Table 1.

Summary of study characteristics (n = 43)

Author	Country	ED	Sample size (N)	BMI (SD) (kg/m2) or % of IBW	Amenorrhoea (%)	OCP use (%)	Diagnostic method	Design	ED population	Controls	Scanning method	Bone density outcome measures	Body composition outcome measures	NOS-Scale stars of 9
Andersen et al. [28]	USA	AN-R	28	68.1 (14.2) % of IBW	Yes	Unknown	DSM-IV	Case–control	Hospital outpatients	Similar age and white controls Referred to radiology for non-orthopaedic, non-eating disorder problems to ‘rule out’ osteoporosis	DXA	Spine BMD		5
					Unknown %
		AN-BP	36	73.3 (11.7) % of IBW	Unknown	Unknown
		BN-PAN	18	111.3 (27.8) % of IBW	Unknown	Unknown
		BN-NPAN	14	109.4 (11.3) % of IBW	Unknown	Unknown
		HC	15	Unknown	Unknown	Unknown
Bachmann et al. [29]	USA	AN	65	15.9	Yes	Yes	DSM-5	Case–control	Premenopausal women who had undergone quantitative computed tomography	Premenopausal women who had undergone quantitative computed tomography	DXA QCT	Spine BMD	Fat mass Lean mass	5
					Unknown %	Unknown %
		HC	45	19.8
Bacopoulou et al. [26]	GRE	AN	17	17.0 (1.1)	Yes	Unknown	DSM-5	Case–control	Adolescent medicine clinic	Healthy controls Age-matched girls presented for annual health examination in an adolescent medicine clinic	DXA	Total body BMD Spine aBMD		5
					100%
		AN-Atypical	5	18.3 (16.0–22.9)	Yes	Unknown	DSM-5
					60%
		HC	15	22.1 (1.7)
Bratland-Sanda et al. [27]	NOR	AN	8	15.6 (1.4)	Unknown	Unknown	DSM- IV	Case–control	ED hospitalised patients	Healthy controls Age-matched Randomly selected from a national pool	DXA	Total body BMD Spine BMD Femoral neck BMD	Fat mass Lean mass	8
		BN	29	22.3 (3.5)	Unknown	Unknown	DSM- IV
		EDNOS	22	21.0 (3.2)	Unknown	Unknown	DSM- IV
		HC	53	25.3 (4.8)
Bredella et al. [30]	USA	AN	10	18.4 (1.0)	Yes	Yes	DSM-IV	Case–control	ED clinics	Healthy controls Recruited through clinic advertisements	DXA	Total body BMD Spine BMD Femoral BMD Hip BMD	Fat mass Lean mass	7
					100%	20%
		HC	10	20.5 (2.6)
Bredella et al. [31]	USA	AN	10	17.6 (1.0)	Unknown	Unknown	‘Psychiatric diagnostic criteria for AN’	Case–control	Clinic referrals	Healthy controls Recruited through community advertisements	DXA MRI	Total body BMD Spine BMD Hip BMD		5
		HC	10	21.9 (1.7)
Bredella et al. [32]	USA	AN	5	18.3 (0.9)	Unknown	No	DSM-IV	Case–control	Clinic referrals	Healthy Controls Recruited through community advertisements	DXA Fluorodeoxyglucose- PET CT	Total body BMD Spine BMD Femoral neck BMD Hip BMD Lateral spine BMD	Fat mass Lean mass	5
		HC	5	21.9 (1.7)
Čagalová et al. [33]	SLO	AN	63	14.5 (1.8)	Yes	No	DSM-5	Case–control	Hospital patients of the Department of Paediatrics	General paediatric practice	DXA	Total body BMD Spine BMD Hip BMD		5
					81.4%
		HC	20	20.2 (1.7)
Davies et al. [34]	USA	AN	26	Unknown	Yes	Yes	DSM-III-R	Case–control	Clinic records of eating disorder patients from medical centre	Healthy controls Select for fitness and diet study or for a DXA and DPA compar ison group	DPA	Spine BMD Forearm BMD		4
					80.8%	15.4%
		AN-BN	26	Unknown	Yes	Yes	DSM-III
					61.5%	23.1%
		BN	11	Unknown	Yes	Yes
					18.2%	9.1%
		HC	211	Unknown
