Table 3.
Adjusted direct and indirect associations of three types of beverages with CKD mediated via metabolic syndrome (MetS)a.
| Sugar-sweetened beverages unit/day | Artificially sweetened beverages unit/day | Natural juices unit/day | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 0 | 0–1 | >1 | 0 | 0–1 | >1 | 0 | 0–1 | >1 | |
| MetS | |||||||||
| Total associationa | Ref | 1.04 [0.89, 1.21] | 1.42 [1.02, 1.98] | Ref | 1.44 [1.19, 1.74] | 1.59 [1.26, 2.01] | Ref | 0.88 [0.77, 1.00] | 1.03 [0.81, 1.30] |
| Natural direct association | 1.03 [0.88, 1.20] | 1.37 [0.97, 1.94] | 1.38 [1.12, 1.69] | 1.48 [1.14, 1.93] | 0.88 [0.77, 1.00] | 1.01 [0.79, 1.29] | |||
| Natural indirect association | 1.01 [1.00, 1.02] | 1.04 [1.00, 1.08] | 1.04 [1.01, 1.07] | 1.07 [1.02, 1.13] | 1.00 [1.00, 1.00] | 1.01 [1.00, 1.03] | |||
| Proportion mediated (%)b | NA | 12.5 | 13.2 | 18.0 | 0.6 | NA | |||
| MetS + hyperuricemia | |||||||||
| Total associationa | Ref | 1.06 [0.90, 1.24] | 1.47 [1.05, 2.07] | Ref | 1.44 [1.18, 1.75] | 1.58 [1.24, 2.00] | Ref | 0.87 [0.76, 1.00] | 1.06 [0.83, 1.34] |
| Natural direct association | 1.04 [0.89, 1.23] | 1.42 [0.99, 2.04] | 1.37 [1.11, 1.70] | 1.46 [1.11, 1.92] | 0.87 [0.76, 1.00] | 1.04 [0.81, 1.33] | |||
| Natural indirect association | 1.01 [1.00, 1.02] | 1.04 [1.00, 1.07] | 1.05 [1.02, 1.08] | 1.08 [1.02, 1.14] | 1.00 [0.99, 1.00] | 1.01 [1.00, 1.02] | |||
| Proportion mediated (%)b | NA | 10.9 | 14.7 | 20.1 | 1.0 | NA | |||
The HRs and 95% CI of the total effect, natural direct effect, and natural indirect effect were calculated using the VanderWeele counterfactual-framework approach (31, 32). The models were adjusted for sex, age, race, education levels, household income, socioeconomic status, smoking status, alcohol consumption, physical activity, fruit and vegetable intake, red and processed meat intake, sleep time, medications use (aspirin, non-aspirin NSAIDs, and proton pump inhibitors use), total sugar intake, total energy, and mutually adjusted for another two beverages. Total effect estimates may vary between models for different mediators due to differences in the number of missing values for each mediator and exposure-mediator interactions.
Proportion mediated not given in cases where the null effect (i.e., 1) is contained in the 95% CI of the HR of the total effect.