
==== Front
BMC Med
BMC Med
BMC Medicine
1741-7015
BioMed Central London

39232779
3586
10.1186/s12916-024-03586-7
Research Article
Effect of salt substitution on fracture—a secondary analysis of the Salt Substitute and Stroke Study (SSaSS)
Wang Faxuan 12
Pi Yangyang 1
http://orcid.org/0000-0002-0958-0350
Zhao Yi zhaoyi751114@hotmail.com

12
Zhang Yuhong 1
Zhou Bo 3
Li Zhifang 4
Sun Jixin 5
Yu Yan 6
Tian Maoyi 7
Yang Minghui 8
Huang Liping 9
Song Hongyi 10
Neal Bruce 911
Kissock Katrina R. 9
1 https://ror.org/02h8a1848 grid.412194.b 0000 0004 1761 9803 School of Public Health, Ningxia Medical University, Yinchuan, Ningxia China
2 https://ror.org/02h8a1848 grid.412194.b 0000 0004 1761 9803 NHC Key Laboratory of Metabolic Cardiovascular Diseases Research, Ningxia Medical University, Yinchuan, Ningxia China
3 https://ror.org/04wjghj95 grid.412636.4 Department of Evidencebased Medicine, First Hospital of China Medical University, Shenyang, China
4 https://ror.org/0340wst14 grid.254020.1 0000 0004 1798 4253 School of Public Health, Changzhi Medical College, Changzhi, China
5 https://ror.org/04hsdam77 grid.417579.9 0000 0004 0627 9655 Department of Noncommunicable Disease Prevention and Control, Center for Disease Control of Hebei Province, Shijiazhuang, China
6 https://ror.org/017zhmm22 grid.43169.39 0000 0001 0599 1243 School of Public Health, Xi’an Jiaotong University School of Medicine, Xi’an, China
7 https://ror.org/05jscf583 grid.410736.7 0000 0001 2204 9268 School of Public Health, Harbin Medical University, Harbin, China
8 grid.24696.3f 0000 0004 0369 153X Department of Orthopaedics and Traumatology, Beijing Jishuitan Hospital, Capital Medical University, Beijing, China
9 grid.1005.4 0000 0004 4902 0432 The George Institute for Global Health, University of New South Wales, Sydney, NSW Australia
10 https://ror.org/05e1zqb39 grid.452860.d The George Institute for Global Health China, Beijing, China
11 https://ror.org/041kmwe10 grid.7445.2 0000 0001 2113 8111 School of Public Health, Imperial College London, London, UK
4 9 2024
4 9 2024
2024
22 3662 12 2023
27 8 2024
© The Author(s) 2024
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Background

Associations of dietary sodium and potassium intake with fracture risk are inconsistent and the effects of salt substitute on fracture incidence are unknown. We assessed the effect of salt substitute compared to regular salt intake on fracture incidence using data from the Salt Substitute and Stroke Study (SSaSS).

Methods

SSaSS was a cluster-randomized controlled trial conducted in 600 villages in northern China. Villages were randomly allocated into intervention and control groups in a 1:1 ratio. Salt substitute was provided to intervention villages and control villages continued regular salt use for 5 years. The primary outcome for this secondary analysis was the incidence of all fractures. Secondary outcomes included incidence of vertebral fracture, non-vertebral fracture, and fracture of unknown or non-specific location.

Results

20,995 participants were included in this study, and 821 fractures occurred during follow-up. Intention-to-treat analyses showed no differences between the salt substitute and regular salt groups in the incidence of all fractures (rate ratio (RR) 0.96; 95% CI 0.81 to 1.14), vertebral fracture (RR 0.82; 95% CI 0.53 to 1.26), non-vertebral fracture (RR 1.05; 95% CI 0.86 to 1.29), or fracture of unknown or non-specific location (RR 0.80; 95% CI 0.54 to 1.18).

Conclusions

Use of salt substitute compared to regular salt had no detectable effect on the incidence of fracture in a population at high risk of cardiovascular disease and fracture.

Trial registration

ClinicalTrials.gov, NCT02092090. Registered on March 12, 2014.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12916-024-03586-7.

Keywords

Low-sodium salt
Salt substitute
Fracture
Osteoporosis
Randomized-controlled trial
issue-copyright-statement© BioMed Central Ltd., part of Springer Nature 2024
==== Body
pmcBackground

The global population is aging with an increase in the number and proportion of older adults [1], and this has led to a rise in fractures associated with fragility due to reduced bone mineral density (BMD) and microstructural degeneration of bone tissue [2]. In 2019, there were 178 million new fractures globally, an increase of 33.4% from 1990 [3]. Fractures in older adults pose a serious economic and disease burden [3–6], causing increased risks of disability and mortality and an increased need for outpatient and inpatient medical services [7].

Reduced BMD is a leading cause of fracture, due to increases in bone fragility that compounds the increased fracture risk attributed to falls. BMD is associated with nutritional, hormonal, lifestyle, genetic, and environmental factors [8]. In regard to nutrition, increased dietary calcium intake is associated with reduced age-related bone loss and reduced risks of fragility fractures [9]. High sodium chloride (regular salt) intake may be associated with reduced BMD because sodium competes with calcium for reabsorption by the kidneys [10]. Several cross-sectional [11] and animal [12] studies show an inverse relationship between sodium intake and BMD, though longitudinal studies show no association of sodium intake with BMD [8]. On the other hand, studies have found that bones release sodium during prolonged hyponatremia leading to bone loss at the cellular level and osteoporosis due to increased bone turnover [13–15]. Higher levels of potassium intake appear to improve calcium homeostasis and reduce bone catabolism and are associated with higher BMD levels [16–18].

