Skip to main content

Table 1 Baseline characteristics of study participants according to sex-specific quartiles of ultra-processed food consumption (n = 70,421)a, b

From: Ultra-processed food and incident type 2 diabetes: studying the underlying consumption patterns to unravel the health effects of this heterogeneous food category in the prospective Lifelines cohort

 

Quartiles of ultra-processed food consumption

 
 

First (n = 17,604)

Second (n = 17,606)

Third (n = 17,606)

Fourth (n = 17,606)

Total (n = 70,421)

Age, years

52.3±9.1

50.2±8.8

48.3±8.4

45.7±7.6

49.1±8.8

Sex, %

 Women

58.6

58.6

58.6

58.6

58.6

 Men

41.4

41.4

41.4

41.4

41.4

Ultra-processed food intake, weight%

23.7 (20.3, 26.0)

31.6 (29.0, 34.1)

38.4 (35.6, 40.5)

48.7 (45.2, 53.9)

34.9 (28.1, 42.7)

Lifeline diet score

28.6±5.1

25.4±4.7

22.8±4.6

19.2±4.7

24.0±5.9

Total energy intake, kcal/day

1811±520

2032±543

2150±579

2261±647

2063±598

Total alcohol intake, grams/day

6.2 (1.4, 12.1)

5.8 (1.3, 11.4)

4.4 (1.0, 10.4)

2.9 (0.4, 9.9)

4.7 (0.9, 11.2)

Fasting glucose, mmol/L

4.96±0.51

4.96±0.50

4.96±0.50

4.97±0.51

4.96±0.50

HbA1c, %

5.56±0.30

5.56±0.30

5.54±0.30

5.54±0.30

5.55±0.30

HbA1c, mmol/mol

37.3±3.2

37.2±3.2

37.1±3.3

37.0±3.3

37.2±3.3

BMI, kg/m2

25.6±3.8

25.6±3.9

26.2±4.0

26.7±4.5

26.2±4.1

Highest tertile of PROCAM diabetes risk algorithm, %

37.2

33.2

31.9

30.4

33.2

MVPA, minutes/weekc

240 (90, 420)

210 (80, 380)

180 (60, 360)

150 (60, 330)

190 (60, 365)

Educational level, %

 Low

28.4

29.5

30.5

33.1

30.4

 Middle

34.8

37.7

40.1

42.8

38.9

 High

36.2

32.4

29.0

23.7

30.3

Smoking status, %

 Never

39.8

43.6

45.4

48.7

44.4

 Former

44.2

40.0

36.8

30.6

37.9

 Current

15.4

15.8

17.4

20.0

17.2

TV watching time, hours/day

2.4±1.3

2.4±1.3

2.5±1.2

2.6±1.4

2.5±1.3

  1. aData are expressed as unadjusted mean ± standard deviation for age, Lifelines diet score (no unit), total energy intake, fasting glucose, HbA1c, BMI, and TV watching time; data are expressed as median (interquartile) for ultra-processed food intake (weight percentage), total alcohol intake, and MVPA; data are expressed as observed percentage for sex, highest tertile of PROCAM diabetes risk algorithm, educational level, and smoking status
  2. bTests for significant differences in baseline characteristics (except for sex) across quartiles of ultra-processed food consumption were performed using Kruskal-Wallis test or χ2 test for proportion, as appropriate; P < 0.001 for all baseline characteristics except for fasting glucose P = 0.019
  3. cMVPA denotes non-occupational moderate-to-vigorous physical activity level