN predicted overweight & obesity OR=3.43 (1.06-11.07)
• OSFED 1 and 2 both predicted overweight and obesity

respec+vely OR=1.80 (1.36-2.39) OR=1.74 (1.31-2.30)

Summary I

Specific associa&on pa`erns:

• binge/purge disorders associated with later drug use
• strong associa+on between ED and DSH, driven by

purging?

• all ED with depression and anxiety

Underweight in AN persistent BED and BN and OSFED-other associated with

• besity/overweight

Summary II

Low behaviour frequency OSFED and higher behaviour frequency OSFED similarly associated with adverse outcomes Preven+ng more common ED behaviours likely to impact a large part of the popula+on Are current diagnos+c frequency thresholds jus+fied?

26/04/17 15

Conclusions

• ED and disordered ea+ng are common in the

general popula+on

• Sub-threshold disorders are associated with

adverse outcomes

• Need to be aware of presenta+ons in boys

and girls

Acknowledgements

Dr Alison Field Dr Nicholas Horton Dr Ross Crosby Prof Janet Treasure Francesca Solmi Sonja Swanson All ALSPAC par+cipants All GUTS par+cipants R01-MH087786

THANK YOU!