Association of sweetened beverages consumption with all-cause mortality risk among Dutch adults: the Lifelines Cohort Study (the SWEET project)

In this study, each additional serving/day of SSB was associated with a 9% higher all-cause mortality risk, which was most pronounced in women. For LNCB, each additional serving/day was associated with 6% higher all-cause mortality risk, but replacing SSB by LNCB was likely to be associated with a lower all-cause mortality risk particularly in women. Finally, although there is no strong evidence of nonlinear association, an inverse association between FJ and all-cause mortality was observed at moderate consumption (< 1 serving/day), but not in higher consumption levels.

Positive associations between SSB and all-cause mortality risk have also been reported in recent meta-analyses of cohort studies [4, 7, 8, 23], where particularly the largest studies support our findings [11, 12, 24,25,26,27]. A pooled prospective analysis of the Nurses’ Health Study and Health Professional Follow-up Study [12] yielded a 7% higher mortality risk for each additional serving/day. Comparing the highest consumption group (≥ 2 servings/day) vs reference (< 1/month) also showed a positive association that was more pronounced in women than in men (HR 1.25, 95% CI 1.16–1.34 vs 1.12, 95% CI 1.00–1.26) [12]. The sex-specific difference might be explained by physiologic differences between men and women, e.g., sex hormones and lipid profile, but further research is warranted [28,29,30]. However, not all studies are in line with our findings, which might be due to differences in study methodology, such as in SSB definition (i.e., including juice, added sugar) and age of included participants [27, 31,32,33,34,35].

Our data showed an association between LNCB and higher all-cause mortality risk. A recent meta-analysis by Pan et al. [8] including eight prospective cohort studies, also showed a higher all-cause mortality with LNCB consumption (HR 1.04, 95% CI 1.00–1.09 for each 250 ml/day). In line with our findings, several studies in this meta-analysis reported attenuation of positive associations after excluding those with diseases history, which may indicate the presence of reverse causation [12, 25]. We also identified other potential signs of reverse causation in our study, i.e., participants with higher LNCB consumption tended to have BMI ≥ 25 than participants with lower consumption. People might have switched to healthier diet once diagnosed with relevant risk factors to control their health, i.e., BMI. Taking this all together, our results need to be interpreted with caution.

We observed an inverse association between FJ and all-cause mortality at moderate intake level of < 1 serving/day, but not at higher consumption levels, when compared to no consumption. Similarly, the UK Biobank study [25] (n = 161,415) showed a 9% lower mortality risk when comparing those consuming ≤ 1 serving/day of FJ with non-consumers, which was not observed among those consuming > 1 serving/day. However, in the UK Biobank study, the association in the moderate consumption (≤ 1 serving/day) group disappeared after adjustment for diet quality. A recent meta-analysis of two prospective data by D’Elia et al. [10] suggested a nonlinear association between low to moderate 100% FJ consumption with stroke (≤ 200 ml/day) and CVD events (≤ 170 ml/day) (reference: no consumption), while no significant association was observed in the higher consumption category. In addition, Khan et al. [9] also demonstrated an inverse association between FJ consumption below 150 ml and CVD incidence, which was not present at higher consumption.

To date, limited studies investigated the replacement of SSB with LNCB or FJ and mortality risk [11,12,13]. In the present study, we observed nonsignificant 9%lower all-cause mortality risk when replacing SSB with LNCB in women, which is similar to previous findings in total population (HR 0.93, 95% CI 0.87–1.00) [11]. Pooled analysis of the Nurses’ Health Study and Health Professional Follow-up Study showed a 4% lower risk of all-cause mortality when replacing SSB with LNCB (HR 0.96, 95% CI 0.94–0.98) [12]. Various experimental studies also showed beneficial effects of replacing SSB with LNCB, especially with respect to weight loss or weight maintenance and some cardiometabolic profiling, i.e., body fat percentages and intrahepatocellular lipid [36]. In terms of replacing SSB with FJ, we observed no association with all-cause mortality, which is also in line with the previous study [11].

Adverse association between SSB and mortality risk can be explained by several biological mechanisms. SSB consumption may induce hepatic de novo lipogenesis, hyperuricemia, and insulin resistance by the high fructose content [37]. SSB consumption is also associated with decreased satiety and insufficient adjustment of energy reduction after liquid calories consumption compared to isocaloric solid food, which subsequently contribute to weight gain [5]. LNCB has been suggested to disturb the reward system, sweetness perception, and induce gut microbiota dysbiosis, which may lead to metabolic homeostasis disruption and insulin resistance [38,39,40]. However, evidence supporting these suggested undesirable effects is limited and more human experimental study is still needed [41]. Underlying mechanisms explaining the association between juice consumption and mortality risk are also not yet clear. FJ may contain a high amount of antioxidants (i.e., polyphenols) and other bioactive components (i.e., vitamin and mineral) [9, 10, 42] that can be beneficial for health, but food processing may affected antioxidant content [43]. Like SSB, FJ also has a high sugar content, which may counteract its benefits at higher consumption [9, 44].

Strengths of this study include its large sample size and long follow-up period, allowing for well-powered stratified analyses conducted in a unique three-generation design. Moreover, our study population is representative of the Dutch population in terms of socioeconomic, lifestyle, and disease prevalence [15]. The theoretical substitution analysis is another strength of this study as it provides insight into public health implications of using LNCB as an alternative for SSB. One of the limitations of our study is that we were unable to distinguish between various types (brands) of LNCB and FJ, which requires further study with respect to their potential differential impact on cardiometabolic health and mortality risks. In addition, dietary consumption was only assessed at baseline while repeated dietary assessment over time could have further reduced the potential of reverse causality and provided more precise risk estimates [12]. Furthermore, deaths due to non-chronic conditions might have attenuated all-cause mortality. However, in the Netherlands, the proportion of deaths in the Netherlands due to other than non-communicable disease was low (< 7%) [45, 46]. This proportion was higher (16%) in 2020 due to COVID-19, which was mainly among older people that also often suffered from co-existing non-communicable disease. Therefore, major attenuation of our findings due to death other than chronic conditions was unlikely. Finally, although we were able to account for a wide range of confounders, residual confounders cannot be excluded.

In conclusion, our findings suggest a positive association between SSB consumption and all-cause mortality risk, which was more pronounced in women than in men. Replacing SSB with LNCB might be associated with a lower all-cause mortality risk, particularly in women. Finally, an inverse association with all-cause mortality risk was observed at moderate consumption of FJ.

留言 (0)

沒有登入
gif