It is also possible that reducing sodium intake and increasing potassium intake could cause fractures through excessive blood pressure lowering in susceptible individuals that leads to falls [19]. This is a consideration that has also been raised for blood pressure-lowering drugs more broadly, though adverse effects of blood pressure-lowering drugs on fracture risk are inconsistent [20, 21].

Given the very positive results of the Salt Substitute and Stroke Study (SSaSS) for cardiovascular protection, and subsequent likely wider use of potassium-enriched salt over coming years, it is important to understand the balance of risks and benefits for fracture risk and how positive effects on cardiovascular disease (CVD) might trade off against possible harms through increased risks of fractures [12, 22]. To the knowledge of the authors, there is no research exploring the effect of salt substitute on fracture risk. While SSaSS was designed primarily to determine the effect of salt substitute on the risk of stroke, it also collected data on all other serious adverse events including fracture. Accordingly, the aim of the current study was to assess the effect of salt substitute versus regular salt intake on fracture incidence over a 5-year period in a population at high risk for CVD.

Methods

Study design

This is an ad hoc analysis using data from SSaSS. SSaSS was an open-label, cluster-randomized trial conducted in 20,995 participants from 600 villages across five northern provinces in China, including Hebei, Liaoning, Ningxia, Shanxi, and Shaanxi. Each village recruited about 35 individuals at elevated risk of stroke. The trial commenced in 2013 and concluded in 2021. The detailed design of the study has been described previously [23, 24].

Written informed consent was obtained from all participants and the trial was registered with ClinicalTrials.gov (NCT02092090) and approved by ethics committees at the Peking University Health Science Center in China and the University of Sydney in Australia. The George Institute for Global Health sponsored the trial and completed the data analysis.

Participants

Individuals were eligible to participate if they had a previous history of stroke or were aged 60 years or older and had uncontrolled high blood pressure (SBP ≥ 140 mmHg if receiving blood pressure lowering medication; SBP ≥ 160 mmHg if not). Participants were required to be able to be contacted by phone or through nominated relatives or friends.

Individuals were excluded if they or someone living in the household was using potassium-sparing diuretics or potassium supplements, had a severe renal impairment, or had other reasons to be concerned about the use of salt substitute. Participants whose life expectancy was expected to be < 6 months or who ate the majority of meals outside the home were also excluded. Only one family member from a household was included.

Randomization

After completion of the baseline survey for all participants, villages were randomly assigned in a 1:1 ratio through a central computerized process to either the intervention group, in which the participants used a salt substitute, or the control group, in which the participants continued to use regular salt. Participants were not blinded to the intervention.

Procedures

Participants in the intervention group were provided with sufficient quantities of salt substitute to cover all salt use at home for the entire household free of charge. The salt substitute was produced according to Chinese national standards comprising 75% ± 10% sodium chloride and 25% ± 10% potassium chloride. Use of salt substitute is considered generally safe but can lead to hyperkalemia if consumed in very large amounts (e.g., more than 50 g/day) or in the presence of serious kidney disease [23]. Recommendations to reduce salt intake were provided to all villages at the start of the trial.

Participants were followed up every 6 months throughout the trial to collect data about all events leading to hospitalization (including fracture) or death as well as other serious adverse events. Follow-up visits during the first 24 months and the final 5-year follow-up were conducted face-to-face. Follow-up after 24 months and up to the final 5-year follow-up were conducted through linkage to the administrative databases of the New Rural Cooperative Medical Scheme (NCMS) and the National Mortality Surveillance System which together covers hospital admissions for 98% of the rural population in China. Data linkage was also completed retrospectively for the first 24 months of the trial and at the final visit to collect all hospitalization and death events since trial commencement. All events were coded using the Medical Dictionary for Regulatory Activities (MedDRA) to the lowest level term. Where possible, MedDRA coding was done automatically; however, manual coding was employed by a clinician familiar with MedDRA where there were no matching terms. Once a term was coded, all future diagnoses of the same nature were coded with the same MedDRA terms for consistency. Investigators and outcome assessors were blinded to participant allocation.

Outcomes

The primary outcome of interest in this paper was the incidence of all fractures. The key secondary outcomes were the incidence of vertebral fracture, non-vertebral fracture, and fracture of unknown or non-specific location.

All fracture events were identified using hospitalization data searched for relevant MedDRA codes. Within the hierarchy of MedDRA coding, any preferred term including “fracture” was included. Preferred terms including “fracture” were found under the “Injury, poisoning and procedural complications” and the “Musculoskeletal and connective tissue disorders” system organ classes. All fractures were classified into fracture types and mapped to vertebral fracture, non-vertebral fracture and fracture of unknown or non-specific location classification. Vertebral fractures were defined as fracture events where the MedDRA preferred term included the words “vertebral” or “spinal.” Non-vertebral fractures were defined as all fracture events excluding vertebral fractures and fractures of unknown or non-specific location. Fractures of unknown or non-specific location were defined as fracture events where the MedDRA preferred term was “fracture,” “compression fracture,” “multiple fractures,” or “osteoporotic fracture.” More detail around relevant MedDRA codes for fracture is provided in the Additional file.

Statistical analysis

All statistical analyses applied the same statistical methodology as used for the main trial analyses [24] and were performed using an intention-to-treat approach. Baseline characteristics were summarized as mean (SD) for continuous variables and counts and proportions (%) for categorical variables by randomized group and overall.

The effect of the intervention on the total incidence of the primary and secondary outcomes were analyzed using a hierarchical Poisson regression model with adjustment for clustering at the village level as a random intercept and follow-up time as an offset. Rate ratios and 95% CIs were calculated. All events of fracture were included in this analysis, which was based on rates, not time to the first event. Cumulative time-to-event curves by randomized group for the primary and secondary outcomes were generated with the use of the Kaplan–Meier method, with censoring at time of death or end of last follow-up. All time-to-event curves were based on the first recorded fracture event for an individual during the trial period that was relevant to each analysis. Specifically, participants with more than one fracture recorded throughout the trial period contributed only once to the analysis of all fractures but might have contributed to multiple analyses of different fracture types. Subgroup analyses to explore differences in intervention effects between participant populations were based on characteristics pre-defined in the main trial statistical analysis plan, including age, sex, education, history of diabetes, history of hypertension, use of antihypertensive medication, systolic blood pressure, diastolic blood pressure, and body mass index. Statistical tests were two-tailed, and a p-value < 0.05 was considered statistically significant. All statistical analyses were conducted with R and RStudio (version 4.2.2).

