Table 2.

Subgroup analysis of prospective studies of linoleic acid and breast cancer risk for highest v. lowest category of linoleic acid

Study	No. of results	No. of cases	No. of participants	RR	95 % CI	I 2 (%)	P *	P †
Overall	14	10 410	358 955	0·98	0·93, 1·04	17·80	0·26
Exposure assessment								0·90
Dietary linoleic acid intake	6	9354	335 952	0·96	0·92, 1·01	0·00	0·56
Serum linoleic acid level	8	1056	23 003	0·98	0·93, 1·04	40·90	0·11
Study design								0·67
Prospective cohort	9	9483	356 396	0·97	0·91, 1·04	45·60	0·07
Nested case–control	5	927	2559	1·00	0·91, 1·11	0·00	0·93
Menopausal status								0·41
Pre- and postmenopausal	4	4119	219 137	0·95	0·85, 1·05	0·00	0·56
Premenopausal	2	118	8287	0·64	0·20, 2·09	56·50	0·13
Postmenopausal	8	6173	131 531	1·01	0·94, 1·08	28·60	0·20
Nations								0·77
USA	6	6423	205 419	1·00	0·92, 1·09	0·00	0·71
Europe	7	3275	80 965	0·95	0·88, 1·03	47·20	0·08
Asia	1	712	72 571	1·13	0·83, 1·55	0·00	0·00
Quality scores (stars)								0·75
6	4	564	1494	0·99	0·90, 1·11	0·00	0·89
7	10	9846	357 461	0·98	0·91, 1·05	40·20	0·09
Follow-up duration (years)								0·54
>5	9	9134	284 890	0·97	0·90, 1·04	43·50	0·08
≤5	5	1276	74 065	1·01	0·92, 1·12	0·00	0·88
Covariates adjusted
Family history of breast cancer								0·77
Yes	9	9914	337 214	0·99	0·92, 1·06	0·00	0·83
No	5	496	21 741	0·69	0·41, 1·14	65·00	0·02
BMI								0·21
Yes	8	9143	335 775	0·99	0·94, 1·05	0·00	0·69
No	6	1267	23 180	0·82	0·64, 1·06	44·50	0·11
Smoking								0·28
Yes	4	5668	160 586	1·03	0·94, 1·13	0·00	0·80
No	10	4742	198 369	0·96	0·89, 1·03	32·30	0·15
Alcohol								0·36
Yes	7	8242	281 397	1·01	0·94, 1·09	0·00	0·72
No	7	2168	77 558	0·98	0·93, 1·04	45·60	0·09
Total energy intake								0·50
Yes	5	7749	279 945	1·00	0·93, 1·09	0·00	0·57
No	9	2661	79 010	0·96	0·88, 1·04	34·80	0·14

RR, relative risk.

^* P value for heterogeneity within subgroup.

^† P value for heterogeneity between subgroups with a meta-regression analysis.