Baseline characteristics

After excluding participants with incomplete baseline data, a total of 48632 subjects, with a mean age of 52 years, were enrolled in the current investigation; Of whom, about 57.5% (n = 27974) were women and 42.5% (n = 20658) were men. After 13.8 years of follow-up on average, 10,774 deaths were recorded, of which 3,868 deaths were due to CVDs, 1,826 due to cancer (with 985 GI cancer death), and 3,160 death cases were due to other causes (e.g., respiratory and infectious diseases).

Table 1 describes the baseline characteristics and dietary intakes of studied participants according to the first and last quintiles of dietary fiber, fruits and vegetable. Compared to the subjects in the first quintile, individuals in the highest quintile of dietary fiber, fruits and vegetables were more likely to be men, younger, more rural, less hypertensive, or to have higher educational levels, wealth score (socio-economic status), and BMI; At the same time, they were more likely to smoke cigarettes, and be alcohol drinkers. Regarding the physical activity level and dietary intakes, the participants with higher fiber consumption tended to be more active, have higher daily intakes of energy and carbohydrate and lower daily intakes of total fat and total protein. The participants with higher fruits consumption tended to be less active, have higher daily intakes of energy, protein and total fat and lower intake of carbohydrate. Furthermore, those with higher vegetables consumption were more likely to be active and have greater daily intakes of energy and total fat and lower protein and carbohydrate consumption (Table 1).

Table 1 Baseline characteristics of studied subjects according to the first and last quintiles of dietary fiber, fruits and vegetables intakes in a large prospective cohort study with 14 years of follow-upDietary fiber intake and all-cause and cause-specific mortality

Tables 2, 3 and 4 presents the Cox multiple regression analysis findings on the association between dietary intakes of fiber, total fruits and vegetables, and hazard ratios (HRs) of all-cause and cause-specific mortality.

Table 2 Hazard ratios for all-cause and cause-specific mortality, according to quintiles of dietary fiber intake in a large prospective cohort study with 14 years of follow-upTable 3 Hazard ratios for all-cause and cause-specific mortality, according to quintiles of fruits intake in a large prospective cohort study with 14 years of follow-upTable 4 Hazard ratios for all-cause and cause-specific mortality, according to quintiles of vegetables intake in a large prospective cohort study with 14 years of follow-up

The median intake of fiber (13.66, 18.94, 22.49, and 25.67 g/d, and 31.37 g/d, respectively for quintiles 1 to 5), fruits (41.26, 81.74, 120.86, 175.81, and 300.05 g/d, and vegetables (98.16, 146.49, 183.05, 224.97, and 306.51g/d, respectively for quintiles 1 to 5), respectively for quintiles 1 to 5), are also illustrated in Tables 2, 3 and 4.

According to the age and gender-adjusted regression analysis (Model 1), the greater fiber consumption (median intake from about 19 to 31 g/d) was shown to be associated with a lower risk of all-cause mortality by 13- 21% (2nd quintile: HR = 0.87, 95%CI: 0.82, 0.92; 3rd quintile: HR = 0.82, 95%CI: 0.77, 0.87; 4th quintile: HR = 0.83, 95%CI: 0.78, 0.88; and 5th quintile: HR = 0.79, 95%CI: 0.74, 0.84), CVD mortality by 11–14% (2nd quintile: HR = 0.89, 95%CI: 0.80, 0.97; 3rd quintile: HR = 0.86, 95%CI: 0.78, 0.95; 4th quintile: HR = 0.89, 95%CI: 0.81, 0.99; and 5th quintile: HR = 0.88, 95%CI: 0.80, 0.98), and other cause mortality 21–23% (2nd quintile: HR = 0.79, 95%CI: 0.71, 0.87; 3rd quintile: HR = 0.79, 95%CI: 0.71, 0.88; 4th quintile: HR = 0.79, 95%CI: 0.70, 0.88; and 5th quintile: HR = 0.77, 95%CI: 0.69, 0.86), compared to the lowest fiber intake as the reference category (median intake ~ 14 g/d) (Table 2).

When taking into account the effects of additional confounding variables including educational level, ethnicity, BMI, cigarette smoking, opiate use, alcohol consumption, history of diabetes, history of hypertension, wealth score, residential area, and physical activity level in Model 2 of Cox regression analysis, similar findings were noted as those with greater dietary fiber intake were found to have a lower risk of all-cause by nearly 9–13% (2nd quintile: HR = 0.91, 95%CI: 0.86, 0.96; 3rd quintile: HR = 0.88, 95%CI: 0.83, 0.93; 4th quintile: HR = 0.91, 95%CI: 0.86, 0.97; and 5th quintile: HR = 0.87, 95%CI: 0.82, 0.92), and a reduced risk for other cause of death (2nd quintile: HR = 0.84, 95%CI: 0.75, 0.93; 3rd quintile: HR = 0.87, 95%CI: 0.78, 0.97; 4th quintile: HR = 0.88, 95%CI: 0.79, 0.99; and 5th quintile: HR = 0.86, 95%CI: 0.77, 0.96), when comparing with the first quintile of fiber intake (Table 2).

