Table 2.

Estimates of association (from logistic regression model) of potential predictors for prevalence of breast cancer or precursor lesion

Possible Predictor	Invasive Cancer (IC)		In-situ cancer (DCISa + LCISb)		Atypical Hyperplasia (ADHc + ALHd)		All categories (IC + DCIS + LCIS + ADH + ALH)
	Odds Ratio	P Value	Odds Ratio	P Value	Odds Ratio	P Value	Odds Ratio	P Value
Mean number of histologic sections examined per woman		0·75		<0·001*		<0·001*		0·002*
per unit on logscale	0.9 (0.7–1.3)		3·2 (1·6–6·4)*		1·6 (1·1–2·4)*		1·79 (1·3–2·4)*
> 20 sections taken on average	1·1 (0·5–2·7)		127.0 (20·2–793)*		21·3 (9·7–46·7)*		22·4 (17·2–49·8)*
Mean age		0·48		0·75		1·0		1·00
per decade increase	1.1 (0.8–1.5)		1·3 (0·34–5·30)		1·1 (0·4–2·7)		1·07 (0·53–2·1)
Mean age > 50	0·8 (0·3–1.7)		1·9 (0·03–105.1)		2·2 (0·2–25·7)		1·24 (0·08–7·9)
Mean age > 70	1.7 (0.7–4.1)		0·6 (0·003–4·6)		0·5 (0·02–13·1)		0·91 (0·08–2·4)
Forensic vs hospital based	0.6 (0.3–1.4)	0.27	2·8 (0·05–168.0)	0·53	0·5 (0·05–5·7)	0·65	0·60 (0·10–3·7)	0·58
Consecutive cases	1.8 (0.7–4.3)	0.19	37·7 (0·5–2777.0)	0·06	1·5 (0·1–16·3)	1·0	2·3 (0·48–11·2)	0·32
Adjusted for number of sections examined			4·5 (0·77–26·3)	0·11
Breast cancer may have caused death	6·6 (1·9–23·3)*	0·02*	0·0 (0·0-∞)	0·21	0·5 (0·0–23·1)	0·75	1·2 (0·1–17·1)	1·00
Peer Review	1·4 (0·5–4.0)	0.58	0·33 (0·04–2·5)	0·29	0·3 (0·02–4·3)	0·44	0·39 (0·06–2·8)	0·37
Internationally recognized reporting method?	1.0 (0.4–2.3)	1.0	8·6 (0·26–288)	0·37	8·6 (0·26–288)	0·37	2·3 (0·42–12·4)	0·37
Time trend (decade of autopsies)	1·0 (0·5–2·0)	1·0	1·0 (0·2–4·4)	0·65	1·4 (0·7–3·0)	0·40	1·6 (0·93–2·7)	0·09
Adjusted for number of sections examined							1·2 (0·81–2·5)	0·37

*P value <0·05

^a DCIS Ductal carcinoma in-situ

^b LCIS Lobular Carcinoma in-situ

^c ADH Atypical Ductal Hyperplasia

^d ALH Atypical Lobular Hyperplasia