Results

Participant characteristics

Between April 2014 and January 2015, there were 20,995 participants enrolled from 600 villages and randomly assigned at the village level to the salt substitute group (n = 10,504) or the control (regular salt) group (n = 10,491). The overall mean and median durations of follow-up were 4.74 years and 5.12 years, respectively. Vital status at trial conclusion was verified for all participants; there were 8,535 and 8,288 people alive at the end of the study in the intervention and control groups, respectively (Fig. 1).Fig. 1 Flow of participants during the trial including enrolment, randomization, and follow-up

The demographic and clinical characteristics of the participants at baseline were similar between the randomized groups (Table 1). Among participants, 72.6% had a history of stroke, and 59.3% were > 60 years old and had hypertension. The mean age was 65.4 ± 8.5 years, and 49.5% were female. Mean systolic blood pressure was 154.0 ± 23.5 mmHg, and diastolic blood pressure was 89.2 ± 14.0 mmHg. The mean BMI was 24.8 ± 3.6; 33.5% of participants smoked in the past, and 18.8% were current smokers. Most participants had an education level of primary school or below (72.5%). A high proportion of participants had a history of hypertension (88.4%), followed by ischemic heart disease (16.0%) and transient ischemic attack (14.1%). Overall, 79.3% of participants were taking antihypertensive medication, while 12.3% were taking lipid-lowering medication and 39.7% were taking aspirin or other antiplatelet agents. Table 1 Baseline characteristics of participants by randomized group

	Intervention
(n = 10,504)	Control
(n = 10,491)	Overall
(n = 20,995)	
Inclusion criteria (%)	History of prior stroke with diagnosis made in hospital	73.3	71.9	72.6	
Age > 60 years and two measures of SBP ≥ 140 mmHg and using BP lowering treatment or two measures of SBP ≥ 160 mmHg and not using BP lowering treatment	59.4	59.2	59.3	
Sex (%)	Female	49.7	49.2	49.5	
Male	50.3	50.8	50.5	
Age

mean (SD)

		65.2 (8.5)	65.5 (8.5)	65.4 (8.5)	
BMI

mean (SD)

		24.8 (3.6)	24.9 (3.7)	24.8 (3.6)	
SBP

mean (SD)

		153.8 (23.37)	154.2 (23.6)	154.0 (23.5)	
DBP

mean (SD)

		89.1 (14.0)	89.4 (14.0)	89.2 (14.0)	
Smoking (%)	Past smoker	32.8	34.2	33.5	
Current smoker	18.7	18.9	18.8	
Level of education (%)	Primary school or lower	72.3	72.6	72.5	
Junior high school	23.2	22.4	22.8	
Senior high school or above	4.5	5.0	4.7	
Disease history (%)	Transient ischemic attack	13.2	14.9	14.1	
Ischemic heart disease	16.4	15.7	16.0	
Congestive heart failure	2.5	2.5	2.5	
Peripheral arterial disease	5.1	3.6	4.4	
Hypertension	88.6	88.3	88.4	
Diabetes mellitus	10.6	10.5	10.6	
Medication (%)	Diuretic	11.5	11.2	11.3	
ACE inhibitor or ARB	23.0	23.0	23.0	
α-blocker	0.7	1.1	0.9	
β-blocker	5.0	6.3	5.6	
Calcium antagonist	43.3	40.0	42.0	
Any of the above medications	59.1	58.7	58.9	
Other antihypertensive agents	38.2	37.5	37.8	
Any antihypertensive medication	79.9	78.7	79.3	
Statin or other lipid-lowering agents	12.6	12.0	12.3	
Aspirin or other antiplatelet agents	40.0	39.4	39.7	
Abbreviations: BMI body mass index, BP blood pressure, DBP diastolic blood pressure, SBP systolic blood pressure

Outcome events

A total of 821 fractures occurred over the duration of the trial, of which 132 were vertebral fractures, 540 were non-vertebral fractures, and 149 were fractures of unknown or non-specific location. Six hundred sixty-six participants had one fracture event, and 72 participants had two or more fracture events (Table 2). Table 2 Number of fracture events during the intervention period

	n	
All fractures	821	
 Vertebral fracture	132	
    Cervical vertebral fracture	1	
    Lumbar vertebral fracture	78	
    Thoracic vertebral fracture	41	
    Spinal fracture of unknown or non-specific locationa	12	
 Non-vertebral fracture	540	
    Ankle fracture	37	
    Clavicle fracture	25	
    Femoral neck fracture	41	
    Femur fracture	166	
    Fibula fracture	4	
    Foot fracture	15	
    Forearm fracture	3	
    Fractured Ischium	1	
    Hand fracture	11	
    Hip fracture	9	
    Humerus fracture	39	
    Ilium fracture	1	
    Jaw fracture	1	
    Lower limb fracture	21	
    Patella fracture	25	
    Pelvic fracture	5	
    Pubis fracture	14	
    Radius fracture	38	
    Rib fracture	43	
    Scapula fracture	2	
    Skull fractured base	1	
    Sternal fracture	2	
    Tibia fracture	24	
    Ulna fracture	5	
    Upper limb fracture	6	
    Wrist fracture	1	
 Fracture of unknown or non-specific location	149	
    Compression fracture	2	
    Fracture	130	
    Multiple fractures	5	
    Osteoporotic fracture	12	
n for all participants