Estour et al. [35]	FRA	AN-R	40	16.0 (0.8)	Yes	No	DSM-IV	Case–control	Hospital outpatients	Healthy controls ranging 18.6–25.0 kg/m2 BMI	DXA	Spine BMD Femoral BMD	Fat mass Lean mass	5
					100%
		HC	54	20.9 (2.2)
Faje et al. [36]	USA	AN	21	17.8 (0.2)	Unknown	Unknown	DSM-IV	Cohort	Hospital outpatients	Healthy controls with normal weight	DXA	Total body aBMD Spine aBMD Hip aBMD Distal radius aBMD	Fat mass Lean mass	6
		HC	23	22.4 (0.5)
Fazeli, Klibanski [37]	USA	AN	26	16.7 (0.5)	Yes	No	DSM-IV	Cohort	ED referrals	Healthy controls Recruited through community advertisements	DXA MRI	Spine BMD Femoral BMD Hip BMD		5
		HC	20	22.6 (0.3)
Fernández-Soto et al. [38]	SPA	AN-R	16	16.8 (1.5)	Yes	No	DSM-IV	Cohort	Clinic outpatients	Healthy controls Caucasian women	DXA MRI	Total body BMD Spine BMD	Fat mass Lean mass	5
		AN-R	31	19.1 (2.0)	Yes	No	DSM-IV
		HC	25	21.8 (0.9)
Frølich et al. [39]	DEN	AN	25	16.2 (1.3)	Yes	Yes	ICD-10	Case–control	ED unit outpatients	Healthy controls Age and height-matched females Randomly selected from a national pool	DXA	Spine BMD Femoral BMD Hip BMD	Fat mass Lean mass	7
					100%
		HC	25	22.8 (2.7)
Guo et al. [40]	CHI	AN	26	16.3 (0.6)	Yes		DSM-IV	Cohort	Clinical psychiatrist referrals	Healthy controls Age-matched	DXA	Total body BMD Spine BMD Hip BMD	Fat mass Lean mass	5
					100%
		HC	24	20.8 (0.7)
Haas et al. [41]	GER	AN	103	16.5 (1.6)	Yes	Yes	DSM-IV	Longitudinal observational	Patients of adolescent and youth medicine clinic	Normal weight controls Recruited from a local high school	DXA	Total body BMD	Fat mass Lean mass	5
					94.2%	5.6%
		HC	51	20.8 (1.8)
Iketani et al. [42]	JAP	AN	22	14.4 (2.3)	Yes	No	DSM-III-R	Cohort	Inpatients and outpatients	Healthy controls Age-matched healthy females	DPA	Total body BMD Spine BMD		5
					Unknown %
		AN-BN	23	14.4 (1.6)	Yes	No	DSM-III-R
					Unknown %
		BN-PAN	10	19.0 (2.6)	Yes	No	DSM-III-R
					40%
		HC	10	19.5 (0.8)
Kandemir et al. [43]	USA	AN	262	17.2 (0.1)	Yes		DSM-IV / DSM-5	Case–control	AN participants from Childhood Study	Healthy controls participants from Childhood Study	DXA	Total body BMD Spine BMD Hip BMD		5
					85.9%
		HC	90	21.0 (0.2)
Karlsson et al. [44]	AUS	AN	77	15.6 (0.2)	Unknown	Unknown	ICD-10	Case–control	AN patients untreated with oestrogen therapy	Healthy controls	DXA	Spine BMD Femoral neck BMD	Fat mass Lean mass	5
		HC	205	23.1 (0.3)
Kooh et al. [45]	JAP	AN	22	15.9 (2.2)	Yes	Yes	DSM-III-R	Cohort	Clinic referrals: Adolescent medicine clinic	Healthy controls School and university students, no oral contraceptives	DXA	Spine BMD Femoral neck BMD	Fat mass Lean mass	5
					72.3%	4.50%
		HC	24	21.6 (2.2)
Maïmoun et al. [46]	FRA	AN-R	206	15.8 (1.7)	Yes	No	DSM-IV	Case–control	Hospital outpatients	Healthy controls Recruited through community advertisement	DXA	Total body aBMD Spine aBMD Hip BMD Proximal femur aBMD	Fat mass Lean mass	6
					100%
		AN-R	99	16.3 (1.4)	No	Yes	DSM-IV
						100%
		HC	121	21.6 (2.3)
Masala et al. [47]	ITA	AN	17	17.5 (1.6)	Unknown	Unknown	ICD-10	Cohort	Patients in weight gain programme	Healthy controls Exclusion included medication or illness to affect bone	DXA QCT	Spine BMD		4
		HC	27	24.3 (4.8)
