# Larsson & Wolk 2006 (Int J Cancer 119:2657): red meat, highest vs lowest, 15 cohort studies

All pooled estimates: DerSimonian–Laird random effects; weights are random-effects weights (%). Leave-one-out after each run.

## CRC HvL – published per-study values (Table I / Fig 1)

Published: 1.28 (1.15–1.42). **Pooled: 1.28 (1.15–1.42)** (fixed 1.28 (1.15–1.42); I² 0 %; k = 15). **Reproduced.**

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 4.1 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 3.2 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 3.7 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| hsing1998 | 1.90 (0.90–4.30) | 1.8 | Table I p2658: 145 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 5.5 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 1.10 (0.70–1.70) | 5.6 | Table I p2658: 185 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| jarvinen2001 | 1.50 (0.77–2.94) | 2.5 | Table I p2658: 109 cases; excluded from type-of-meat strata (items not specified) | f |
| tiemersma2002 | 1.60 (0.90–2.90) | 3.2 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 14.3 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 1.21 (0.72–2.03) | 4.1 | Table I p2658: 876 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_hpfs | 1.24 (0.78–1.96) | 5.2 | Table I p2659: 602 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| english2004 | 1.40 (1.00–1.90) | 10.8 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.32 (1.03–1.68) | 18.5 | Table I p2659: 733 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| chao2005 | 1.36 (0.93–2.00) | 7.6 | Table I p2659: 1667 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| norat2005 | 1.35 (0.96–1.88) | 9.8 | Table I p2659: 1329 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |

Leave-one-out: without bostick1994: 1.29 (1.16–1.44); without kato1997: 1.28 (1.15–1.43); without chen1998: 1.29 (1.16–1.43); without hsing1998: 1.27 (1.14–1.42); without singh1998: 1.27 (1.14–1.42); without pietinen1999: 1.29 (1.16–1.44); without jarvinen2001: 1.28 (1.15–1.42); without tiemersma2002: 1.27 (1.14–1.42); without flood2003: 1.32 (1.17–1.47); without wei2004_nhs: 1.29 (1.15–1.43); without wei2004_hpfs: 1.28 (1.15–1.43); without english2004: 1.27 (1.13–1.42); without larsson2005: 1.27 (1.13–1.43); without chao2005: 1.28 (1.14–1.42); without norat2005: 1.27 (1.14–1.42)

## CRC HvL – corrected: inclusion

**Pooled: 1.27 (1.14–1.42)** (fixed 1.27 (1.14–1.42); I² 0 %; k = 14). 

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 4.2 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 3.2 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 3.8 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 5.6 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 1.10 (0.70–1.70) | 5.7 | Table I p2658: 185 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| jarvinen2001 | 1.50 (0.77–2.94) | 2.5 | Table I p2658: 109 cases; excluded from type-of-meat strata (items not specified) | f |
| tiemersma2002 | 1.60 (0.90–2.90) | 3.3 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 14.5 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 1.21 (0.72–2.03) | 4.2 | Table I p2658: 876 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_hpfs | 1.24 (0.78–1.96) | 5.3 | Table I p2659: 602 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| english2004 | 1.40 (1.00–1.90) | 11.0 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.32 (1.03–1.68) | 18.9 | Table I p2659: 733 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| chao2005 | 1.36 (0.93–2.00) | 7.7 | Table I p2659: 1667 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| norat2005 | 1.35 (0.96–1.88) | 10.0 | Table I p2659: 1329 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |

Leave-one-out: without bostick1994: 1.28 (1.15–1.43); without kato1997: 1.27 (1.14–1.42); without chen1998: 1.28 (1.15–1.42); without singh1998: 1.27 (1.13–1.41); without pietinen1999: 1.28 (1.15–1.43); without jarvinen2001: 1.27 (1.14–1.41); without tiemersma2002: 1.26 (1.13–1.41); without flood2003: 1.30 (1.16–1.46); without wei2004_nhs: 1.28 (1.14–1.42); without wei2004_hpfs: 1.27 (1.14–1.42); without english2004: 1.26 (1.12–1.41); without larsson2005: 1.26 (1.12–1.42); without chao2005: 1.27 (1.13–1.41); without norat2005: 1.26 (1.13–1.41)