Some participants had more than one fracture event throughout the intervention period

aCombined “spinal fracture” and “spinal compression fracture” MedDRA terms

There was no detectable effect of salt substitute compared to regular salt on the risk of all fractures (RR 0.96, 95% CI 0.81 to 1.14), vertebral fracture (RR 0.82, 95% CI 0.53 to 1.26), non-vertebral fracture (RR 1.05, 95% CI 0.86 to 1.29), or fracture of unknown or non-specific location (RR 0.80, 95% CI 0.54 to 1.18) (Figs. 2 and 3).Fig. 2 Effects of salt substitute on all fractures and fracture sub-types

Fig. 3 Cumulative incidence of fracture classifications over the 5-year intervention period. Cumulative incidence curves using the Kaplan–Meier method of all fractures (A), vertebral fracture (B), non-vertebral fracture (C), and fracture of unknown or non-specific location (D)

Subgroup analysis

There was no detectable modification of age, sex, education, history of diabetes, history of hypertension, use of antihypertensive medication, blood pressure, or body mass index on the association between salt substitute and all fracture risk (all p for heterogeneity ≥ 0.05) (Table 3). Table 3 Subgroup analysis of the effects of salt substitute on all fractures

	Regular salt
(N = 10,491)	Salt substitute
(N = 10,504)	Rate ratio (95% CI)a	p value for heterogeneity	
No. of events per 1000 person-years	
Subgroups	
Age	
 ≤ 65 years	6.95	6.65	0.97	0.76 to 1.23	0.96	
 > 65 years	10.12	9.72	0.96	0.77 to 1.20	
Sex	
 Female	11.24	10.64	0.96	0.79 to 1.17	0.93	
 Male	5.62	5.41	0.98	0.74 to 1.30	
Education	
 Primary school or lower	9.30	9.33	1.01	0.84 to 1.21	0.17	
 Junior high school or higher	6.21	4.78	0.77	0.52 to 1.15	
History of diabetes	
 Yes	9.04	8.39	0.90	0.55 to 1.48	0.92	
 No	8.38	8.02	0.96	0.80 to 1.15	
History of hypertension	
 Yes	8.50	8.23	0.97	0.81 to 1.16	0.54	
 No	8.00	6.68	0.92	0.50 to 1.72	
Antihypertensive use	
 Yes	8.68	8.69	1.00	0.83 to 1.20	0.11	
 No	7.54	5.48	0.79	0.51 to 1.23	
Systolic blood pressure	
 ≤ 153 mm Hg	7.60	7.12	0.94	0.74 to 1.20	0.75	
 > 153 mm Hg	9.33	9.05	0.97	0.78 to 1.22	
Diastolic blood pressure	
 ≤ 89 mm Hg	8.08	8.61	1.07	0.86 to 1.33	0.12	
 > 89 mm Hg	8.81	7.49	0.86	0.67 to 1.10	
Body-mass index	
 ≤ 24.6	8.83	7.93	0.89	0.70 to 1.13	0.28	
 > 24.6	8.08	8.18	1.02	0.82 to 1.27	
aAll non-significant (p ≥ 0.05)

Discussion

In this large-scale cluster randomized controlled trial, consumption of salt substitute compared with regular salt had no detectable effect on the incidence of fracture in participants at high risk of cardiovascular disease during the 5-year trial period. This was true for vertebral fractures which often have a fragility cause as well as non-vertebral fractures which are often caused by trauma. These observations and their related likely cause of fracture suggest that the null overall effect was not a consequence of balanced effects of protection against fragility fractures through enhanced bone metabolism or traumatic fractures due to hypotension-related falls or other trauma; rather, it appears that there was no detectable effect for either mechanism. While the absence of a beneficial effect on fracture may be viewed as disappointing, the absence of any adverse effect on fracture is a key observation. The very positive finding for cardiovascular protection observed in the main report from SSaSS will likely result in largely expanded use of salt substitute for cardiovascular protection among elderly and frail patients. As such, it is important for clinicians and policymakers to know that advocating the use of salt substitute for cardiovascular prevention will not be achieved at the expense of increased fracture risk [24, 25].

In regard to possible metabolic impacts of salt substitute on fracture, multiple studies have shown that higher urinary sodium excretion (a proxy for dietary sodium (salt) intake) is associated with reduced BMD in pre- and post-menopausal women, reduced bone mineral content, and positively associated with an increased risk of osteoporosis [11, 22, 26]. Another smaller study found that increased urinary sodium excretion was strongly associated with increased markers of bone resorption providing a plausible mechanism for protective effects of dietary sodium reduction on fracture risk [27]. However, BMD, bone mineral content, and osteoporosis are intermediate markers for fracture risk, and all are affected by multiple different factors. Observational studies are also prone to confounding and the absence of an association between dietary sodium intake and fracture risk in another large study may reflect the methodological limitations of this type of research [28]. There is similar uncertainty about the effects of dietary potassium intake on bone metabolism and fracture risk. There are some limited data suggesting supplementary potassium intake may increase BMD and bone mineral content [16, 29, 30], but no studies show clear associations with fracture risk [29]. On balance, the findings of no detectable change in fracture risk with salt substitute in the current study are consistent with the prior data which have shown inconsistent associations of sodium and potassium with markers of fracture risk and no clear associations with fracture itself [31, 32].

Observational data have also identified associations of high blood pressure with increased fracture risk that may be attributed to reductions in BMD [33], while other studies show an association independent of BMD [34]. It has also been noted that blood pressure lowering following initiation of antihypertensive therapy is a risk for fracture which is believed to be mediated by dizziness, fainting, and alterations in postural balance, leading to an increased risk of falls [20]. On this basis, salt substitute could have increased the risk of falls and fracture through its blood pressure-lowering effect [25]. However, the associations of blood pressure lowering and fracture risk are all derived from observational analyses, while secondary investigations of multiple randomized trials of blood pressure lowering therapies compared to control have not identified any increased risk of fracture.