However, after additionally controlling for daily energy, and macronutrients intakes in the Model 3 of Cox regression analysis, only those in the second and third quintiles of dietary fiber intake showed a reduced risk of all-cause (by 7–10%) and other cause of death (by 15%-17%) (for all-cause mortality: 2nd quintile: HR = 0.93, 95%CI: 0.87, 0.99; and 3rd quintile: HR = 0.90, 95%CI: 0.83, 0.97; for other cause of death: 2nd quintile: HR = 0.83, 95%CI: 0.74, 0.94; and 3rd quintile: HR = 0.85, 95%CI: 0.74, 0.98), while the overall associations were not statistically significant (Table 2 and Fig. 1). The dose–response relationship was non-significant for all of the outcomes (Fig. 2).

Fig. 1

The association between dietary fiber intake (quintiles 2–5 versus 1) and all-cause and cause-specific mortality according to fully adjusted Cox regression analysis (Model 3 adjusted for gender, age, educational level, ethnicity, body mass index, cigarette smoking, opiate use, alcohol consumption, having a history of diabetes, having a history of hypertension, wealth score, residential area, physical activity level, and dietary data including daily energy intake (kcal/d), and energy-adjusted carbohydrate, protein and fat consumption)

Fig. 2

Association between different amounts of fiber intake and death from all-cause, CVD, all cancer, GI cancer, and other cause. CVD, cardiovascular disease; GI, gastrointestinal

Fruits intake and all-cause and cause-specific mortality

When total fruits consumption was explored, it was observed that according to Model 1, increasing intakes of fruits from the third to fifth quintiles (median intake from 121 to 300 g/d) was linked to a lower risk of all-cause mortality (3rd quintile: HR = 0.82, 95%CI: 0.77, 0.87; 4th quintile: HR = 0.83, 95%CI: 0.79, 0.88; and 5th quintile: HR = 0.77, 95%CI: 0.72, 0.81), all cancer mortality (3rd quintile: HR = 0.83, 95%CI: 0.72, 0.96; 4th quintile: HR = 0.81, 95%CI: 0.70, 0.94; and 5th quintile: HR = 0.76, 95%CI: 0.70, 0.94), GI cancer mortality (3rd quintile: HR = 0.71, 95%CI: 0.58, 0.87; 4th quintile: HR = 0.81, 95%CI: 0.67, 0.99; and 5th quintile HR = 0.75, 95%CI: 0.61, 0.91), CVD mortality (3rd quintile: HR = 0.82, 95%CI: 0.74, 0.90; 4th quintile: HR = 0.86, 95%CI: 0.78, 0.95; and 5th quintile HR = 0.87, 95%CI: 0.79, 0.96), and other cause mortality (3rd quintile: HR = 0.81, 95%CI: 0.73, 0.90; 4th quintile: HR = 0.88, 95%CI: 0.80, 0.98; and 5th quintile HR = 0.77, 95%CI: 0.69, 0.86) compared to those in the lowest quintile consuming about 41 g/d of fruits (Table 3).

In Model 2 of Cox regression analysis, additional adjustment for other confounding variables revealed that increasing the intake of fruits was accompanied by 10–13% reduced risk for all-cause mortality in comparison with the lowest intake (3rd quintile: HR = 0.88, 95%CI: 0.82, 0.93; 4th quintile: HR = 0.90, 95%CI: 0.84, 0.96; and 5th quintile HR = 0.87, 95%CI: 0.81, 0.93). Besides, the subjects in the last quintile of fruits demonstrated a lower risk of all cancer mortality by 15% (HR = 0.85, 95%CI: 0.73, 0.99) compared to those in the first quintile (Table 3).

Following taking into consideration the daily energy and macronutrients intakes in addition to other potentials variables in Cox regression Model 3, increasing consumption of fruits was associated with a decreased risk of mortality for all-cause mortality by 9–11% (3rd quintile: HR = 0.89, 95%CI: 0.83, 0.94; 4th quintile: HR = 0.91, 95%CI: 0.86, 0.97; and 5th quintile HR = 0.89, 95%CI: 0.83, 0.96), and all cancer cause mortality by 15–20% (3rd quintile: HR = 0.85, 95%CI: 0.73, 0.99; 4th quintile: HR = 0.83, 95%CI: 0.71, 0.98; and 5th quintile HR = 0.80, 95%CI: 0.67, 0.95) compared to the lowest intake. Also, a significant decreasing trend was observed across quintiles of fruits consumption for GI cancer mortality risk (Table 3 and Fig. 3).A significant dose response relationship was observed for death from all-cause, all cancer and GI cancer, and a marginally significant relationship for CVD mortality. The relationship between fruit intake and other cause of death was not dose–response (Fig. 4).