## CRC HvL – corrected: exposure

**Pooled: 1.21 (1.10–1.34)** (fixed 1.21 (1.10–1.34); I² 0 %; k = 14). 

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 3.7 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 2.9 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 3.4 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| hsing1998 | 1.90 (0.90–4.30) | 1.6 | Table I p2658: 145 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 5.0 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 0.80 (0.50–1.20) | 5.2 | exposure: unprocessed red meat Q4 vs Q1, multivariate |
| tiemersma2002 | 1.60 (0.90–2.90) | 2.9 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 12.9 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 1.21 (0.72–2.03) | 3.7 | Table I p2658: 876 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_hpfs | 1.24 (0.78–1.96) | 4.7 | Table I p2659: 602 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| english2004 | 1.40 (1.00–1.90) | 9.7 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.22 (0.98–1.53) | 20.2 | exposure: red meat main dish (unprocessed) ≥4 servings/week |
| chao2005 | 1.36 (0.93–2.00) | 6.8 | Table I p2659: 1667 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| norat2005 | 1.17 (0.92–1.49) | 17.2 | exposure: fresh red meat ≥80 vs <10 g/day, model ‡ |

Leave-one-out: without bostick1994: 1.22 (1.10–1.35); without kato1997: 1.21 (1.10–1.34); without chen1998: 1.21 (1.10–1.34); without hsing1998: 1.20 (1.09–1.33); without singh1998: 1.20 (1.09–1.33); without pietinen1999: 1.24 (1.12–1.38); without tiemersma2002: 1.20 (1.09–1.33); without flood2003: 1.23 (1.11–1.37); without wei2004_nhs: 1.21 (1.10–1.34); without wei2004_hpfs: 1.21 (1.09–1.34); without english2004: 1.19 (1.07–1.33); without larsson2005: 1.21 (1.08–1.35); without chao2005: 1.20 (1.08–1.33); without norat2005: 1.22 (1.09–1.36)

## CRC HvL – corrected: model

**Pooled: 1.28 (1.16–1.41)** (fixed 1.28 (1.16–1.41); I² 0 %; k = 15). 

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 3.4 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 2.7 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 3.2 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| hsing1998 | 1.90 (0.90–4.30) | 1.5 | Table I p2658: 145 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 4.6 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 1.10 (0.70–1.70) | 4.8 | Table I p2658: 185 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| jarvinen2001 | 1.50 (0.77–2.94) | 2.1 | Table I p2658: 109 cases; excluded from type-of-meat strata (items not specified) | f |
| tiemersma2002 | 1.60 (0.90–2.90) | 2.7 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 12.0 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 1.21 (0.72–2.03) | 3.5 | Table I p2658: 876 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_hpfs | 1.24 (0.78–1.96) | 4.4 | Table I p2659: 602 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| english2004 | 1.40 (1.00–1.90) | 9.1 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.32 (1.03–1.68) | 15.6 | Table I p2659: 733 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| chao2005 | 1.28 (1.04–1.58) | 22.1 | model: fixed-effect pool of colon and rectum as fn 6 states |
| norat2005 | 1.35 (0.96–1.88) | 8.3 | Table I p2659: 1329 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |

Leave-one-out: without bostick1994: 1.29 (1.17–1.42); without kato1997: 1.28 (1.16–1.41); without chen1998: 1.28 (1.16–1.41); without hsing1998: 1.27 (1.15–1.40); without singh1998: 1.27 (1.15–1.40); without pietinen1999: 1.29 (1.17–1.42); without jarvinen2001: 1.27 (1.15–1.40); without tiemersma2002: 1.27 (1.15–1.40); without flood2003: 1.30 (1.18–1.45); without wei2004_nhs: 1.28 (1.16–1.41); without wei2004_hpfs: 1.28 (1.16–1.41); without english2004: 1.27 (1.14–1.40); without larsson2005: 1.27 (1.14–1.41); without chao2005: 1.28 (1.14–1.42); without norat2005: 1.27 (1.15–1.41)