The present study was done among a study population consisting mostly of older people, about half of which were post-menopausal women and many with cerebrovascular disease who would be susceptible to both falls and fragility fractures. It is among these populations that any adverse effect of salt substitute on fracture would be expected to be identified. However, the null result for fracture is highly reassuring in regard to the use of salt substitute for cardiovascular protection within these same populations.

There was no detectable difference in the effects of salt substitute on fracture risk across the participant subgroups studied, including for groups defined by sex and body mass index. Post-menopausal women are at particularly high risk of low BMD and fracture, but while there was an anticipated higher incidence of fracture among females compared to males [35], there was no evidence that the effects of salt substitute on fracture risk varied by sex. Malnutrition, which may be indicated by body mass index, is another key risk factor for falls and fractures among older adults [36, 37], mediated by muscle wastage, bone fragility, and gait instability [38, 39]. There was, however, no evidence of differences in fracture incidence or the effects of salt substitute on fracture risk between participant subgroups defined on the basis of body mass index.

This analysis benefits from being based upon data from a large-scale randomized trial such that it is likely to have had adequate power to detect plausible effects of salt substitute on total fracture risk; however, power would have been reduced for subgroup analyses of fracture type. The study population included key subsets of interest including females and the elderly. The randomized design means that the results are unlikely to be importantly biased, and the large number of all fractures available for analysis means that while small effects might have been missed, moderate to large effects have been reliably excluded.

The study also has several limitations. The study population is not representative of the global population, so direct generalizability of the results to other populations and nations cannot be guaranteed. The constancy of the findings across participant subgroups does, however, provide some reassurance about the likely applicability of the findings beyond the trial population. This was a secondary analysis of data therefore these investigations need to be interpreted in light of their post hoc nature and how they fit within the broader evidence base. Fracture was a secondary outcome, and event data were obtained mostly through hospitalization records, excluding emergency and outpatient data, so some cases of fracture that were mild in nature may be missed. There is likely to be misclassifications of some fracture events, though in general, fractures are more likely to be missed from a hospital record than classified as another form of event. Misclassification of types of fractures is more likely and any conclusions based on fracture causation are more uncertain. Both misclassification and omission may reduce power to detect an effect and create bias towards the null hypothesis. Current analyses were unable to account for whether the fracture was due to high-energy trauma or force from a chance accident or hypotensive fall or due to a fall secondary to low BMD and/or osteoporosis in which consumption of sodium through regular salt may be a contributing factor. However, in the present study, analyses of vertebral fractures which are likely caused by fragility and non-vertebral fractures which most are likely caused by trauma indicate that salt substitute has no detectable effect on fracture or contributes to any cause of fracture.

Conclusions

These analyses identify no detectable effect of salt substitute on fracture risk and provide further reassurance about the safety of salt substitute use for cardiovascular prevention among individuals at high risk of fracture. While the data cannot exclude any harmful effect or any protective effect of salt substitute on the incidence of fracture, the findings indicate that any real effect is likely to be small. Concerns about fracture risk should not be a clinical or public health contraindication to the use of salt substitute.

Supplementary Information

Additional file 1: Identification of all fractures.

Abbreviations

SSaSS Salt Substitute and Stroke Study

BMD Bone mineral density 

CVD Cardiovascular disease

HR Hazard ratio

BMI Body mass index

DBP Diastolic blood pressure

SBP Systolic blood pressure

Acknowledgements

The authors would like to thank Jiangsu Sinokone Technology Co Ltd for providing salt substitutes free of charge in years 3 and 4 of the study. Thanks are also extended to all participants and local collaborators.

Authors' contributions

FW, YZ1, MT, and BN contributed to the concept and design of the study. FW, YZ1, YZ2, BZ, ZL, JS, YY, MT, MY, LH, HS, and KRK contributed to the acquisition, analysis, and interpretation of data. LH, HS, and KRK had full access to all the data and take responsibility for the integrity of the data and the accuracy of the data analysis. FW, YP, LH, BN, and KRK drafted the manuscript. All authors contributed to the final version of the manuscript and accept responsibility for submitting for publication. All authors read and approved the final manuscript. YZ1, MT, and BN obtained funding to support this study.

Funding

This study was funded by grants APP1049417 and APP1164206 from the Australian National Health and Medical Research Council and funded by grants 2021BEG02026 from the Key Research and Development Program of Ningxia Hui Autonomous Region. The funders of the study had no role in design, data collection, data analysis, data interpretation, or writing of the report.

Availability of data and materials

Complete deidentified patient data set collected for the study can be requested from professor Bruce Neal, email: bneal@georgeinstitute.org.au. Once the data access is granted, data can be accessed via a secured environment hosted by The George Institute for Global Health, China, in line with China’s Cyber Security Law. The SSASS protocol and statistical analysis plan are available from https://osf.io/2u9jm/.

Declarations

Ethics approval and consent to participate

The ethics committees of Peking University Health Science Center in China (IRB00001052-13069) and the University of Sydney in Australia (2013/888) reviewed and approved the study, and all participants willingly provided written informed consent.

Consent for publication

All authors critically revised drafts of the manuscript and approved the final version.

