Skip to main content

Table 3 Threshold effect analysis of WWI, BMI, and WC on all-cause mortality by two-piecewise linear regression model

From: Association of weight-adjusted waist index with all-cause mortality among non-Asian individuals: a national population-based cohort study

Variables

number of participants/ number of events

Person-Years (years)

Adjusted HR (95%CI), P-value

Fitting by the standard linear model (WWI)

1,065/18,592

90311.67

1.13 (1.03, 1.25)

Fitting by the two-piecewise

linear model

   

Inflection point (cm/√kg)

  

10.46

WWI < 10.46 (cm/√kg)

94/4,621

23816.83

0.62 (0.41, 0.94)

WWI > 10.46 (cm/√kg)

971/13,971

66494.83

1.20 (1.08, 1.33)

Log likelihood ratio

  

0.007

Fitting by the standard linear model (BMI)

1,065/18,592

90311.67

0.98 (0.97, 0.99)

Fitting by the two-piecewise

linear model

   

Inflection point (kg/m2)

  

28.50

BMI < 28.50 (kg/m2)

583/9,246

45673.00

0.92 (0.90, 0.94)

BMI > 28.50 (kg/m2)

472/9,326

44562.83

1.02 (1.00, 1.03)

Log likelihood ratio

  

< 0.001

Fitting by the standard linear model (WC)

1,065/18,592

90311.67

0.99 (0.99, 1.00)

Fitting by the two-piecewise

linear model

   

Inflection point (cm)

  

101.40

WC < 101.40 (cm)

528/10,233

50935.00

0.98 (0.97, 0.98)

WC > 101.40 (cm)

537/8,359

39376.67

1.01 (1.00, 1.01)

Log likelihood ratio

  

< 0.001

  1. Age, gender, race, education level, ratio of family income to poverty, alcohol drinking status, smoking status, stroke status, coronary heart disease status, renal failure status, liver diseases status, myocardial infarction status, and cancer status were adjusted
  2. Abbreviation Q, quartile; WWI, weight-adjusted-waist index; BMI, body mass index; WC, waist circumference