## CRC HvL – corrected: all classes

**Pooled: 1.21 (1.10–1.33)** (fixed 1.21 (1.10–1.33); I² 0 %; k = 13). 

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 3.2 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 2.5 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 2.9 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 4.3 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 0.80 (0.50–1.20) | 4.5 | exposure: unprocessed red meat Q4 vs Q1, multivariate |
| tiemersma2002 | 1.60 (0.90–2.90) | 2.5 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 11.2 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 1.21 (0.72–2.03) | 3.2 | Table I p2658: 876 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_hpfs | 1.24 (0.78–1.96) | 4.1 | Table I p2659: 602 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| english2004 | 1.40 (1.00–1.90) | 8.4 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.22 (0.98–1.53) | 17.5 | exposure: red meat main dish (unprocessed) ≥4 servings/week |
| chao2005 | 1.28 (1.04–1.58) | 20.5 | model: fixed-effect pool of colon and rectum as fn 6 states |
| norat2005 | 1.17 (0.92–1.49) | 15.0 | exposure: fresh red meat ≥80 vs <10 g/day, model ‡ |

Leave-one-out: without bostick1994: 1.22 (1.11–1.34); without kato1997: 1.21 (1.10–1.33); without chen1998: 1.21 (1.10–1.33); without singh1998: 1.20 (1.09–1.32); without pietinen1999: 1.23 (1.12–1.36); without tiemersma2002: 1.20 (1.09–1.32); without flood2003: 1.23 (1.11–1.35); without wei2004_nhs: 1.21 (1.10–1.33); without wei2004_hpfs: 1.21 (1.10–1.33); without english2004: 1.19 (1.08–1.32); without larsson2005: 1.21 (1.09–1.34); without chao2005: 1.19 (1.07–1.32); without norat2005: 1.22 (1.10–1.35)

## CRC HvL – corrected: errors only (model choices as published)

**Pooled: 1.20 (1.09–1.33)** (fixed 1.20 (1.09–1.33); I² 0 %; k = 13). 

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 3.7 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 2.9 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 3.4 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 5.0 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 0.80 (0.50–1.20) | 5.3 | exposure: unprocessed red meat Q4 vs Q1, multivariate |
| tiemersma2002 | 1.60 (0.90–2.90) | 3.0 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 13.1 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 1.21 (0.72–2.03) | 3.8 | Table I p2658: 876 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_hpfs | 1.24 (0.78–1.96) | 4.8 | Table I p2659: 602 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| english2004 | 1.40 (1.00–1.90) | 9.9 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.22 (0.98–1.53) | 20.5 | exposure: red meat main dish (unprocessed) ≥4 servings/week |
| chao2005 | 1.36 (0.93–2.00) | 6.9 | Table I p2659: 1667 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| norat2005 | 1.17 (0.92–1.49) | 17.5 | exposure: fresh red meat ≥80 vs <10 g/day, model ‡ |

Leave-one-out: without bostick1994: 1.21 (1.09–1.34); without kato1997: 1.20 (1.09–1.33); without chen1998: 1.20 (1.09–1.34); without singh1998: 1.19 (1.08–1.32); without pietinen1999: 1.23 (1.11–1.37); without tiemersma2002: 1.19 (1.08–1.32); without flood2003: 1.22 (1.09–1.36); without wei2004_nhs: 1.20 (1.09–1.33); without wei2004_hpfs: 1.20 (1.08–1.33); without english2004: 1.18 (1.06–1.32); without larsson2005: 1.20 (1.07–1.34); without chao2005: 1.19 (1.07–1.32); without norat2005: 1.21 (1.08–1.35)