Competing interests

The authors declare no competing interests.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References

1. Partridge L Deelen J Slagboom Pe Facing up to the global challenges of ageing Nature 2018 561 45 56 10.1038/s41586-018-0457-8 30185958
Partridge L, Deelen J, Slagboom Pe. Facing up to the global challenges of ageing. Nature. 2018;561:45–56.30185958 10.1038/s41586-018-0457-8
2. Compston JE McClung MR Leslie WD Osteoporosis Lancet 2019 393 10169 364 376 10.1016/S0140-6736(18)32112-3 30696576
Compston JE, McClung MR, Leslie WD. Osteoporosis. Lancet. 2019;393(10169):364–76.30696576 10.1016/S0140-6736(18)32112-3
3. GBD 2019 Fracture Collaborators Global, regional, and national burden of bone fractures in 204 countries and territories, 1990–2019: a systematic analysis from the Global Burden of Disease Study 2019 Lancet Healthy Longev 2021 2 9 e580 e92 10.1016/S2666-7568(21)00172-0 34723233
GBD 2019 Fracture Collaborators. Global, regional, and national burden of bone fractures in 204 countries and territories, 1990–2019: a systematic analysis from the Global Burden of Disease Study 2019. Lancet Healthy Longev. 2021;2(9):e580–92.34723233 10.1016/S2666-7568(21)00172-0
4. Polinder S, Haagsma J, Panneman M, Scholten A, Brugmans M, E. VB. The economic burden of injury: health care and productivity costs of injuries in the Netherlands. Accid Anal Prev. 2016;93:92–100.
5. Borgström F Karlsson L Ortsäter G Norton N Halbout P Cooper C Fragility fractures in Europe: burden, management and opportunities Arch Osteoporos 2020 15 1 59 10.1007/s11657-020-0706-y 32306163
Borgström F, Karlsson L, Ortsäter G, Norton N, Halbout P, Cooper C, et al. Fragility fractures in Europe: burden, management and opportunities. Arch Osteoporos. 2020;15(1):59.32306163 10.1007/s11657-020-0706-y
6. Veronese N Stubbs B Crepaldi G Solmi M Cooper C Harvey NC Relationship between low bone mineral density and fractures with incident cardiovascular disease: a systematic review and meta-analysis J Bone Miner Res 2017 32 5 1126 1135 10.1002/jbmr.3089 28138982
Veronese N, Stubbs B, Crepaldi G, Solmi M, Cooper C, Harvey NC, et al. Relationship between low bone mineral density and fractures with incident cardiovascular disease: a systematic review and meta-analysis. J Bone Miner Res. 2017;32(5):1126–35.28138982 10.1002/jbmr.3089
7. Watts NB Manson JE Osteoporosis and fracture risk evaluation and management: shared decision making in clinical practice JAMA 2017 317 3 253 254 10.1001/jama.2016.19087 27942715
Watts NB, Manson JE. Osteoporosis and fracture risk evaluation and management: shared decision making in clinical practice. JAMA. 2017;317(3):253–4.27942715 10.1001/jama.2016.19087
8. Cohen AJ Roe FJ Review of risk factors for osteoporosis with particular reference to a possible aetiological role of dietary salt Food Chem Toxicol 2000 38 2–3 237 253 10.1016/S0278-6915(99)00145-3 10717363
Cohen AJ, Roe FJ. Review of risk factors for osteoporosis with particular reference to a possible aetiological role of dietary salt. Food Chem Toxicol. 2000;38(2–3):237–53.10717363 10.1016/S0278-6915(99)00145-3
9. Heaney RP Bone health Am J Clin Nutr 2007 85 1 300S S303 10.1093/ajcn/85.1.300S 17209214
Heaney RP. Bone health. Am J Clin Nutr. 2007;85(1):300S–S303.17209214 10.1093/ajcn/85.1.300S
10. Harrington M Cashman KD High salt intake appears to increase bone resorption in postmenopausal women but high potassium intake ameliorates this adverse effect Nutr Rev 2003 61 5 Pt 1 179 183 12822707
Harrington M, Cashman KD. High salt intake appears to increase bone resorption in postmenopausal women but high potassium intake ameliorates this adverse effect. Nutr Rev. 2003;61(5 Pt 1):179–83.12822707
11. Devine A Criddle RA Dick IM Kerr DA Prince RL A longitudinal study of the effect of sodium and calcium intakes on regional bone density in postmenopausal women Am J Clin Nutr 1995 62 4 740 745 10.1093/ajcn/62.4.740 7572702
Devine A, Criddle RA, Dick IM, Kerr DA, Prince RL. A longitudinal study of the effect of sodium and calcium intakes on regional bone density in postmenopausal women. Am J Clin Nutr. 1995;62(4):740–5.7572702 10.1093/ajcn/62.4.740
12. Dar HY Singh A Shukla P Anupam R Mondal RK Mishra PK High dietary salt intake correlates with modulated Th17-Treg cell balance resulting in enhanced bone loss and impaired bone-microarchitecture in male mice Sci Rep 2018 8 1 2503 10.1038/s41598-018-20896-y 29410520
Dar HY, Singh A, Shukla P, Anupam R, Mondal RK, Mishra PK, et al. High dietary salt intake correlates with modulated Th17-Treg cell balance resulting in enhanced bone loss and impaired bone-microarchitecture in male mice. Sci Rep. 2018;8(1):2503.29410520 10.1038/s41598-018-20896-y
13. Bergstrom WH Wm W Bone as a sodium and potassium reservoir J Clin Invest 1954 33 6 867 873 10.1172/JCI102959 13163179
Bergstrom WH, Wm W. Bone as a sodium and potassium reservoir. J Clin Invest. 1954;33(6):867–73.13163179 10.1172/JCI102959
