. 2021 Mar 9;4(1):115–131. doi: 10.1136/bmjnph-2020-000188

Table 3.

Fragility Index of the included randomised controlled trials

First author	Outcome	Intervention	Intervention arm (n)	Events in intervention arm (n)	Control arm (n)	Events in Control arm (n)	P-value (Intervention vs control)	Fragility Index (FI)	Reverse Fragility Index (RFI)
Al Wattar⁵⁸	Maternal composite outcome*	MD	486	111	500	143	0.04	2	–
	Offspring composite outcome†	MD	531	92	564	118	0.145	0	7
Assaf-Balut⁵⁹	GDM incidence	MD + EVOO + Pistachio	434	74	440	103	0.012	4	–
Assaf-Balut⁶⁰	Composite maternal-fetal outcome‡	MD + EVOO + Pistachio	360	32	337	87	0.0001	39	–
Babio⁶¹	Incidence of MetS	MD + EVOO	1982	29§	1934	75§	<0.001	26	–
	Incidence of MetS	MD + Nuts	1885	58	1934	75	0.186	0	7
de Lorgeril⁶²	Composite outcome¶	MD	219	14	204	44	0.0001	17	–
de Lorgeril³⁴	CV mortality	MD	302	6	303	19	0.01	3	–
	Non-fatal ΜΙ	MD	302	8	303	25	<0.001	5	–
de Lorgeril^{35 36}	CV mortality	MD	302	3	303	16	0.004	4	–
	Non-fatal MI	MD	302	5	303	17	0.015	2	–
Díaz-López^{37 38}	Diabetic retinopathy	MD + EVOO	7830	22	6856	32	0.075**	0	2
Díaz-López^{37 38}	Diabetic retinopathy	MD + Nuts	6622	20	6856	32	0.126	0	3
Esposito^{31 32}	Need for T2DM medication	LCMD	108	48	107	75	<0.001	14	–
Estruch^{18 19}	Composite CV events††	MD + EVOO	11 852	96	9763	109	0.024	5	–
	Composite CV events††	MD + Nuts	10 365	83	9763	109	0.024	4	–
García-Gavilán^{33 53}	Osteoporotic fractures	MD + EVOO	291	40	290	37	0.807	0	13
García-Gavilán^{33 53}	Osteoporotic fractures	MD + Nuts	289	37	290	37	1	0	16
García-Layana⁴⁰	Cataract surgery incidence	MD + EVOO	11 728	206	10 633	179	0.681	0	28
García-Layana⁴⁰	Cataract surgery incidence	MD + Nuts	10 719	174	10 633	179	0.748	0	29
Gené Huguet⁴¹	Reversion to robustness	MD	85	14	88	1	<0.001	5	–
Greenberg⁵⁷	Weight loss	MD (vs LFD)	92	41	91	35	0.454	0	8
	Weight loss	MD (vs LCD)	92	41	89	39	1	0	13
Marcos-Forniol⁴²	Optimal CV risk factor control‡‡	MD	54	34	52	15	<0.001	7	–
Martínez-González⁴³	Atrial fibrillation incidence	MD + EVOO	10 634	72	8851	89	0.014	7	–
Martínez-González⁴³	Atrial fibrillation incidence	MD + Nuts	9333	92	8851	89	0.94	0	27
Papadaki^{44 63}	Heart failure incidence	MD + EVOO	11 737	29	9664	32	0.303	0	7
	Heart failure incidence	MD + Nuts	10 279	33	9664	32	0.902	0	14
Pintó⁵²	Presence of steatosis	MD + EVOO	34	3	30	10	0.027	1	–
	Presence of steatosis	MD + Nuts	36	12	30	10	1	0	7
Properzi⁴⁵	NAFLD resolution	MD	26	3	25	9	0.046§§	0	1
Ruiz-Canela⁵⁴	New symptomatic PAD cases	MD + EVOO	2539	18	2444	45	<0.001	12	–
	New symptomatic PAD cases	MD + Nuts	2452	26	2444	45	0.023	3	–
Salas-Salvadó⁵⁵	Reversion of MetS	MD + EVOO	252	53	250	41	0.208	0	6
	Reversion of MetS	MD + Nuts	249	63	250	41	0.015	4	–
	New MetS Incidence	MD + EVOO	157	36	154	36	1	0	15
	New MetS Incidence	MD + Nuts	162	29	154	36	0.266	0	6
Salas-Salvadó^{39 46}	Incidence of T2DM	MD + EVOO	570	14	515	24	0.068	0	1
Salas-Salvadó^{39 46}	Incidence of T2DM	MD + Nuts	598	16	515	24	0.105	0	2
Salas-Salvadó^{47 48}	New-onset of T2DM	MD + EVOO	4990	80	4271	101	0.01	9	–
Salas-Salvadó^{47 48}	New-onset of T2DM	MD + Nuts	4876	92	4271	101	0.126	0	6
Sánchez-Villegas⁵⁶	Depression	MD + EVOO	7715	88	6096	77	0.018	5	–
		MD + Nuts	6803	59	6096	88	0.003	13	–
Singh⁴⁹	Total cardiac events	Indo-MD	499¶¶	39	501	76	<0.001	16	–
Toledo⁵⁰	First invasive breast cancer	MD + EVOO	7031	8	5829	17	0.027	2	–
	First invasive breast cancer	MD + Nuts	5492	10	5829	17	0.253	0	4
Tuttle⁵¹	Total outcome endpoints***	MD	51	8	50	8	1	0	7

*gestational diabetes mellitus (GDM) or preeclampsia.

†Stillbirth, small-for-gestational age (SGA) fetus, or admission to the neonatal care unit.

‡Emergency caesarean section, perineal trauma, pregnancy-induced hypertension and preeclampsia, prematurity, large-for-gestational age, and SGA.

§The number of events was calculated by subtracting the baseline n from the follow-up n count.

¶Cardiac death and non-fatal myocardial infarction (MI).

**In the manuscript, the comparison is reported as ‘significant’ based on the multivariable-adjusted model; however, analysis using Fischer’s exact test does not reveal a significant difference.

††MI, stroke, or death from cardiovascular causes.

‡‡Defined as the achievement of ≥5 risk factor goals: blood pressure <140/90 mm Hg, low-density lipoprotein <2.6 mmol/L, smoking cessation, body mass index <25 kg/m², physical activity of moderate intensity >30 min/day, 3 days/wk (≥6 metabolic equivalents h/wk) and HbA_1c <7% in patients with diabetes.

§§In the manuscript, the comparison is reported as significant (p=0.046); analysis using Fischer’s exact test does not reveal a significant difference.

¶¶In the manuscript, table 4 reports that the size of the intervention group (n) was 999; instead, the count reported in the main manuscript text and the other tables was used for the calculation of the FI (n=499).

***All-cause and cardiac deaths, MI, hospital admissions for heart failure, unstable angina pectoris or stroke.

CV, cardiovascular; EVOO, extra-virgin olive oil; LCD, low-carbohydrate diet; LCMD, low-carbohydrate MD; LFD, low-fat diet; MD, Mediterranean diet; MetS, metabolic syndrome; NAFLD, non-alcoholic fatty liver disease; PAD, periphery artery disease; T2DM, type 2 diabetes mellitus.