Negative symptoms of schizophrenia may improve over time, says meta-analysis


The negative symptoms experienced by people with a diagnosis of schizophrenia typically involve an absence or decline of normal functioning and behaviour, including anhedonia, amotivation, asociality and alogia.

While the course of negative symptoms is typically heterogeneous, many treatments for schizophrenia have been shown to provide limited benefit for negative symptoms.

To account for this heterogeneity, classifications of secondary and primary negative symptoms have been developed:

  • Secondary negative symptoms are defined as being typically responsive to the treatment of the underlying cause, for example medication side-effects or hospitalization (W.T. Carpenter et al, 1985)
  • However, primary negative symptoms are ongoing at varying intensities over long periods of time even in the absence of other symptoms of schizophrenia (W.T. Carpenter et al 1985)

The identification of effective treatments for primary negative symptoms which persist during periods of clinical stability is required to meet this therapeutic need.

The objective of this meta-analysis was to investigate how negative symptoms change over time in schizophrenia outpatients, identify any trends in how symptoms change, and explore the impact of factors known to induce secondary negative symptoms.

People experiencing the negative symptoms of schizophrenia may appear emotionless, flat and apathetic.

People experiencing the negative symptoms of schizophrenia may appear emotionless, flat and apathetic.


An electronic search was conducted which included the MedLine, PsycINFO, EMBASE and CENTRAL databases dating back to 1962. The search used three parameters:

  • Diagnosis,
  • Negative symptoms, and
  • Whether an indicator study occurred over at least two time-points

The electronic search was accompanied by a hand search (including reference lists) and authors were contacted where necessary. Screening was duplicated and any discrepancies were resolved without adjudication. The criteria for exclusion included:

  • Clearly not relevant
  • Studies involving less than 50 participants
  • A study was excluded if it didn’t include repeated assessments of negative symptoms at set time-points
  • Studies including participants who were not exclusively schizophrenia patients
  • Studies including children or older adults, and
  • Any study duration which was either to short (under 10 weeks) or to0 long (over 3 years)

Studies were included if the symptoms measured used a validated scale and if there was at least one time-point where participants were outpatients. The first stage of analysis used the Der Simonion and Laird random-effects model and included five intervention types:

  1. Treatment as usual (TAU)/control
  2. Non-drug intervention
  3. Drug: second generation antipsychotic
  4. Drug: adjunctive medications
  5. Drug: first generation antipsychotic

The second stage of analysis involved planned univariate and multivariate meta-regressions.


The search identified 9,480 articles which were screened to identify 41 relevant articles containing 5,944 participants, which were included in the final analysis. The 41 studies included in the meta-analysis used the following scales:

  • Positive and Negative Syndrome Scale (PANSS): 23 studies
  • Scale to Assess Negative Symptoms (SANS): 14 studies
  • Brief Psychiatric Rating Scale (BPRS): 4 studies

All five intervention types reported a significant reduction in negative symptoms between baseline and follow-up assessment stages including:

  • Second generation antipsychotics (ES = 1.09, 95% CI 0.86 to 1.32, I2 = 95.5%)
  • Adjunctive medication (ES = 0.97, 95% CI 0.68 to 1.26, I2 = 91.7%)
  • TAU/placebo (ES = 0.33, 95% CI 0.17 to 0.49, I2 = 91.8%)

The significant effect size and high heterogeneity noted for negative symptom change was examined using the planned univariate meta-regression and was associated with the scale used, study duration, intervention type and minimum negative symptom inclusion criteria. The scale used and type of intervention received were also shown to be significant by the multivariate meta-regression. A significantly greater reduction in negative symptoms was reported for studies using SANS relative to those using PANSS:

  • SANS (ES = 1.02, 95% CI 0.77 to 1.28)
  • PANSS (ES = 0.66, 95% CI 0.56 to 0.77)

The scale used and intervention type accounted for 43.65% of the variance reported.

The authors also reported on changes in individual negative symptoms based on nine studies using PANSS and SANS. These studies demonstrated a significant reduction in affective blunting, alogia, avolition-apathy and anhedonia-asociality. The greatest symptom reduction reported was for avolition-apathy, and the smallest decrease noted was for alogia.

The study found that negative symptoms were significantly reduced in all treatment interventions, including in placebo and treatment as usual conditions.

The study found that negative symptoms were significantly reduced in all treatment interventions, including in placebo and treatment as usual conditions.


The authors concluded that:

Based on the available data of almost 6,000 outpatients, negative symptoms of schizophrenia do not tend to be stable or deteriorate, but are instead likely to improve over time. This finding offers a further critique of the historical argument which suggests schizophrenia is a disorder of continual decline and instead provides further support to the recovery model of schizophrenia.

Overall, these findings suggest that negative symptoms may not be as resistant to change as what has previously been assumed, and perhaps offer new hope to those who may experience such symptoms.


A broad inclusion criterion was used for this study resulting in high levels of clinical and methodological heterogeneity. However, the authors note that the findings show a consistent trend of negative symptom improvement and describe their findings as ‘relatively robust’. They also note that due to the within-group design the effect sizes obtained cannot be used to assess the effectiveness of any one treatment.

The univariate meta-regression identified a number of factors responsible for the high levels of heterogeneity including the scale used and intervention type. The same two factors were identified by the multivariate meta-regression and accounted for 44% of the variance identified in this study. It should also be noted that the authors did not include an examination of publication bias in their meta-regression analysis.

The authors also state that their findings include no evidence relating specifically to either primary or secondary negative symptoms, only the severity of the negative symptoms experienced. However, attempts were made to address this concern, for example, by using studies where participants experienced chronic symptoms and who were described as treatment-resistant etc.

Subsequently, the authors offer a cautious conclusion suggesting their findings reflect an ongoing trend of improvement in negative symptoms which lends support to the recovery model of schizophrenia. To provide more conclusive results the authors recommended that future studies are conducted with inclusion criteria specific to persistent negative symptoms which include appropriate controls for secondary negative symptoms.

The study findings support the recovery model of schizophrenia.

The study findings support the recovery model of schizophrenia.


Primary paper

Savill, C. Banks, H. Khanom and S. Priebe (2014). Do negative symptoms of schizophrenia change over time? A meta-analysis of longitudinal data (PDF). Psychological Medicine, 1-15. doi:10.1017/S0033291714002712

Other references

Blanchard JJ, Kring AM, Horan WP, Gur R (2011). Toward the next generation of negative symptom assessments: the collaboration to advance negative symptom assessment in schizophrenia. Schizophrenia Bulletin 37, 291–299.

Carpenter WT, Heinrichs DW, Alphs LD (1985). Treatment of negative symptoms. Schizophrenia Bulletin 11, 440–452.

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+