Mathisen et al. [48]	SWE	BN-NPAN	100		Unknown	Unknown	DSM-5	Cross-sectional	Patients from a randomised controlled trial	No healthy control group	DXA	Total body BMD Spine BMD	Fat mass Lean mass	6
		BN-PAN	37		Unknown	Unknown	DSM-5
Misra et al. [49]	USA	AN	23	16.7 (1.2)	Unknown	Unknown	DSM-IV	Cohort	Clinic referrals	Healthy controls Age-matched and bone age- matched Recruited through adverts through healthcare providers and newspapers	DXA	Spine BMD Hip BMD	Fat mass Lean mass	6
		HC	21	21.7 (3.7)
Misra et al. [50]	USA	AN	17	16.7 (1.3)	Unknown	Unknown	DSM-IV	Cohort	Paediatrician referrals	Healthy controls Age-matched and bone age- matched Recruited through mailings to paediatricians	DXA	Total body BMD Spine BMD Femoral neck BMD Hip BMD	Fat mass Lean mass	6
		HC	19	21.8 (3.4)
Misra et al. [51]	EUA	AN	110	17.4 (0.1)	Yes	No	DSM-IV	RCT	Hospital outpatient treatment programme	Healthy controls Recruited through mailings to paediatricians	DXA	Spine BMD Hip BMD	Fat mass Lean mass	7
					100%
		HC	40	21.4 (0.5)
Morris et al. [52] 	UK	AN	51	15.1 (1.9)	Yes	Yes	DSM-IV	Cohort	ED specialist referrals	Control group data from department of medical physics	DXA X-Ray	Total body BMD Spine BMD	Fat mass	5
					100%
		BN	47	21.9 ( 3.4)		Yes	DSM-IV
		BN-NPAN	26	22.3 (2.7)		Yes	DSM-IV
		BN-PAN	21	21.4 (3.0)		Yes	DSM-IV
		HC	40	23.3 (4.4)
Naessén et al. [53]	SWE	BN	77	22	Yes	No	DSM-IV	Cohort	Recruited from hospital advertisements	Healthy controls Hospital advertising: hospital staff and students, no current diseases or medication prior to 3 months before the study	DXA	Total body BMD Spine BMD Leg BMD	Fat mass Lean mass	8
					7.8%
		BN-NPAN	59				DSM-IV
		BN-PAN	18				DSM-IV
		HC	56	22.2
Newman, Halmi [54]	USA	AN	18	73.2 (8.4)	Yes	No	DSM-III-R	Cross-sectional	Inpatients from an ED unit	Normal weight women	DXA	Spine BMD Femoral BMD		4
					100%
		BN	12		Yes	Yes	DSM-III-R
					66.7%	8.4%
		HC	12
Newton et al. [55]	AUS/UK	BN	20	21.8 (3.5)	Yes	Yes	DSM-III-R	Cohort	ED outpatient treatment programme	Healthy controls Age and sex-matched controls from hospital staff notice boards	DXA	Spine BMD		7
					20%
		HC	16	21.9 (1.8)
Olmos et al. [56]	SPA	AN	51	17.3 (2.4)	Yes		DSM-IV	Prospective longitudinal cohort study	ED unit outpatients	Healthy controls Hospital advertisements	DXA	Spine BMD Femoral neck BMD Hip BMD		5
					100%
		HC	40	21.8 (2.7)
Poet et al. [57]	FRA	AN	18	Unknown	Yes	No	DSM-III-R	Cohort	Hospital outpatients	Healthy controls Volunteers	DXA	Spine BMD		5
					100%
		HC	36	Unknown
Resch et al. [58]	AU	AN	20	16.0 (13–18)	Unknown	Unknown	DSM-III-R	Cohort	Hospital outpatients	Healthy controls Age-matched nursing school students	DXA	Spine BMD Hip BMD		5
		HC	20	23.0 (19–29)
Schorr et al. [59]	USA	AN	46	16.7 (1.8)	Yes	Yes	DSM-5	Cross-sectional	Patients of National Institutes Health trials	Lean healthy controls From the National Institutes Health trial	DXA HRpQCT	Spine BMD Femoral BMD Hip BMD		6
					60.9%	15.3%
		HC	29	22.6 (1.4)
Seeman et al. [60]	AUS	AN	12	15.9 (1.3)	Yes	No	DSM-III-R	Cohort	Patients with AN	Healthy controls Volunteers with no illness that affects the bone, no drugs, no medication	DXA	Total body BMD Spine BMD Proximal femur Femoral neck BMD Ward’s triangle BMD Trochanter BMD	Fat mass Lean mass	4