14. Barsony J Sugimura Y Verbalis JG Osteoclast Response to Low extracellular sodium and the mechanism of hyponatremia-induced bone loss J Biol Chem 2011 286 12 10864 10875 10.1074/jbc.M110.155002 21135109
Barsony J, Sugimura Y, Verbalis JG. Osteoclast Response to Low extracellular sodium and the mechanism of hyponatremia-induced bone loss. J Biol Chem. 2011;286(12):10864–75.21135109 10.1074/jbc.M110.155002
15. Verbalis JG Barsony J Sugimura Y Tian Y Adams DJ Carter EA Hyponatremia-induced osteoporosis J Bone Miner Res 2010 25 3 554 563 10.1359/jbmr.090827 19751154
Verbalis JG, Barsony J, Sugimura Y, Tian Y, Adams DJ, Carter EA, et al. Hyponatremia-induced osteoporosis. J Bone Miner Res. 2010;25(3):554–63.19751154 10.1359/jbmr.090827
16. Tucker KL Hannan MT Chen H Cupples LA Wilson PW Kiel DP Potassium, magnesium, and fruit and vegetable intakes are associated with greater bone mineral density in elderly men and women Am J Clin Nutr 1999 69 4 727 736 10.1093/ajcn/69.4.727 10197575
Tucker KL, Hannan MT, Chen H, Cupples LA, Wilson PW, Kiel DP. Potassium, magnesium, and fruit and vegetable intakes are associated with greater bone mineral density in elderly men and women. Am J Clin Nutr. 1999;69(4):727–36.10197575 10.1093/ajcn/69.4.727
17. Ha J, Kim SA, Lim K, S. S. The association of potassium intake with bone mineral density and the prevalence of osteoporosis among older Korean adults. Nutr Res Pract. 2020;14(1):55–61.
18. Zhu K Devine A Prince RL The effects of high potassium consumption on bone mineral density in a prospective cohort study of elderly postmenopausal women Osteoporos Int 2009 20 2 335 340 10.1007/s00198-008-0666-3 18575949
Zhu K, Devine A, Prince RL. The effects of high potassium consumption on bone mineral density in a prospective cohort study of elderly postmenopausal women. Osteoporos Int. 2009;20(2):335–40.18575949 10.1007/s00198-008-0666-3
19. Hannon MJ Verbalis JG Sodium homeostasis and bone Curr Opin Nephrol Hypertens 2014 23 4 370 376 10.1097/01.mnh.0000447022.51722.f4 24840297
Hannon MJ, Verbalis JG. Sodium homeostasis and bone. Curr Opin Nephrol Hypertens. 2014;23(4):370–6.24840297 10.1097/01.mnh.0000447022.51722.f4
20. Bromfield SG Ngameni CA Colantonio LD Bowling CB Shimbo D Reynolds K Blood pressure, antihypertensive polypharmacy, frailty, and risk for serious fall injuries among older treated adults with hypertension Hypertension 2017 70 2 259 266 10.1161/HYPERTENSIONAHA.116.09390 28652459
Bromfield SG, Ngameni CA, Colantonio LD, Bowling CB, Shimbo D, Reynolds K, et al. Blood pressure, antihypertensive polypharmacy, frailty, and risk for serious fall injuries among older treated adults with hypertension. Hypertension. 2017;70(2):259–66.28652459 10.1161/HYPERTENSIONAHA.116.09390
21. Velliou M, Sanidas E, Zografou A, Papadopoulos D, Dalianis N, J. B. Antihypertensive drugs and risk of bone fractures. Drugs Aging. 2022;39(7):551–7.
22. Kwon SJ Ha YC Park Y High dietary sodium intake is associated with low bone mass in postmenopausal women: Korea National Health and Nutrition Examination Survey, 2008–2011 Osteoporos Int 2017 28 4 1445 1452 10.1007/s00198-017-3904-8 28074252
Kwon SJ, Ha YC, Park Y. High dietary sodium intake is associated with low bone mass in postmenopausal women: Korea National Health and Nutrition Examination Survey, 2008–2011. Osteoporos Int. 2017;28(4):1445–52.28074252 10.1007/s00198-017-3904-8
23. Neal B Tian M Li N Elliott P Yan LL Labarthe DR Rationale, design, and baseline characteristics of the Salt Substitute and Stroke Study (SSaSS)-a large-scale cluster randomized controlled trial Am Heart J 2017 188 109 117 10.1016/j.ahj.2017.02.033 28577665
Neal B, Tian M, Li N, Elliott P, Yan LL, Labarthe DR, et al. Rationale, design, and baseline characteristics of the Salt Substitute and Stroke Study (SSaSS)-a large-scale cluster randomized controlled trial. Am Heart J. 2017;188:109–17.28577665 10.1016/j.ahj.2017.02.033
24. Neal B Wu Y Feng X Zhang R Zhang Y Shi J Effect of salt substitution on cardiovascular events and death N Engl J Med 2021 385 12 1067 1077 10.1056/NEJMoa2105675 34459569
Neal B, Wu Y, Feng X, Zhang R, Zhang Y, Shi J, et al. Effect of salt substitution on cardiovascular events and death. N Engl J Med. 2021;385(12):1067–77.34459569 10.1056/NEJMoa2105675
25. Yin X Rodgers A Perkovic A Huang L Li KC Yu J Effects of salt substitutes on clinical outcomes: a systematic review and meta-analysis Heart 2022 108 20 1608 1615 10.1136/heartjnl-2022-321332 35945000
Yin X, Rodgers A, Perkovic A, Huang L, Li KC, Yu J, et al. Effects of salt substitutes on clinical outcomes: a systematic review and meta-analysis. Heart. 2022;108(20):1608–15.35945000 10.1136/heartjnl-2022-321332
26. Park Y Kwon SJ Ha YC Association between urinary sodium excretion and bone health in male and female adults Ann Nutr Metab 2016 68 3 189 196 10.1159/000444536 26967579
Park Y, Kwon SJ, Ha YC. Association between urinary sodium excretion and bone health in male and female adults. Ann Nutr Metab. 2016;68(3):189–96.26967579 10.1159/000444536