## CRC HvL – corrected + follow-ups

**Pooled: 1.15 (1.05–1.25)** (fixed 1.15 (1.05–1.25); I² 0 %; k = 13). 

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 1.04 (0.62–1.76) | 2.6 | Table I p2658: 212 cases; total red meat (both reported) | from Fig 1 p. 2660 (render |
| kato1997 | 1.23 (0.68–2.22) | 2.0 | Table I p2658: 100 cases; excluded from type-of-meat strata (items not specified) | f |
| chen1998 | 1.17 (0.68–2.02) | 2.4 | Table I p2658: 212 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| singh1998 | 1.41 (0.90–2.21) | 3.6 | Table I p2658: 157 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| pietinen1999 | 0.80 (0.50–1.20) | 3.7 | exposure: unprocessed red meat Q4 vs Q1, multivariate |
| tiemersma2002 | 1.60 (0.90–2.90) | 2.1 | Table I p2658: 102 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| flood2003 | 1.10 (0.83–1.45) | 9.2 | Table I p2658: 487 cases; total red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| wei2004_nhs | 0.96 (0.76–1.21) | 13.3 | follow-up: NHS → Bernstein 2015 women >1 svg/d vs lowest, unprocessed red meat |
| wei2004_hpfs | 0.98 (0.75–1.27) | 10.3 | follow-up: HPFS → Bernstein 2015 men >1 svg/d vs lowest, unprocessed red meat |
| english2004 | 1.40 (1.00–1.90) | 7.0 | Table I p2659: 451 cases; fresh red meat only | from Fig 1 p. 2660 (rendered): RR (95 |
| larsson2005 | 1.22 (0.98–1.53) | 14.5 | exposure: red meat main dish (unprocessed) ≥4 servings/week |
| chao2005 | 1.28 (1.04–1.58) | 16.9 | model: fixed-effect pool of colon and rectum as fn 6 states |
| norat2005 | 1.17 (0.92–1.49) | 12.3 | exposure: fresh red meat ≥80 vs <10 g/day, model ‡ |

Leave-one-out: without bostick1994: 1.15 (1.05–1.26); without kato1997: 1.15 (1.05–1.25); without chen1998: 1.15 (1.05–1.25); without singh1998: 1.14 (1.04–1.24); without pietinen1999: 1.16 (1.07–1.27); without tiemersma2002: 1.14 (1.05–1.24); without flood2003: 1.15 (1.05–1.26); without wei2004_nhs: 1.18 (1.08–1.29); without wei2004_hpfs: 1.17 (1.07–1.28); without english2004: 1.13 (1.04–1.23); without larsson2005: 1.14 (1.04–1.25); without chao2005: 1.12 (1.02–1.23); without norat2005: 1.14 (1.04–1.26)

## DR reconstruction per 120 g – our GL values for Larsson's DR studies where gram doses exist

Published: 1.28 (1.18–1.39). **Pooled: 1.27 (1.10–1.46)** (fixed 1.27 (1.10–1.46); I² 0 %; k = 8). **Not reproduced.**

not reproducible from the paper; studies without gram-based categories in our table: chen1998, hsing1998, flood2003, wei2004, chao2005

| Study | RR (95 % CI) | Weight % | Source / choice |
|---|---|---|---|
| bostick1994 | 0.93 (0.36–2.35) | 2.4 | database GL/continuous per 100 g → per 120 g (colon, red_main_dish) |
| singh1998 | 2.99 (0.40–22.51) | 0.5 | database GL/continuous per 100 g → per 120 g (colon, red_undefined) |
| pietinen1999 | 0.80 (0.36–1.77) | 3.3 | database GL/continuous per 100 g → per 120 g (colorectal, red_unprocessed) |
| jarvinen2001 | 1.47 (0.90–2.39) | 8.7 | database GL/continuous per 100 g → per 120 g (colorectal, red_undefined) |
| tiemersma2002 | 2.19 (0.83–5.76) | 2.2 | database GL/continuous per 100 g → per 120 g (colorectal, red_unprocessed) |
| english2004 | 1.23 (0.87–1.71) | 17.9 | database GL/continuous per 100 g → per 120 g (colorectal, red_unprocessed) |
| larsson2005 | 1.28 (0.88–1.85) | 14.8 | database GL/continuous per 100 g → per 120 g (colorectal, red_main_dish) |
| norat2005 | 1.26 (1.02–1.54) | 50.2 | database GL/continuous per 100 g → per 120 g (colorectal, red_unprocessed) |

Leave-one-out: without bostick1994: 1.27 (1.10–1.47); without singh1998: 1.26 (1.09–1.45); without pietinen1999: 1.28 (1.11–1.49); without jarvinen2001: 1.25 (1.07–1.45); without tiemersma2002: 1.25 (1.08–1.45); without english2004: 1.27 (1.09–1.49); without larsson2005: 1.26 (1.08–1.48); without norat2005: 1.27 (1.04–1.56)