					100%
		AN	37	16.6 (0.4)	Yes	No	DSM-III-R
					100%
		AN OCP	16	17.2 (0.8)		Yes	DSM-III-R
						100%
		HC	52	23.3 (0.5)
Singhal, Bredella [61]	USA	AN	55	18.7 (0.2)	Yes	No	DSM-IV / DSM-5	Cross-sectional	Ongoing studies assessing bone outcomes	Normal-weight controls From ongoing studies assesing bone outcomes	DXA HRpQCT	Total body BMD Spine BMD Femoral BMD Hip BMD	Fat mass Lean mass	6
					27.3%
		HC	48	21.7 (0.3)
Soyka et al. [62]	USA	AN	19	16.5 (0.4)	Yes	No	DSM-IV	Cohort	Healthcare provider referrals	Healthy controls Recruited through advertisement in primary care providers and newspapers, BMI 25th-90th percentile, one participant in pre- menarche	DXA	Total body BMD Spine BMD Lateral spine BMD	Fat mass Lean mass	6
					73.7%
		HC	19	21.8 (0.4)
Strumila et al. [63]	FRA	BN-NPAN	50	21.9 (2.6)	Irregular menstrual cycle	Unknown	DSM-5	Cross-sectional	ED unit	No	DXA	Spine BMD Hip BMD	Fat mass	5
					35%
		BN-PAN	35	21.9 (2.9)	Irregular menstrual cycles	Unknown	DSM-5
					46.7%
Sundgot-Borgen et al. [64]	NOR	AN	13	15.8 (0.9)	Yes	No	DSM-IV	Case–control	Clinic referrals	Healthy controls University information board recruitment Comprehensive inclusion criteria for dietary, exercise and ED symptoms	DXA	Total body BMD Spine BMD Femoral neck BMD Leg BMD Arm BMD	Fat mass	7
					100%
		BN	43	20.7 (2.0)	Yes	No	DSM-IV
					32.6%
		HC	17	22.0 (2.4)
van Marken Lichtenbelt et al. [65]	NL	AN-R	12	16.5 (1.7)	Yes	Yes	DSM-III-R	Cohort	Non-hospitalised outpatients	Healthy controls Normal weight participating from a study on energy expenditure	DXA	Total body BMD	Fat mass Lean mass	3
					58.3%	33.3%
		HC	16	23.7 (2.1)
Walsh et al. [66]	USA	AN	8	17.4 (1.4)	Unknown	Unknown	DSM-IV	Cohort	Hospital outpatients	Healthy controls 90–100% ideal weight for age	DXA CT	Spine BMD Femoral neck BMD Hip BMD	Fat mass Lean mass	4
		HC	6	24.4 (2.5)
Wojcik et al. [67]	USA	AN	15	17.6 (0.2)	Unknown	No	DSM-IV	Cohort	Healthcare referrals and community adverts	Healthy controls Recruitment through community advertisement	DXA	Total body BMD Spine BMD Femoral neck BMD Hip BMD	Fat mass Lean mass	4
		HC	16	23.5 (0.6)
Wu et al. [68]	CHI	AN	25	17.5 (1.3)	Yes	Unknown	DSM-5	Case–control	ED providers referrals	Healthy controls Recruited through mailings to paediatricians	DXA	Total body BMD Spine BMD Femoral neck BMD	Fat mass Lean mass	4
					100%
		HC	31	20.2 (1.3)

ED eating disorder, HC healthy controls, BMI body mass index, IBW ideal body weight, SD standard deviation, AN anorexia nervosa, BN bulimia nervosa, AN-R AN-restrictive subtype, BN-NPAN BN with no previous AN history, BN-PAN BN with previous AN, AN-BN AN with an additional diagnosis of BN (according to DSM-III-R), AN-PB AN-binge purging subtype, EDNOS ED not otherwise specified, OCP oral contraceptive pill, DSM Diagnostic and Statistical Manual of Mental Disorders, CID International Classification of Diseases, DXA dual-energy X-ray absorptiometry, DPA dual photon absorptiometry, CT computerised tomography, MRI magnetic resonance imaging, PET positron emission tomography, BMD bone mineral density