27. Jones G Beard T Parameswaran V Greenaway T Witt Rv A population-based study of the relationship between salt intake, bone resorption and bone mass Eur J Clin Nutr 1997 51 8 561 5 10.1038/sj.ejcn.1600452 11248883
Jones G, Beard T, Parameswaran V, Greenaway T, Witt Rv. A population-based study of the relationship between salt intake, bone resorption and bone mass. Eur J Clin Nutr. 1997;51(8):561–5.11248883 10.1038/sj.ejcn.1600452
28. Carbone L Johnson KC Huang Y Pettinger M Thomas F Cauley J Sodium intake and osteoporosis. Findings from the Women’s Health Initiative J Clin Endocrinol Metab 2016 101 4 1414 21 10.1210/jc.2015-4017 26863423
Carbone L, Johnson KC, Huang Y, Pettinger M, Thomas F, Cauley J, et al. Sodium intake and osteoporosis. Findings from the Women’s Health Initiative. J Clin Endocrinol Metab. 2016;101(4):1414–21.26863423 10.1210/jc.2015-4017
29. Nieves JW Melsop K Curtis M Kelsey JL Bachrach LK Greendale G Nutritional factors that influence change in bone density and stress fracture risk among young female cross-country runners PM R 2010 2 8 740 750 10.1016/j.pmrj.2010.04.020 20709302
Nieves JW, Melsop K, Curtis M, Kelsey JL, Bachrach LK, Greendale G, et al. Nutritional factors that influence change in bone density and stress fracture risk among young female cross-country runners. PM R. 2010;2(8):740–50.20709302 10.1016/j.pmrj.2010.04.020
30. Jehle S Hulter HN Krapf R Effect of potassium citrate on bone density, microarchitecture, and fracture risk in healthy older adults without osteoporosis: a randomized controlled trial J Clin Endocrinol Metab 2013 98 1 207 2017 10.1210/jc.2012-3099 23162100
Jehle S, Hulter HN, Krapf R. Effect of potassium citrate on bone density, microarchitecture, and fracture risk in healthy older adults without osteoporosis: a randomized controlled trial. J Clin Endocrinol Metab. 2013;98(1):207–2017.23162100 10.1210/jc.2012-3099
31. Fatahi S Namazi N Larijani B Azadbakht L The association of dietary and urinary sodium with bone mineral density and risk of osteoporosis: a systematic review and meta-analysis J Am Coll Nutr 2018 37 6 522 532 10.1080/07315724.2018.1431161 29617220
Fatahi S, Namazi N, Larijani B, Azadbakht L. The association of dietary and urinary sodium with bone mineral density and risk of osteoporosis: a systematic review and meta-analysis. J Am Coll Nutr. 2018;37(6):522–32.29617220 10.1080/07315724.2018.1431161
32. Greendale G Barrett-Connor E Edelstein S Ingles S Haile R Dietary sodium and bone mineral density: results of a 16-year follow-up study J Am Geriatr Soc 1994 42 10 1050 1055 10.1111/j.1532-5415.1994.tb06208.x 7930328
Greendale G, Barrett-Connor E, Edelstein S, Ingles S, Haile R. Dietary sodium and bone mineral density: results of a 16-year follow-up study. J Am Geriatr Soc. 1994;42(10):1050–5.7930328 10.1111/j.1532-5415.1994.tb06208.x
33. Vestergaard P Rejnmark L Mosekilde L Hypertension is a risk factor for fractures Calcif Tissue Int 2009 84 2 103 111 10.1007/s00223-008-9198-2 19067019
Vestergaard P, Rejnmark L, Mosekilde L. Hypertension is a risk factor for fractures. Calcif Tissue Int. 2009;84(2):103–11.19067019 10.1007/s00223-008-9198-2
34. Yang S Nguyen ND Center JR Eisman JA Nguyen TV Association between hypertension and fragility fracture: a longitudinal study Osteoporos Int 2014 25 1 97 103 10.1007/s00198-013-2457-8 23892585
Yang S, Nguyen ND, Center JR, Eisman JA, Nguyen TV. Association between hypertension and fragility fracture: a longitudinal study. Osteoporos Int. 2014;25(1):97–103.23892585 10.1007/s00198-013-2457-8
35. Bergh C Wennergren D Möller M Brisby H Fracture incidence in adults in relation to age and gender: A study of 27,169 fractures in the Swedish Fracture Register in a well-defined catchment area PLoS ONE 2020 15 12 e0244291 10.1371/journal.pone.0244291 33347485
Bergh C, Wennergren D, Möller M, Brisby H. Fracture incidence in adults in relation to age and gender: A study of 27,169 fractures in the Swedish Fracture Register in a well-defined catchment area. PLoS ONE. 2020;15(12): e0244291.33347485 10.1371/journal.pone.0244291
36. Malgorzata KU Ewa MS Malnutrition in older adults-effect on falls and fractures: a narrative review Nutrients 2022 14 15 3123 10.3390/nu14153123 35956299
Malgorzata KU, Ewa MS. Malnutrition in older adults-effect on falls and fractures: a narrative review. Nutrients. 2022;14(15):3123.35956299 10.3390/nu14153123
37. Dent E, Olivia R L Wright, Jean Woo, Hoogendijk EO. Malnutrition in older adults Lancet. 2023;401(10380):951–66.
38. DiNicolantonio JJ Mehta V Zaman SB O’Keefe JH Not salt but sugar as aetiological in osteoporosis: a review Mo Med 2018 115 3 247 252 30228731
DiNicolantonio JJ, Mehta V, Zaman SB, O’Keefe JH. Not salt but sugar as aetiological in osteoporosis: a review. Mo Med. 2018;115(3):247–52.30228731
39. Hong S Choi JW Park JS Lee CH The association between dietary sodium intake and osteoporosis Sci Rep 2022 12 1 14594 10.1038/s41598-022-18830-4 36028560
Hong S, Choi JW, Park JS, Lee CH. The association between dietary sodium intake and osteoporosis. Sci Rep. 2022;12(1):14594.36028560 10.1038/s41598-022-18830-4