## Deviations (synthesis vs primary paper)

| Outcome | Study | Class | Deviation | Page | Alternative used |
|---|---|---|---|---|---|
| colorectal | hsing1998 | inclusion | Lutheran Brotherhood: fatal cases only (mortality); Larsson's own sensitivity analysis excluding mortality studies gives 1.27 (1.14–1.42) | Table I fn7; text p2660 | removed |
| colorectal | pietinen1999 | exposure | Larsson's row 203 vs 79 g/day is ATBC's red meat *including* processed (Chan 2011 S1; Pietinen Table 5); the unprocessed row (99 vs 35 g/day) gives 0.8 (0.5–1.2) | Pietinen 1999 Table 5 | 0.80 (0.50–1.20) – unprocessed red meat Q4 vs Q1, multivariate |
| colorectal | norat2005 | exposure | Larsson chose EPIC's total red meat (fresh + processed, ≥160 vs <20 g) 1.35 (0.96–1.88); the fresh red meat row (≥80 vs <10 g) is 1.17 (0.92–1.49) — as WCRF 2007 Fig 4.3.1 uses | Norat 2005 Table 3; Larsson Table I fn 4 (stated rule: total red meat preferred) | 1.17 (0.92–1.49) – fresh red meat ≥80 vs <10 g/day, model ‡ |
| colorectal | larsson2005 | exposure | SMC total red meat (incl. processed) ≥94 g vs <47 g 1.32 (1.03–1.68); the paper's unprocessed 'red meat as main dish' top category gives 1.22 (0.98–1.53) | Larsson 2005 Table II | 1.22 (0.98–1.53) – red meat main dish (unprocessed) ≥4 servings/week |
| colorectal | chao2005 | exposure | CPS-II Nutrition Cohort 'red meat' includes processed meat (beef, pork, lamb and processed); no unprocessed row exists | Chao 2005 Methods | none available (documented only) |
| colorectal | chao2005 | model | Larsson's CRC row for Chao, 1.36 (0.93–2.00), is a random-effects pool of Chao's colon 1.15 (0.90–1.46) and rectum 1.71 (1.15–2.52); footnote 6 says inverse-variance weighting, which as a fixed-effect pool gives 1.28 (1.04–1.58) | Larsson Table I fn 6; Chao 2005 Table 2 | 1.28 (1.04–1.58) – fixed-effect pool of colon and rectum as fn 6 states |
| colorectal | bostick1994 | exposure | Table II fn 4 lists Bostick as reporting both total and fresh red meat; Larsson's row 1.04 (0.62–1.76) is the paper's 'red meat' (>11 vs <4 servings/week) and the main-dish variable includes ham — neither is a clean unprocessed row | Larsson Table II fn 4; Bostick 1994 Table 3 | none available (documented only) |
| colorectal | flood2003 | exposure | BCDDP 'red meat' includes processed meat; no unprocessed row | Flood 2003 Methods | none available (documented only) |
| colorectal | singh1998 | cases | 157 colon cases include 22 rectosigmoid cases | Singh 1998 Methods | none available (documented only) |
| colorectal | chen1998 | inclusion | Physicians' Health Study nested case-control (212 cases); Larsson's sensitivity excluding nested case-control studies gives 1.28 (1.15–1.42) | Table I; text p2660 | none available (documented only) |
| colorectal | wei2004_nhs | exposure | 'red meat as main dish' excludes hamburger and mixed dishes (both NHS and HPFS) | Wei 2004 Table II | none available (documented only) |
| colorectal | jarvinen2001 | exposure | Järvinen's 'red meat' row is near-total meat (quartile cut-points 91–94 % of those for 'Meat, all'); Larsson's own Table II fn 4 excludes it from the type-of-meat strata (items not specified). Same deviation as in CUP 2010/2016 (correction R106) | Järvinen 2001 Tables 2–3; Larsson Table II fn 4 | removed |

## Notes

- Per-study per-120 g estimates (Table III, 1.28 (1.18–1.39), 14 studies) are not printed anywhere in the paper; the dose-response analysis cannot be reproduced from printed values (protocol R3). Our own GL per-100 g reconstruction for the same 14 studies is reported under 'DR reconstruction'.