Fig. 3

The association between dietary fruits intake (quintiles 2–5 versus 1) and all-cause and cause-specific mortality according to fully adjusted Cox regression analysis (Model 3 adjusted for gender, age, educational level, ethnicity, body mass index, cigarette smoking, opiate use, alcohol consumption, having a history of diabetes, having a history of hypertension, wealth score, residential area, physical activity level, and dietary data including daily energy intake (kcal/d), and energy-adjusted carbohydrate, protein and fat consumption)

Fig. 4

Association between different amounts of fruit intake and death from all-cause, CVD, all cancer, GI cancer, and other cause. CVD, cardiovascular disease; GI, gastrointestinal

Vegetables intake and all-cause and cause-specific mortality

Increasing vegetables consumption from about 146 to 307 g/d as compared to 98 g/d was found to reduce the risk of mortality from all-cause (2nd quintile: HR = 0.88, 95%CI: 0.83, 0.93; 3rd quintile: HR = 0.84, 95%CI: 0.79, 0.89; 4th quintile: HR = 0. 81, 95%CI: 0.77, 0.86; and 5th quintile: HR = 0.78, 95%CI: 0.73, 0.82), CVD (2nd quintile: HR = 0.81, 95%CI: 0.74, 0.90; 3rd quintile: HR = 0.76, 95%CI: 0.69, 0.84; 4th quintile: HR = 0.76, 95%CI: 0.69, 0.84; and 5th quintile: HR = 0.83, 95%CI: 0.75, 0.91), and other cause (2nd quintile: HR = 0.88, 95%CI: 0.80, 0.98; 3rd quintile: HR = 0.85, 95%CI: 0.76, 0.94; 4th quintile: HR = 0.80, 95%CI: 0.72, 0.89; and 5th quintile: HR = 0.77, 95%CI: 0.69, 0.86, compared to the first quintile as the refrence category) (Table 4).

This inverse relationship remained significant after controlling for additional confounding variables in Model 2 of Cox regression analysis for all-cause and other cause mortality. The risk of all-cause mortality was reduced by almost 9–10% by increasing the intake of vegetables (3rd quintile: HR = 0.91, 95%CI: 0.86, 0.97; 4th quintile: HR = 0.90, 95%CI: 0.85, 0.95; and 5th quintile: HR = 0.90, 95%CI: 0.85, 0.96) in comparison with the lowest vegetables consumption. Furthermore, greater vegetables consumption showed a decreased risk of CVD mortality by 10–16%, and (3rd quintile: HR = 0.90, 95%CI: 0.81, 0.99; 4th quintile: HR = 0.84, 95%CI: 0.76, 0.93 and 5th quintile: HR = 0.84, 95%CI: 0.76, 0.93) for those in the fifth quintiles of vegetables intake compared to the first quintile. The risk of other cause of death was also demonstrated to reduce by 12% for those in the third and fourth quintiles of vegetables consumption in comparison with the lowest quintile (3rd quintile: HR = 0.88, 95%CI: 0.79, 0.98 and 4th quintile: HR = 0.88, 95%CI: 0.79, 0.99) (Table 4).

However, after considering the daily energy and macronutrients intake in the fully adjusted Cox regression model (Model 3), it failed to detect these associations as significant, except that those in the third and fourth quintiles of vegetables intake were found to have 11–12% lower risk of CVD mortality (3rd quintile: HR = 0.88, 95%CI: 0.80, 0.98; 4th quintile: HR = 0.89, 95%CI: 0.80, 0.99; compared to the first quintile as the reference category) (Table 4 and Fig. 5). No significant dose–response relationship was observed between vegetable intake and death from all-cause, CVD, all cancer, and other cause. Only a marginally significant dose–response relationship was observed for GI cancer (Fig. 6).

Fig. 5

The association between dietary vegetables intake (quintiles 2–5 versus 1) and all-cause and cause-specific mortality according to fully adjusted Cox regression analysis (Model 3 adjusted for gender, age, educational level, ethnicity, body mass index, cigarette smoking, opiate use, alcohol consumption, having a history of diabetes, having a history of hypertension, wealth score, residential area, physical activity level, and dietary data including daily energy intake (kcal/d), and energy-adjusted carbohydrate, protein and fat consumption)

Fig. 6

Association between different amounts of vegetable intake and death from all-cause, CVD, all cancer, GI cancer, and other cause. CVD, cardiovascular disease; GI, gastrointestinal

Sensitivity analysis

The sensitivity analysis indicated that according to the multivariable Cox regression Model 3, the inverse association between higher intake of dietary fiber (2nd and 3rd quintiles versus 1st quintile) and lower all-cause mortality risk became statistically non-significant following excluding the patients with a history of chronic diseases,: 2nd quintile: HR = 0.93, 95%CI: 0.84, 1.02; and 3rd quintile: HR = 0.90, 95%CI: 0.80, 1.01 (Supplementary Table 1). Similar results were observed when Smokers, opium users and alcohol drinkers were excluded; such that the significant relationship in the 2nd and 3rd quintiles became non-significant (2nd quintile: HR = 1.00, 95%CI: 0.92, 1.09; and 3rd quintile: HR = 0.98, 95%CI: 0.89, 1.09) (Supplementary Table 1).

The multivariable-adjusted HRs for all-cause mortality according to quintiles of fruits and vegetables consumption remained consistent after we excluded the patients with chronic diseases, extreme BMI, those with a history of smoking, opium or alcohol use, or the subjects who were followed-up in the first 2 years of study (Supplementary Tables 2 and 3).